CICTP 2017: Transportation Reform and Change—Equity, Inclusiveness, Sharing, and Innovation

Universities keep a close connection with society, incurring an increasing frequency of vehicle parking on campuses. Tongji University, located in the vicinity of Wujiao Square, which is one of the main CBDs of Shanghai, attracts numerous vehicles commuting or temporarily visiting. With the number of licenses released by the university authority mounting, the parking demand has already spilled over, which leads to severe violation behaviors on parking regulations. The analysis of campus parking characteristics is the prerequisite for proposing parking strategies. Through interpreting field observation data and intelligent vehicle plate detection system data at all university entrances, the research has a sophisticated analysis on the parking characteristics of Siping Campus. Analysis implied the imbalance of campus parking supply and demand, uneven distribution of parking, the rate of high illegally parking, and the presence of a stable peak of weekday parking time, and the amount of corresponding parking.

Many researches on resident activity range come from traditional data collection methods such as artificial or telephone surveys, but those approaches are difficult to update, and the results are limited to a certain point in time. With the progress of mobile-phone positioning technology, mobile-phone positioning data has become a kind of economic and efficient spatiotemporal data source. In this paper, we propose a method of calculating activity-radius to measure the activity range of a mobile-phone user, by using mobile-phone billing data in the city of Shenzhen for a seven-day period. We find that the distribution of activity-radius is well approximated by a truncated power-law. The accuracy and reliability of the activity-radius based on billing data has also been analyzed. This method could provide a new idea for the real-time observation of resident activity range and could also lead to more informed decision making in both city planning and transport system planning.

With the explosion of automobile information, the industry of internet of vehicles (IOV) is rising rapidly, and the vehicle number in IOV is growing exponentially; however, this results in problems like how to store and analyze the massive dynamic data of IOV. Considering the IOV’s characteristics like different data types, uncertain sampling period, and huge size, the study presented in the paper designs an efficient data saving and analysis storage system based on B+ tree technology. The storage system can customize the file type, file path, search keywords, saving data fields, and according to the data features in IOV, the storage system can customize B+ tree’s index key dimension to enhance the search efficiency. Theoretical analysis and practical test results show that the storage system can solve the problems of traditional database in IOV application, such as large physical space, slow querying speed, the difficulty of data backup, and frequent data loss.

Understanding the daily temporal pattern of bus trips is very significant for both operators and passengers. However, few studies have focused on popular areas like the Tian’anmen area. To fill this gap, this paper empirically analyzes the temporal patterns of human bus trips in the Tian’anmen area using smart card data. In addition, we proposed crowding indicators at three levels, the vehicle, link, and network levels, to make a quantitative analysis of bus networks. At the network level, we consider the importance of links. The entropy method is used to assign weights. The results indicate (a) transit ridership increases on May Day on some stops; (b) the temporal patterns of bus travel in weekdays are similar and stable, and there is a significant difference in holidays; (c) the crowding indicators we propose are appropriate. The empirical analysis and the crowding measure give us a deep and comprehensive view of Tian’anmen area bus trips.

Mobile communication network positioning (MCNP) technology was applied to study the off-campus travel patterns of college students. MCNP data was collected by smartphone app from 97 volunteers on Siping Campus and Jiading Campus of Tongji University for a period of 35 days. The students’ travel characteristics are described by 3 measures, including daily activity frequency, number of distinct activity points, and travel distance from the campus. For students on Jiading Campus, their average daily activity frequency and travel distance from the campus are higher/larger than those of students on Siping Campus. The students’ travel often happened during lunch and supper. The activity scopes of students on Jiading Campus are much larger than those of students on Siping Campus. Most students’ travel depended on bus and metro. Finally, we studied their travel predictability by entropy method. We found that major and grade have a certain impact on their travel predictability.

Illegal parking is one of the major causes of traffic congestion. Among many ways to mitigate illegal parking, a high level of direct traffic enforcement has been shown to have the strongest impact. Traffic enforcement must invest substantial human resources in patrolling and issuing parking tickets for illegally parked cards. The work load is overwhelming and efficiency is low since the illegal parking occurs almost everywhere in urban areas. This study tries to help traffic related officers ease their work load but minimize illegal parking by implementing social media networking. Social media in China has a significant number of users which covers nearly every urban area so that the service users can be public police officers to make traffic flow better in the city. The aim of this research is to develop the smartphone-based application and see its possible potential. The first part of this paper will describe the WeChat based application, which collects information on illegal parking cars from the public. It includes user interfaces, an automatic warning system, a parking navigation system, and data visualization. The second part demonstrates the process of the practical operation in Suzhou, China. Finally, the authors conclude the advantages and further development of this application. In addition, features of the illegal parking problem and parking planning problem in Suzhou are discussed.

The accuracy and efficiency of TDOD online prediction have significant impacts on the application of a DTA model. This paper aims to propose the online prediction of time-dependent OD flow for the freeway network based on freeway toll collection data. The number of freeway toll stations in Jiangsu Province is 316, and the total travel demand is around 1 million vehicles per day. The requirements and characteristics of the time-dependent OD for the online application is discussed first. Then, the prediction of time-dependent OD flows in previous and future time periods are proposed respectively. Finally, the actual time-dependent OD flows are utilized to validate the prediction results. The validation indicates that the proposed model can predict the time-dependent OD flow with high accuracy and efficiency, contributing to the online application of a DTA model.

The primary objective of this study was to conduct a spatial analysis for the usage of ride-sourcing services. This study collected the ride-sourcing trip data from Uber in New York City, as well as the road network attributes and the socio-demographic characteristics. The collected data were further aggregated into 167 ZIP code tabulation areas (ZCTA). The geographically weighted regression models were developed to establish the relationships between ride-sourcing usage and various contributing factors on weekday and weekend, respectively. The results suggested that the number of population with bachelors’ degree or higher, the number of parking spaces, the ratio of floor area allocated for commercial use, and the subway accessibility in each ZCTA were positively correlated with the Uber usage. The number of unemployment population and the median household income in each ZCTA were found to be negatively correlated with the Uber usage. The results of the present study could provide useful insights for predicting the demand of zonal-level ride-sourcing services and developing effective policy initiatives for pricing and regulation of ride-sourcing industry.

Traffic state identification and prediction are playing an important role in intelligent transportation systems (ITS) research, especially in the era of big data when a variety of big data resources are available. In this paper, two kinds of real-world collected spatiotemporal datasets are analyzed. These data sets were captured by microwave and video detectors from Haining City of China. First of all, we propose a novel method which transfer one dimensional time series data into two dimensional pattern for missing data compensation and data fusion. After that, we build a special kind of recurrent neural network (RNN) which is usually called long short term memory (LSTM) networks to train the time series traffic flow data sets. We compare the prediction accuracy between the imputed data and the original data. Experiments show that the prediction error has lower RMSE score by using proposed novel method for imputation than original data without using this method.

A non-parametric kernel regression algorithm is developed to analyze duration of traffic information based on traffic information historical data and converted traffic condition value. This non-parametric kernel regression algorithm takes thorough advantage of historical data to analyze the duration of traffic information. Empirical analysis is conducted based on traffic guidance information data from Sept. 1, 2008, to June 30, 2012, released by Beijing variable message signs (VMS) system and converted traffic condition. Empirical analysis shows that there is a gap between predicted and real duration. Limited by current data resources, this work can only be an attempt to analyze duration of traffic information.

Conventional travel behavior research relies on questionnaires to understand individual travel patterns. However, these studies consume considerable manpower and resources. In a connected vehicle environment, on-board diagnostic (OBD) devices can record engine status, trajectories, acceleration, and fuel consumption of a vehicle to capture long-term variability in driver behaviors. In this study, the heterogeneous travel patterns of drivers are modeled using 2-month OBD data. An algorithm called clustering by fast search and find of density peaks is employed to classify drivers into long-distance and occasional, high-frequency, and regular travelers. The average travel distance, travel days, and first and last departure time records are considered in the procedure. A multi-dimensional discrete hidden Markov model is used to predict the category of any driver based on their historical travel behavior. This study provides useful data sources for activity-based modeling and also demonstrates the potential of vehicle OBD data for developing targeted online services.

This paper described a demonstration pilot project that allows the determination of user-focused road condition indicators and road safety concerns by developing a road user oriented smartphone application which calls built-in sensors and extracting information from big data collected by road inspections and other common drivers. This paper investigated the prediction of pavement quality rating based on international roughness index with the Kalman filtering and ordinal logistics regression method which applied to analyze the accelerometer and GPS data. An online map-matching algorithm based on the hidden Markov model was used to relate GPS data to roadways. The map-matching results as well as pavement quality rating indicating pavement condition were visualized in a real-time map. The crowdsourcing-based approach and visual map display technology provides wider coverage of country road networks at frequent intervals and collects uniform data to support strategic and network asset management decision making.

Short-term traffic speed prediction plays a crucial role in intelligent transportation systems, e.g., vehicle navigation. Short-term traffic speed forecasting used to be analyzed only considering temporal factors. In this paper, we take both temporal and spatial factors into account. Features of the upstream and downstream locations are considered to establish the random forest (RF) and support vector regression (SVR) models. In order to improve the prediction accuracy, two brand new variables (i.e., absolute congestion and relative congestion) calculated from raw data are proposed. The RF is built to test the new variables’ influence on the prediction accuracy. Traffic data extracted from the California Freeway Performance Measurement System (PeMS) is used to build and test the speed prediction models. It is verified that both upstream and downstream speed variables are able to improve model accuracy, and the two proposed new features are helpful to further improving the accuracy.

Processing inductance loop detector data for desirable outputs is essential for traffic management and control. Because there are two kinds of correlations in traffic flow data, namely, spatial correlations and temporal correlations, we propose a loop data processing method based on both spatial and temporal dimensions. The performance of the proposed method is evaluated with the data collected on an urban expressway segment in Shanghai. In addition, the comparisons between time series model (ARMA), spatial multivariate linear model, and comprehensive model based on spatio-temporal characteristics are conducted. The results indicate that the last model can fit the data best.

The transportation integration of in urban areas is the main direction of urban development. The regions in Shanghai have strong connection with the Yangtze River Delta mega-city group. This paper analyzes the activity space of nonresidents by using mobile phone data. An identification method has been used to identify nonresidents traveled through the Hongqiao transportation hub for air flights or high-speed trains. After identifying active areas of nonresidents, POI data is introduced to evaluate the built environment to quantify activity space by using the factor analysis method. Three activity space indicators have been proposed to describe the activity space. By using a two-step clustering method, nonresidents are classified into four clusters to better describe their activity space. The result shows that the activity space of nonresidents was distributed in a way that has strong relationship with Metro Line 2, which can strongly support the further connections between Shanghai and the Yangtze River Delta mega-city group.

In this paper, the multiple discrete-continuous extreme value (MDCEV) modeling framework is employed to model daily time allocation behavior in Xiaoshan District in the City of Hangzhou, China. The data for model development are collected from a household survey recently conducted in its urban area based on an advanced survey system using portable tablet computer with web-based map service. The models are developed to identify the constraint from mandatory work schedules. In the worker model, it is found that the total work time, work schedule, commuting time, commuting travel mode significantly affect daily time allocations on other non-mandatory activities. Other impact factors include worker's age, gender, education level, household income, and population density of residential area. In the non-worker model, influential factors include non-workers' age, gender, household size, household income, permanent residency, and population density of residential area. It is found that shifting the work schedule one hour earlier from 8:00 - 17:00 to 7:00 - 16:00 will allow workers to allocate extra about 40 minutes on out-of-home activities but cut the same amount of time on in-home activities. Lengthening (or shortening) daily work time by 1 hour will cut (or add) about 45 minutes on in-home activities and about 15 minutes on out-of-home activities.

A lot of attention has been paid on the differences of operation and service characteristics between traditional taxis and transportation network company (TNC) cars. Based on the multi-source data including taximeter data, GPS data of the traditional taxi, and TNC data in working days in September 2015, this paper proposes taxi operation characteristic index model and compares the operation characteristics from the views of operation intensity, and spatial and temporal distribution. The analysis results indicate that the operation intensity of TNC cars is 40% when compared to that of the traditional taxis in general. More than 50% of the TNC transactions are mainly concentrated during morning and evening peak hours and during the night. Spatial distribution shows that the TNC pick-up hotspots coincide with that of the traditional taxis in the downtown area. But, on the regions outside of the downtown where the taxi supply is not sufficient, TNCs form the significant replacement for the traditional taxi service. The balance coefficient (BC) in these regions could also be increased by 9% in peak hours. This paper provides an important support for the scientific management and service quality improvement of the taxi industry.

Taiwan freeway electronic toll collection system (ETC) is not only replacing manual tolling to high accuracy distance-based toll collection ways, but is collecting specific and timely traffic data. In this study, we utilized the open data from ETC, such as traffic flow information, travel time information, vehicle types, etc., in every 5 minutes for multiple years and used the support vector machine and support vector regression method to get the kernel function to predict the future travel time and future traffic flow. Furthermore, we applied these prediction values as important references to support the strategies decision in freeway road maintenance management system and perform in the freeway management system. Thus, with the help of the big data collection and utilization we could efficiently improve the freeway service performance and develop the freeway system into intelligent transportation system (ITS) in the near future.

Traffic congestion as a worldwide problem not only causes traffic delays but also causes a series of environmental problems. Evaluating traffic congestion impacts more efficiently, reasonably, and systematically becomes a fundamental issue. In this context, this paper strives to propose a comprehensive estimation model based on a novel parameter–“continuous congestion,” which is a comprehensive evaluation index by combining all characteristics of traffic congestion. The most important feature of this method is the formation-dissipation process of congestion that is used as the evaluation target. The method estimates full-scale congestion as a whole, breaks the restriction of road classifications and regions, and expands the comparable range of congestions. Finally, a case study is conducted on the West 2^nd Ring Road in Beijing.

The taxi is a common way of travelling in cities, which is an important supplement to other public travel modes. Thus, it is necessary to understand the characters of taxi operation and its passengers. The individual vehicle trajectory information has been obtained easily and economically because of the widely-used GPS devices, which makes mining the meaningful information behind the trajectory feasible and more convenient. In this paper, the taxi GPS data from Kunshan City, in China, is used to identify the pick-up and drop-off hotspots. Firstly, a procedure is proposed to extract the taxi trip trajectories from the vast amounts of GPS data. Then, with the help of data mining technology, the pick-up and drop-off hotspots are analyzed using the density-based spatial clustering method. It comes to a conclusion that the pick-up and drop-off hotspots can share a similar distribution pattern and overlap very well.

The rapid development of urbanization has led to the gradual emergence of more traffic problems emerging; traffic incidents continue to increase in number and scope, and safety has become a serious issue. With the steady increase of vehicles and the contradiction between the supply and demand of road transportation, efficient solutions to traffic congestion and safety are paramount; intelligent transportation systems are the key to improving urban traffic. Based on the intelligent transportation decision system, this paper briefly discusses how to use business intelligence to conduct big data mining and analysis using data warehouses and online analytical processing technology to realize the automation of data reports, analyze the causes of traffic accidents, make scientific predictions of possible accidents, improve road capacity, and enhance the intelligence and informatization of the transportation decision system.

This paper identifies traffic zone division and location of users' residence and work place in Songjiang District of Shanghai City based on mobile phone data, which can be used to analyze residents' commuting characteristics. Then, we proposed a novel way to discuss human activity space by means of anchor point theory. Classifying the important arrest points as anchor points helps locating the important points of residence in the living space as an anchor point. This paper described human activity space by convex polygons whose size represent the spatial range, length-width ratio represents the ductility. Meanwhile, we found that the size obeys an exponential distribution. This conclusion has very important practical application value for traffic and urban planning.

The analysis for vehicle travel characteristics has always been of great interest to transport authorities, since it has a significant impact on logistic and operational decisions. Moreover, vehicle clustering can improve market research through more targeted access to groups of interest and facilitate planning through better survey design. This paper directly clustered 392,856 vehicles in a week using k-means clustering algorithm based on license plate recognition (LPR) data obtained in Shenzhen, China. First, several corresponding temporal variables are applied in weekdays and weekends respectively to identify homogeneous clusters. Second, Davies Bouldin index (DBI) and silhouette coefficient (SC) are utilized to find the optimal cluster number. Finally, seven groups in weekdays and three in weekends are classified. Meanwhile, detailed analysis of the characteristics for each group in terms of temporal changes in cluster characteristics is presented.

Heavy rainfall, as a kind of adverse weather condition, can have an extremely negative impact on subway ridership. This paper proposed a comprehensive framework to analyze the rainfall impact on subway transit ridership. Firstly, rainfall events were detected from the precipitation data. Then, heterogeneous characteristics of subway transit trips, such as the trip duration, the functionality of subway stations, and the trip elasticity, were extracted from smart card data. Finally, ridership fluctuation coefficient was calculated to measure the rainfall-induced ridership fluctuation degree quantitatively. It figured out that heavy rainstorm could cause as much as forty percent of reduction on ridership. Meanwhile, different kinds of trips showed different levels of sensitivity to rainfall. Long or elastic trips were more vulnerable, and trips flowing to airports or tourist spots could be sharply cut down by rainfall events. The research results can help metro operators improve their service quality under certain forecasted weather conditions.

More and more attention has been given to understanding the uncertainty regarding the travel time. In this paper, travel time is studied by collecting video monitoring data on Beijing Jingtong Expressway. The data shows that vehicle travel time is influenced by departure time, bus lane, vehicle restriction, and the number of exits and entrances and anomaly events. This paper determines three travel time reliability evaluation models to study 15 min interval travel time reliability based on Florida Department of Transportation’s research results, Beijing urban road traffic operation evaluation system and the results of questionnaire survey of traveler. The results suggest there are significant differences among the reliabilities under different models. Last, on the basis of dynamic optimization method, traffic management department can optimize traffic operation management scheme and improve traffic operation quality. Results of this paper can provide reference to road planning, operation management, and information publication.

Compared to other types of cellular signaling data, cellular network recording data has several unique advantages on extracting travel characteristics, including comprehensive samples, high location accuracy, elaborate trajectory, and high data stability. With such advantageous attributes, a series of methods are proposed in order to extract valuable information. First, in order to build candidate default trajectory point set, Bowyer–Watson algorithm is applied to delimit Voronoi diagram of base station coverage. Second, a user’s trajectory is segmented based on motion state discriminant algorithms. This is done in which sample filtering and noise reduction are helpful for parameter extraction. Third, the MBR of trajectory point subset is introduced to improve conventional piecewise map-matching algorithms. Lastly, the proposed algorithm will be evaluated by verifying the accuracy of the parameters. Preliminary results show that the extracted parameters are acceptable for further traffic modeling application at both macro and micro levels and are expected to replace the traditional individual trip survey.

A better understanding of travel patterns and demands of transit users helps to evaluate current transit services and improve transit service. However, it’s challenging to identify passengers’ long-term travel characteristics and demands simultaneously through a single dataset. This paper investigates the classification of transit users and interprets passenger segmentation through the analysis of travel patterns and demands using combined smart card data and survey data. Transit users were grouped into segments based on the travel patterns revealed in smart card data, and the users’ demands were subsequently analyzed using survey data to improve the understanding of demands on the transit system. The finding indicate that the clustering method based on travel patterns work well to distinguish the transit users and their demands. The segmentation of users enables authorities to precisely reach users with similar demands. The paper’ method and conclusion could support more precise operational and strategic decisions.

One of the major challenges transportation agencies face when applying big data to address interchange safety or operational issues is not having sufficient inventory to provide information on the location, operations, or geometrics of the highway interchanges. The objective of this paper is to establish a prototype of the comprehensive database that includes all domestic system (expressway) interchanges. The paper presented an efficient approach to create and maintain the data of interchanges in terms of geological position, interchange type, and horizontal alignment. 1398 system interchanges were precisely located and inventoried in the paper. 1306 types of interchanges were successfully recognized, with the accuracy of 94.49%. A plugin application was embedded on Amap for real-time information retrieval. The database developed in this paper aims at providing a good reference to interchange design, operation, maintenance, and safety. The study is also expected to pave the way for follow-up researches on highway interchange area, which may further use supplement data and examine the sensitivity of interchange design out of safety and operation.

Large-scale taxi GPS data contains spatial and temporal information of urban systems, which reflects time and space distribution of traffic flow and characteristics of urban land use. This paper adopts a k-means clustering method to identify the home-work zone based on the origins and destinations extracted from the taxi GPS trajectories data. The original data used in the study were collected from more than 13,000 taxis in Shanghai City. Starting with the analysis of origin-destination distribution both temporally and spatially, three features (e.g., the hourly origins, the hourly destinations, and the hourly increment) were selected to evaluate the likelihood of a zone to be a residential area or workplace. Further, after dividing the urban area into grid cells of equal size and calculating the feature value of each grid cell, k-means algorithm was used to cluster the residential area and workplace locations. Finally, the result against the reality area was evaluated. Validation results demonstrate that the proposed method was able to identify home-work zone.

Holidays have been found to have a large impact on travelers’ mobility patterns. Understanding the holiday effects on airport landside passenger volume is important to prepare access mode services and manage landside traffic. However, the effects have little been investigated in the literature. This study analyzed the data from daily landside passenger volumes collected from year 2012 to 2014 at Shanghai Hongqiao International Airport Terminal 2. A SARIMAX model for passenger volume prediction was used to quantify the holiday effects, and the results indicated that passenger volume decreases during holiday period. A comparison between SARIMAX and SARIMA model is presented to compare accuracies and it is shown that the SARIMAX model is capable of interpreting both internal and external influences on passenger volume prediction.

Short-term traffic speed prediction is an important technique for advanced traffic management systems. In this paper, a deep architecture combining convolutional neural networks (CNNs) and restricted Boltzmann machines (RBMs) with Gaussian units (GRBM-CNN) is proposed to predict short-term traffic speed; it combines the advantages of RBMs and CNNs and achieves a powerful capability to learn complex features of traffic data. Furthermore, this paper uses a graphics processing unit (GPU) to accelerate the learning process. Experimental results show that the proposed deep architecture has a lot of advantages over state-of-arts on the short-term traffic speed prediction task.

In data mining of trains’ actual timetable, data cleaning is a key step that helps to improve data quality and data utilization. This study presented a data cleaning method based on data mining technology to preprocess train actual timetable data. Using the outlier detection method, first abnormal data was detected, then the missing data were classified as missing timetable data and missing outbound and inbound train information. Finally, an interpolation method was proposed based on the planned timetable to process two types of missing data. This methodology was examined a month of actual timetable for a rail transit line. Experimental results demonstrated the efficiency of the proposed method.

There are some joint activities in families due to interdependence among individuals, such as substitution, assistance, etc. This paper starts with the spatio-temporal characteristics of joint activities and interactive relationships between members. Firstly, space-time path are extended, which are able to express the temporal change process of group interaction behavior. Secondly, based on object-oriented modeling methods, the concepts of "behavior event" and "person event" are defined and used to describe "active" objects and build semantic organization of member interaction based on spatio-temporal GIS, achieving the expression of individual activities processes and group interaction processes at the same time. Thirdly, on the basis of semantic organization relationships of member interaction, a spatio-temporal data model of family activities is built. This paper proposes a 3D visualization tool for spatio-temporal data of family joint activities, and visualization of the space and time path and joint activities of family is achieved through experimental data.

With the development of information technology, data have been changed on many aspects, such as volume, structure, production, process, which significantly reform the field of statistics. The government statistics on road freight transportation also experiences challenges in the era of big data. To better understand the value of road freight big data, the difference and the relationship between big data statistics and the government statistics was described based on the statistical procedures. In addition, the structure of road freight big data was proposed according to the participants (e.g., people, vehicles, goods, roads, and environment), along with its influence on government statistics. By analyzing the scope, technology, and services, the value of road freight big data to government statistics is to increase the indexes of statistics, to produce meaningful reports and to benefit vast population.

As the big data first arrived from car sharing systems, the rapid development of transportation big data has had a huge influence on people's way of considering transportation. However, information such as user location, home address, travel path, and individual preferences can be contained in big data. This raises valid privacy and security concerns in the analysis of big data. This paper aims to analyze the existing framework of data privacy protection protocol as well as general technology used in data privacy efforts. Then, the paper proposes a composition of the lifecycle of transportation big data, summarizes the challenges of data privacy based on this lifecycle, and makes suggestions for successful data protection at each stage of the data lifecycle. Finally, big data privacy protection system has to be built to ensure the fulfillment of the privacy protection of transportation big data.

In this paper, a novel person re-identification method is introduced under the problem of path selection. Unlike other supervised person re-identification algorithms, our method can identify persons in the video without artificial markers. Our method divides each of the image sequences into several slices and selects the most distinguished ones, which can improve the performance. More crucially, this model is unsupervised and readily scalable to real-world large scale ReID settings, and more suitable to previous path selection, i.e., no need of exhaustively collecting large numbers of cross-view pairwise labels for each camera pair, as required by most existing ReID models, which need supervised training. Experimental results show that our method can outperform other methods and achieve excellent results.

The unmanned aerial vehicle (UAV) can be an efficient and economical solution to traffic surveillance mission for collecting real-time traffic information of special areas. In some cases, UAV is unable to transmit the traffic information back to the ground control station directly. Deploying multiple UAVs and to build a communication relay link is a feasible method to avoid the building and other obstacles and finally transmit the traffic information successfully. According to the number of traffic surveillance target, the multiple UAVs deployment problem is classified into single target UAV space deployment problem (STUSD) and multi target UAV space deployment problem (MTUSD). This paper focuses on the MTUSD problem, which is modeled as the Steiner tree problem. PGA algorithm is developed based on the combination of Pareto solution and GA algorithm for MTUSD. Moreover, the experiments demonstrate the effectiveness of the PGA algorithm for the multiple UAVs space deployment problem.

In the near future, real traffic scenarios will be randomly mixed with adaptive cruise control (ACC) vehicles and manual vehicles, where fundamental diagram and driving comfort are studied under different ACC penetration rates. The car-following model validated by PATH program based on real experimental data is employed as the surrogate model for current commercial ACC vehicles, and the intelligent driver model is used as the manual driven model. Not only is the analytic fundamental diagram of the mixed traffic flow derived, but also the effect of ACC penetration on the scattered pattern of fundamental diagrams is investigated. Then, passenger comfort is measured using simulations, whose major cause is also discussed. Impacts of ACC and its desired time-gap on fundamental diagram and driving comfort are obtained, which can be as a reference for determining the value of ACC desired time-gap, under the consideration of a better balance between capacity and comfort.

Fog has a negative impact on freeway traffic safety and efficiency due to the reduced visibility. Variable speed limits (VSL) can improve traffic states under foggy weather conditions. However, most of the previous researches calculated the speed limits by stopping sight distance model, which is likely to cause sharp fluctuations. This paper presents a VSL control method based on the real-time road visibility and traffic states. The speed limits of road sections are determined by maximum safe speed, speed difference between adjacent sections, etc. A part of G4 freeway is selected to evaluate the proposed VSL control method, and simulation experiments are conducted by using VISSIM simulator. The simulation results show that the lowest speed has increased by 34.9%, and the maximum speed difference of adjacent links has decreased by 67.77%, compared with no VSL control. The simulation indicated that VSL control can ease the congestion caused by low visibility.

Electronic toll collection systems are convenient for a freeway toll, and it also provides a new resource for the extraction of freeway traffic state information. This paper analyzed the principle of traffic information collection based on the dedicated short range communication (DSRC) technology based ambiguous path identification toll system. A method of estimating traffic volume of road sections and the average travel time is proposed by processing and fusing the path information and toll data in the system. The average travel time of the link is predicted by using BP artificial neural network. Finally, the field experiments verified the feasibility of the proposed method and analyzed the error of average travel time prediction for freeways based on ETC-based path identification toll system on the test section of Nanjing-Hangzhou Expressway. The average percentage error of prediction results is 9.2%, which indicates satisfactory results.

In order to obtain the information of passengers’ public transit transfer behavior, by refining the time-space evolution of public transit transfer process, and utilizing vehicle GPS data and smart card data fusion techniques, this paper calculates the boarding time and alighting time of passengers. Then, principles including transfer distance threshold, transfer time threshold, and bus arriving time matching are used to identify public transit transfer behavior with the method of layers screening. Compared with existing methods, transfer recognition by the above-mentioned method takes more factors into consideration and is more accurate. Finally, four continuous public transit travel records of an actual passenger from Foshan are selected as an example to verify the practicality and efficiency of the method. Results show that the above algorithms can effectively recognize public transit transfer behaviors including transfer inside one station and transfer between different stations. The recognition method can provide data support for planning and management of public transit network, hub location, and public transportation system efficiency improvement.

Adaptive cruise control (ACC) systems can effectively reduce drivers’ workload, improve vehicles’ active safety, and make driving more comfortable. Real-time leading vehicle detection is one of the most important research tasks in the ACCs. The traditional leading vehicle detection methods based on radar or vision have many problems on curved roads, such as limited detection range, low real-time performance, poor adaptability to the environment, and so on. This paper proposes a method of real-time leading vehicle detection on curved roads based on the vehicle-to-vehicle (V2V) communication technique. We collect vehicle state information by equipment on-board and use V2V communication equipment to send, receive, and store this information. Moreover, one vehicle detection algorithm based on contact ratio of vehicle trajectory has been designed. The experimental results indicate that the method can meet the requirements of excellent accuracy, high real-time performance, and good adaptability to the environment.

This paper is based on vehicular networks and analyzes disadvantages of collecting road information through fixed detectors. It presents a basic principle and design framework about collecting traffic datum based on vehicular networks. On the basis of this, it mainly studies a discriminating model about network traffic state based on vehicular networks and analyzes vehicular networks’ specific features. This paper collected urban road traffic state information by using GPS, and then in network traffic information that was collected, after pretreatment, was combined with an improved ID3 algorithm, using C# language, to output road traffic condition. Using the algorithms of vehicular networks combined with ID3 algorithms, the road traffic state can be quickly judged. The result of examples show that the proposed method can effectively discriminate the road traffic state and improve road capacity.

On the basis of reference to the mainstream APPs evaluation system and standards at home and abroad, this paper establishes the evaluation index system for APPs of public transport information from three aspects, include functional quantitative index, user experience index, and developer self-evaluation. The functional quantitative index includes 5 quantitative indicators such as dynamic line information coverage rate, static line information coverage rate, line information accuracy rate, travel time prediction accuracy rate, and arrival prediction accuracy rate. The user experience index includes six aspects, such as convenience, accuracy, fluency, degree of timeliness, stability, and satisfaction. Developer self-evaluation focuses on the data source, information transmission mode, and software features. Through the appropriate algorithm summary evaluation results, we obtained the final rating of APP. Finally, taking the domestic three mainstream navigation software as a case, an example is given which has a certain sense for reference. The evaluation results are consistent with the facts, which verifies the usability and objectivity of the evaluation system.

In order to optimize the problem of carbon emissions in transportation networks, the environmental shortest path problem is addressed in a traffic network based on carbon emissions. Firstly, the suitable calculation model for carbon emissions is suitably selected. Secondly, a nonlinear constraints optimization mathematical model is established to the environmental shortest path problem through defining carbon emission as the objective function. Thirdly, a label algorithm is constructed to solve the environmental shortest path problem. Finally, numerical results in transportation network show the validity and feasibility of the proposed label algorithm. The results suggest that the environmental shortest path based on carbon emissions is significantly more difficult to solve to optimality than the shortest distance path and least travel time path problem, but offers insight on navigation of “environmental-friendly” vehicle routing.

Deep learning theory, as a powerful set of techniques for learning in neural networks, has achieved numerous successes in the domains of artificial intelligence (AI). Based on huge amounts of probe vehicle data, this study attempts to apply the novel deep learning theory to the multi-step travel time forecasting on urban expressways. Serving as state-of-the-art neural network architecture, long short-term memory neural networks (LSTMs) are employed due to their superiority to time series methods. LSTMs enable capture of the underlying structure of data from a pool of historical datasets and excavate the intrinsic traffic features without strong assumptions on their temporal evolution. The proposed approach is investigated on Ring 2, a 33 km urban expressway of Beijing, China. The results demonstrate the advantage of the proposed method, as well as its feasibility and effectiveness compared with other prevailing parametric and nonparametric algorithms.

The car-following model, referring to the way that one car follows another in a single lane under non-free flow conditions, is the core of microscopic traffic simulation. The traditional car-following model cannot capture uncertainties related to driver during driving. These uncertainties can result from the driver’s physiology, psychology, and/or mental state, namely, the driver’s propensities. In this study, driver’s propensity is introduced into the car-following model to account for differences between drivers. The data of driver behaviors, vehicle states, and driving environments are collected and analyzed through real vehicle experiments. The car-following model based on driver’s propensity is established using a simulated annealing neural network algorithm. The simulation results show that the car-following model is accurate and feasible and can successfully imitate the car-following behaviors.

Gray correlation analysis was used to analyze the relationship between cycle length and controlling indexes, such as delay time and queue length, under different saturation conditions. Introducing a delay variation coefficient, with minimum average delay time, queue length, and maximum capacity as control targets (cycle length and minimum green time are used as constraint conditions), a cycle time optimization model for signalized crossing is established. In this model, weight of control index will be determined according to actual situation. To verify model validity, an intersection of Qingdao is taken as an example. R software is used to optimize the model. The F-B method, HCM method, and optimization model were simulated by Vissim software. By comparing simulation results, it shows that the proposed model can reduce vehicle delay, effectively improve intersection operating conditions, and provide a theoretical basis for intersection signal design.

With the development of big data, mobile phone data are used to study the characteristic of transportation system in the last few years. Since the land use is important for urban planning and development, in this paper, an algorithm of determining urban functional area classification is proposed based on mobile phone data. Firstly, the number of residents in different period are selected to characterize semantics of land. Then, an unsupervised k-means clustering algorithm is applied to identify urban functional area. Four patterns of urban functional area could be identified such as workspace, residence, temporary workspace, and mixed area. All these findings could be helpful to the government for urban planning in the future.

Considering the existing problems of high damage rates inflexible settings, and poor data reliability of the traditional loop detector, this paper proposes a dynamic detection method for intersection traffic state based on intelligent video analysis. This method is verified by timing optimization at an actual signalized intersection. Practical operation results show that the traffic state parameters of signalized intersections were detected accurately and effectively, and the problems that existed in the traditional loop detector were overcome. Meanwhile, the traffic efficiency was also significantly improved by this new method. The method of real-time optimization and control of traffic signal based on intelligent video analysis could be widely used in traffic signal control and has a wide range of application and practical value.

Nowadays, with the urbanization of the cities, traffic is becoming a severe problem. An efficient traffic signal control strategy is beneficial for improving the efficiency of the traffic. With the development of intelligent transportation systems (ITS) and intelligent vehicle infrastructure cooperation systems (IVICS), it is possible to improve the traffic performance by implementing real-time and accurate traffic information. This paper proposes an innovative algorithm based on game theory to control traffic signal for single intersection in real time. By utilizing the theory of mixed strategies, Nash equilibrium, an optimum signal timing plan for the isolated intersection can be achieved. Finally, this game theory-based traffic signal timing model is simulated on the VISSIM platform. Compared with the fixed time signal control system, the new signal timing algorithm shows improvement in the intersection traffic performance through reducing the average queue length and therefore improving the effectiveness of signal control.

This paper proposes a dynamic control strategy and switch method for the operation of reversible lanes on freeways, bridges, or tunnels. Five traffic situations were defined considering different traffic saturations on each direction. Average delay of all the vehicles on the road were used to develop operation models or reversible lanes under different states. The proposed method were designed for both one and multiple reversible lane with continuous flows. Enumeration algorithm were used to solve the optimization models. A dynamic switching method of driving direction at reversible lane were proposed using variable lane marks and lane direction indicative signals on gantries under vehicle-to-infrastructure communication system. The proposed methods and models were simulated and tested using data of a five lanes bridge in Guangzhou via microscopic traffic simulation. The results show that the proposed method and models were able to reduce the average delay and traffic congestion on studied road section.

This paper established a framework and measures of effectiveness used for assessing the benefits of auxiliary road metering system (ARMS). A case study was performed to assess the actual system deployed in the southern exit of Dongsishitiao Bridge of Beijing 2^nd Ring Road. As a result, some conclusions are drawn as follows: 1) in some special time, the ARMS can improve the efficiency and throughout, especially when the traffic demand is high; 2) the auxiliary road users are the loser, they encounter some unavoidable delays. The system can also generate a long queue that may affect other neighboring network and the entrance when the auxiliary road volume is high; 3) not all the vehicles which travel from the mainline to auxiliary road are the winner. Some may benefit from the system, others are not; 4) the control strategy needs to be improved, and the simulation test result shows that the long cycle fixed time control strategy which has long time red cycle and green cycle may worsen the traffic condition under some special traffic demand; and 5) not all system that has been deployed operate effectively, i.e., some investigations should be conducted to clarify the reasons for the exit congestion, traffic demand, and surrounding environment before the system is deployed.

Variable speed limit (VSL) control is a relatively common control strategy. By smoothing traffic speed and optimizing traffic status, the efficiency of urban expressway systems would be greatly improved. With the development of intelligent transportation systems (ITS), intelligent vehicle infrastructure cooperative systems (IVICS) can obtain the real-time microscopic information of vehicle more accurately and broadcast the speed limit information to the drivers. By combining IVICS and VSL, we can set up speed limit sign with on board unit (OBU) and obtain better control efficiency. In this paper, we proposed a VSL broadcast strategy based on IVICS. Focusing on the work zone of urban expressway system, this paper built up a simulation scenario based on Microsoft Visual Studio and VISSIM co-simulation platform. The simulation results indicate that in the situation of urban expressway work zone, the proposed VSL broadcast strategy can alleviate congestion, reduce the length of queue, and decrease travel time.

Dedicated short-range communication (DSRC) and smartphone have unique advantages and great potentials in improving traffic safety. DSRC is a potential technology that allows drivers to get driving information of other vehicles in a range beyond their normal vision, which provides support for vehicle driving safety. Smartphones integrate a variety of sensors and interfaces. In this paper, the DSRC-based rear-end collision warning application for android smartphone is proposed, which warns the driver when it detects subject vehicle is in risk of collision. In order to offer the best warning and mitigation mechanisms to interact with the driver, the collision avoidance warning strategy with three levels of warning is developed. The test results show that the proposed DSRC-based rear-end collision warning application has high correct warning rates in different warning levels.

In order to meet the information demands for connected vehicles, this study develops a novel multi-source information fusion algorithm and in-vehicle data collection terminal. Firstly, the intelligent vehicle-mounted terminal can analyze the vehicle bus data according to J1939 protocol and collaborate with other sensors. Secondly, we developed a real-time vehicle data management system to share the vehicle data and send the traffic conditions and guidance information to the terminals by vehicular communication system, which can enhance the safety and improve the traffic information service. Finally, we establish an online traffic evaluation system based on the fuzzy theory for traffic simulation. By integrating the dynamic vehicle-mounted data and static intersection information, this system can generate a set of road evaluation indicators based on real-time data. By comparing the evaluation results and field conditions, the simulation results showed that the algorithm and terminal is feasible.

Transit scheduling is the basis of the operation and organization plan of a public transportation enterprise; it is also the core of bus dispatching. In this paper, we will compile the collaborative departure schedule in the bus route of the region to maximize the convenience for passengers. This paper introduces formulation of a regional schedule as a special kind of 0-1 knapsack problem with a special coordination coefficient of no capacity constraint and with the minimum passenger transfer time as the objective. The synergy coefficient is defined, and the corresponding mathematical optimization model is established, which realizes the regional and coordinated development of the public transit timetable. The enterprise can effectively allocate vehicle resources and human resources, in order to lay the foundation to achieve a comprehensive system of regional bus operators.

Formation control of intelligent vehicles can improve traffic capacity and reduce fuel consumption. The paper proposes a consensus-based formation control protocol, where the spacing and velocity of micro intelligent vehicles (microIVs) are regarded as the consensus states and switchings of multi-modes are introduced to adapt to the change of road structures. The formation algorithm is validated in a cooperative vehicle-infrastructure simulation platform (CVISP) consisting of a micro road network and microIVs. In the experiment, positioning and spacing calculation are addressed. The optimal positions of microIVs are obtained by utilizing a Kalman filter to fuse the absolute positions from RFID labels and the relative positions from dead reckoning (DR). The Romberg method and an assisted spacing algorithm are used to obtain the spacing information under different road conditions. With the relative state information, the consensus-based formation protocol is implemented in CVISP, and its efficiency is validated in the experiment.

Traffic surveillance videos can provide data of multimodal (car, bus, pedestrian, bicyclist, et al.) activities for transportation research in general. However, automated data extraction remains an issue due to illumination variation, size scaling, and angle changes. To tackle these challenges, this paper proposed a deep-learning based multimodal detection and tracking system. The proposed system consists of two modules: 1) a multimodal detection module which is based on the object detection framework of Faster R-CNN, and 2) a multiple object tracking module which is based on the KCF tracking algorithm. A comprehensive experiment based on traffic videos captured in different scenarios is conducted to evaluate the performance of the proposed system. Testing results demonstrate that the proposed system can achieve better performance for multimodal detection and tracking compared with existing methods. The proposed system is tested to be robust to illumination variation, size scaling, and angle changes.

The full-scale ring road (full-scale track) accelerated loading test system can be considered as a closed ITS. The traffic information collection and transmission are very important for the traffic control, especially for test safety and efficiency. This paper shows the key processes of the system design and implementation about the traffic information collection and transmission on full-scale ring road for accelerated loading test. Firstly, we select the scope of traffic information, including vehicles and road information, which is closely related to the test control. Secondly, the information is classified, and the engineering implement methods of traffic information collection and transmission are studied as well. At this stage, equipment capacity and transmission network structure need to be considered. Finally, the system construction needed to be completed in next few months. In current phase, we have already started to the debug scheme, including the research of difficulties in debugging and corresponding solutions.

Real-time traffic flow forecasting is of vital significance to the intelligent transportation system (ITS). Providing real-time and efficient information for travellers, the forecast aims to reduce travel times, control traffic pressure, and reduce pollution. The time-series autoregressive integrated moving average (ARIMA) model has been widely used in advanced traffic management systems. Compared to the traditional ARIMA model, the autoregressive integrated moving average with exogenous inputs (ARIMAX) model can take the impact of covariates on the forecasting into account to improve the comprehensiveness and accuracy of the prediction. Based on data accessibility and correlation analysis, the upstream traffic flow is taken as the covariate X of the model. Data was collected from Interstate Highway 280 in California, with a sampling period of 5 minutes. The results showed that the ARIMAX model outperforms the ARIMA model during morning peak hours in terms of accuracy and comprehensiveness.

In urban transportation network, generally, traffic sensors are unable to fully cover the freeway network, and only local measurement data is available. To accurately and timely estimate traffic density in an urban freeway network, this paper proposes a distributed state observer approach based on a macroscopic traffic flow model. Firstly, the urban freeway network is modeled in the way of distributed systems by using the cell transmission model (CTM). Secondly, based on the model, a distributed observer is designed, which gains are computed by using LMI technique according to the Lyapunov function and S-procedure theory. Thus, the traffic densities of the whole network can be estimated by the distributed observer. This method was applied to Beijing third ring road and experimental results demonstrate the effectiveness and applicability of the proposed approach.

The heavy congestion and pollution caused by the rapid growing vehicles have become the bottleneck of developing harmonious city. With the development of CV (connected vehicle) technology, CV traffic management and control system based on V2V (vehicle to vehicle) and V2I (vehicle to infrastructure) has become an effective means to solve the urban traffic problems. Adapting the variable pattern of traffic demand, the methods of adjusting dynamically the unit extension time and its times are adopted to design three IA-SPaT (intersection actuated signal phase and timing) algorithms. Using the OE-CVS (one engine connected vehicle system), the proposed algorithms are validated and three indexes of delay, queue length, and emission are analyzed. The results indicate that three actuated algorithms have a reduction of around 20.3% for average delay, 5.5% for average queue length, and 9.7% for average pollutants emission comparing to the fixed signal phase and timing algorithm.

The advanced traffic information system can be effective, only when drivers accept and adopt traffic information provided by the advanced traffic information system. The aim of this paper is to examine drivers’ satisfaction to traffic information system and figure out the reasons why drivers would accept or refuse the information by quantifying the impacts and values of various factors that are related to drivers’ compliance and satisfaction with variable message sign (VMS). To achieve these goals, reliability analysis and structural equation modeling (SEM) approach was adopted. The results revealed that traffic information quality assessment was the most significant factor that positively affected drivers’ acceptance intention of traffic information displayed on VMS. Moreover, drivers’ satisfaction with traffic information was found to be an important factor that affected drivers’ acceptance intention of traffic information. Furthermore, traffic information utility assessment was the most significant factor affecting drivers’ satisfaction with traffic information. Based on the findings of present studies, several suggestions are proposed to improve the traffic information service.

Short-term traffic flow forecasting has been the most important application of the intelligent transportation system (ITS). This paper presents a model structure with a coupled hidden Markov model (CHMM) for short-term traffic prediction in the highway system with real-time traffic flows data. Data used in this study was gathered from simulation software. The model defines traffic states in a two-dimensional space with speed and volume observations. The decoding function of CHMM is used in this study to estimate the most likely sequence of traffic states. The forecasting model is accessed by predicting errors. The CHMM is compared to autoregressive moving average (ARIMA), which is one of the most widely used regression techniques. These results present that the CHMM outperforms the regression model. Consequently, the paper concludes that CHMM is more robust and successful in modelling unstable traffic conditions.

This paper identifies the main factors that influence the reliability of the variable message signboard (VMS) are driver board, power supply, and control board through the multi-aspect analysis of the VMS historical fault data. A different factor prediction model is obtained based on fault factors changing with the time. The validation example shows that the accuracy of the model is over 80%. It provides a superior data support and experience reference for VMS fault detection and spare parts which could improve the work efficiency of maintenance personnel.

To better simulate car-following behavior, this paper presents a new car-following model based on the OV model impacted by road gradients. The formulations are determined based on two different driving strategies including the aggressive strategy and the conservative strategy. A stability condition of the proposed model is obtained using the linear system theory. Through the statistical analysis on vehicle trajectory data collected in Guangzhou, it is shown that car-following behavior on the slope is different from those on the flat road. It is necessary to build car-following models considering gradient impacts. The proposed model shows its better performance compared with the OV model, implying that the proposed model is effective and able to capture the difference between gradient road driving and flat road driving.

In this paper, we proposed a method for vehicle localization with low-cost GPS and a monocular camera. Traditionally, the positioning accuracy of a single GPS can merely be reached at the metric-level both laterally and longitudinally. By integrating vision and GPS with the prior position database, which was generated by the bag-of-words method and improved by lane change detection, our method can achieve effective improvement of positioning accuracy, which is able to acquire the location of lane laterally and more precise vehicle coordinates than raw GPS data. In addition, by leveraging the complementary global information, our GPS-aided vision method has the advantages of avoiding inaccurate pose initialization of SLAM systems and obtaining absolute coordinates which can be deployed to navigation directly. The proposal was described by pseudocode in detail and was evaluated by simulation and a dataset collected by our team in an urban environment. The improvement and validity was verified by experiment based on our proposed evaluation criteria.

Predicting urban network congestion and exploring congestion mechanisms are vital for both transportation researchers and practitioners. The state-of-the-art studies rely on either mathematical equations or simulation techniques to depict the traffic congestion evolution. However, most of the existing studies tend to make simplified assumptions since transportation activities involve complex human factors which are difficult to represent or model accurately using mathematics-driven approaches. In this paper, long-short term memory neural networks (LSTM NN) are employed to interpret traffic congestion in terms of traffic speed. Traffic speed predictions are also made by considering both temporal and spatial correlation information. The proposed approach is tested on different links in one road network in Beijing, China. The results demonstrate the advantage of LSTM NN for analyzing the complex non-linear variations of traffic speeds as well as its promising prediction accuracy.

The expressway construction in China has developed dramatically, while the development of traffic information service lags significantly behind, which leads to a series of traffic problems. Therefore, it is urgent to improve the traffic situation by obtaining the needs of traffic information service and realizing the service with high quality and personalization. By analyzing the explicit and implicit requirements, together with the information customization method, travelers’ demand of traffic information service on expressway is modeled by a multinomial logistic model, using statistical software, SPSS. Finally, based on the data obtained from “stated preference” survey in service zone of Yangcheng Lake, it is found that the measured data is consistent with the prediction model, which verifies the accuracy of the prediction model.

The introduction of vehicle-to-vehicle (V2V) communication brings the change to further improve the stability of traffic flow. In this paper, we present a new car-following model, considering the effects of multiple preceding cars’ optimal velocity changes with memory, and backward looking. To address the linear stable judging condition of this model, some linear stability analysis has been done. Also, nonlinear analysis has been conducted to obtain the modified Kortewege-de-Vries (mKdV) equation around the critical point and the corresponding kink-antikink soliton solutions. Then, the phase diagram was divided into stable, unstable, and metastable regions. Moreover, as the multi-preceding cars’ optimal velocity changes and backward looking effects are considered, the stable area is obviously enlarged. Finally, some numerical simulations are carried out under the periodic boundary conditions. Compared with the full velocity (FVD) model and the optimal velocity changes with memory (OVCM) model, our results demonstrate that the proposed model can enhance the stability of traffic flow under the same initial disturbance conditions.

Lane-departure detection is an interesting and challenging task in the field of computer vision, providing an efficient way to improve driving safety. In this paper, we present a vision based lane-departure detection algorithm using local Hough transform and particle filter. A monocular vision based experiment platform also has been designed for implementing our algorithm. There are two main contributions in our work: Firstly, a novel lane detection method using local Hough transform was proposed to obtain the lane marking coordinates location. This method was proved to apply to fast and dynamic detection results. Secondly, we designed a detecting and tracking strategy, which is implemented by a particle filter, to ensure the stability of the system. The actual roads experiments are used to test the performance of our algorithm, and the experimental results shows that the proposed method is of demonstrated feasibility and reliability on actual roads.

In the urban public transport system, the ground bus is always playing an important role in residents’ traveling due to its high accessibility. During the past two years, the attraction of the ground bus has declined because of mismatch between supply and demand. For ground bus, understanding the characteristics of space distribution and evolution of public transit demand will contribute to the optimization of public transport capacity scheduling, as well as the adjustment of network, thus improving the quality and attractiveness of ground bus services. This work uses bus GPS positioning data, IC card data, and AFC card data, based on four time dimensions: 2015 January, June, December, and 2016 June, to analyze the characteristics of regional public transport demand evolution. All data were offered by Transportation Operations Coordination Center. The study uses the daily average boarding and alighting volume of the bus stations, the daily average passenger volume of bus lines, and regional daily average occurrence volume at morning peak to present the demand of ground bus at different levels. Identification methods were established to recognize special region which is under strong demand fluctuation, combined with the case analysis of typical region to explore the cause of the fluctuation. The work provides data support on decision making for the ground public transportation system in order to enhance residents’ attraction and balance demand and supply.

Now unmanned aerial vehicle (UAV) plays a more and more important role in urban traffic management in Shenzhen, because of the advantages of flexible, rapid response, and low cost. One of the challenges of constructing digital urban traffic model is that it cannot obtain the static and dynamic road traffic data quickly, accurately, and effectively. In this paper, a real-time modeling platform for urban road traffic based on 3S (RS, GPS, GIS) and UAV is proposed, which is oriented to urban traffic modeling and simulation analysis. Emphasis is placed on the blurring dynamic reconstruction model and algorithm of low altitude airborne aerial video images. The motion-blurred degraded model and the motion blur reconstruction algorithm based on Radon transform are designed. Finally, the method is verified by the measured data.

The requirements of communication area on lane in ITS-station system for highway status abstraction were analyzed. To expand the communication range and determine the lane that the vehicle is running, a new scheme of vehicle detection was presented. We use a narrow beam to detect one lane at one time and the narrow beam switch direction between the lanes. To implement the scheme, a novel design of 2×8 microstrip phased array antenna is proposed and designed for ITS-station. The phased array antenna can be electrically steerable, and the effective radiation pattern of the array can be reinforced in a desired direction and suppressed in undesired directions. The antenna is composed of a novel phase-shifting network, which consists of several groups of switch combination. By controlling the states of the switch combination, three narrow beam in -16°, 0°, 16° were formed, which respectively cover three lanes.

In order to improve safety of vehicle operation and efficiency of passenger and cargo transportation, a new traffic organization scheme of separating trucks from passenger vehicles is proposed, which is scientific and effective. The traditional intelligent transportation systems (ITS) architecture and its key technologies are suitable for the mix of passenger cars and trucks but lack special research on the two characteristics of “passenger and cargo separation” and “multilane.” In this paper, one ITS architecture and four key technologies are studied systematically. Among them, the ITS architecture represents one special structure designed for multilane freeways, which separates trucks from passenger vehicles. In addition, four key technologies consist of intelligent monitoring, traffic state estimation and prediction, advanced guidance, and coordinated control. Related achievements can be guidance for freeway design, operation, and management.

Development of public transport is the fundamental way to solve urban traffic congestion, and transit way priority is the key to improve bus operating speed and attract passengers. This paper compared the advantages and disadvantages of different kinds of vehicle detection technologies, put forward some signal priority control methods and different control strategies of buses based on automobile electronic registration identification with radio frequency identification technology, and built the communication mechanism between RFID reader devices and signal control machines. Through example applications of some signalized intersections in Wuxi City and Shenzhen City, experimental results show that these signal control methods can make the average delay of bus at intersections without bus-lane reduce obviously and improve transit operating efficiency.

Bus travel decision-making behaviors were analyzed from the aspect of the travelers’ psychology and physiology based on their planned behavior. The main factors affecting the bus travel decision-making were extracted by investigating bus travel decision-making behavior through survey data. A logistic regression model was introduced to analyze the data. Urumqi City bus travel was taken as an example for this study. The results showed the application of planned behavior to solve the bus travel decision-making behavior reflects the effect of the individual differences of the travelers. This was especially true with respect to the travelers’ psychology and physiology. The results facilitate accurate prediction of the bus travel volume in the hub and provide a scientific basis for improving bus service and the bus-sharing rate.

Existing ramp metering approaches can sometimes not proactively offer a traffic control solution, especially when traffic demand is high. However, connected vehicles provide a possibility for active traffic control. Based on the study of connected vehicles, this study proposes a systemic architecture of coordinated ramp metering under the connected vehicle environment. A Paramics-based simulation test bed was applied to a real life network to simulate and evaluate coordinated ramp metering. The simulation found that the ramp metering approach proposed in this study was effective at improving traffic efficiency. Sensitivity analysis results found the designed ramp metering approach was sensitive to control loop length.

Accurate short-term prediction of traffic conditions on freeways has recently become increasingly important because of its vital role in the basic traffic management functions and trip decision making processes. The objective of this research is to utilize traffic and weather data from multiple data sources to develop an integrated model to predict traffic conditions under different rainfall conditions. A set of prediction models are compared and their performances using data from case studies are investigated and reported. The model performance was valuated using prediction errors, which are measured by the relative length of the distance between the predicted state and the observed state.

As an essential part of dynamic traffic assignment in time-dependent road networks, route choice set generation has been studied extensively in recent decades. The K shortest path searching algorithm has been widely used to generate the choice set with the assumption of bounded rationality. Some improved classical path searching algorithms and heuristic approaches have been proposed for the K shortest path searching, but most of them are time-consuming or in some cases, unable to find the optimal results. In this paper, a K shortest path algorithm was designed as a dual-loop framework for urban network route generation. In the framework, the A-star algorithm was used for shortest path search. Meanwhile, the link elimination approach was used as the strategy when searching for the k^th shortest path. In order to improve the efficiency of the proposed model, heuristic function in A-star algorithm was also rebuilt according to the characteristics of urban network. The proposed algorithm was tested using a real urban road network. Empirical analysis shows that the proposed algorithm is computationally appealing and succeeds in finding the global K shortest paths. A genetic algorithm and the traditional Dijskstra algorithm were also conducted for comparison analysis. It is encouraging to find that the proposed algorithm significantly outperforms the genetic algorithm in terms of accuracy and is more efficient than the Dijskstra algorithm.

Survival analysis is used to analyze site data to explore vehicles’ mandatory lane changing behaviors and related influential factors. The traffic data on the mandatory lane changing durations of vehicles were collected by an unmanned aerial vehicle (UAV) in a freeway maintenance construction area. A multiplicative hazard model of mandatory lane change was established using the semi-parametric method of survival analysis. Meanwhile, the lane change data has been analyzed by Cox regression analysis. The results show that about three quarters of vehicles’ MLC duration is less than 10 sec. There is no significant evidence showing that different vehicle types have an effect on the duration of mandatory lane changing. The cumulative survival rate of MLC duration the off-peak period has been found to be significantly lower than that in the peak period and transition period. As it turns out, the cumulative survival rate of vehicles’ MLC duration in the transition period is the highest.

To explore the characteristics of driver reaction time in an Internet of vehicles (IOV) environment, a vehicle-to-vehicle platform was established to perform real-world driving tests. The characteristics of the following driver’s reaction time in an IOV were analyzed based on a Gaussian mixture model (GMM). By comparing fitted distribution models of the reaction times with the actual testing data in both traditional and networked vehicle environments, it was found that the GMM was superior to the normal distribution model and lognormal model in the IOV environment. The simple reaction time was reduced by 7.94%, the complex reaction time was reduced by 25.79%, and single lane basic traffic capacity was improved by 15.39% in the IOV environment. Furthermore, the population standard deviation of the driver’s reaction time was reduced by 32.99% and grew more concentrated. The effects of speed on reaction time were not significant in the IOV environment.

Intelligent transportation systems (ITS) has become a vital means of traffic management and control tool in the entire world. It provides necessary subsystems and functions to classify the traffic state. This paper proposed a dynamic automatic condition identification based local accuracy (DACI-LA) algorithm to classify urban road traffic state level. The new useful information traffic state was obtained from the DACI-LA. The learning ability of integrated classifiers can solve the limitation that pure classification methods can’t train sample data. Finally, the experiment results indicated that ensemble learning method can reduce the influence of noise sample data and provide better identification accuracy.

In recent years, advances in sensor technology have made three-dimensional (3D) pavement automated measurement systems practical for collecting accurate road distresses. However, until now, there has been no information showing that there is an algorithm or software that can be adapted to collect data and achieve automatic recognition rates of more than 90%. The main impetus for this paper is derived from the lack of information on the mainstream methods that have been widely used in commercial applications, among them including laser triangulation, 3D laser scanner, and 3D PIXA based on stereo and fast line scan sensors. In view of the corresponding principle, this paper will focus on the hardware aspect, comparing their respective system specifications in regards to width range, depth range, detection speed, and accuracy. In addition, given the advantages and limitations, the recommended road type and detection range are proposed. Therefore, this paper will be able to serve as a reference for researchers, technical engineers, and managers for the selection of methods, system structures, and the future prospects.

Chinese big cities are facing severe traffic congestion problems and urban taxis are often underutilized and vacated by seats. Vacant seats can be made full use of in taxis by achieving efficient carpooling patterns; decreasing the number of taxis used is a method of mitigating traffic congestion. This paper attempts to establish a dynamic carpooling system based on time windows and path optimization. Real-time reservation mode refers to that the interval between traveler's reserving time and boarding time is quite short, requiring the carpooling system to give a rapid and flexible response and make efficient matching. The model constructed in this paper includes the detailed principle and logic composition, and main formulas. More research is needed to achieve specific programming and applications.

The current research related to traffic state detection in mountainous cities faces many challenges. Due to detection requirements, special road network features, and climate characteristics, traffic detection in mountainous areas requires special considerations. A traffic state detection system using unmanned aerial vehicles (UAV) for traffic management optimization has been developed and the parameters have been explored in this research. Based on a derived image information system, a traffic condition detection model for urban road construction, two kinds of highway system architectures, and a suburban integrated traffic information platform a UAV platform is developed. Finally, an approach for solving information challenges is addressed.

Traditional automatic train operation (ATO) system performs well on automatic driving. In the current researches of ATO, most of the scholars take the whole line as an object. But the final speed curve is very different from the actual speed. This paper proposes a developed strategy based on working condition to analyze and optimize parameter of energy consumption, driving comfort, punctuality, and precise parking. The model is built by optimized genetic algorithm and then get the speed curve. Finally, all speed curves are combined to get ATO multi-objective optimization speed curve. Through the test of example simulation, the strategy is proved to have a good optimization effect and maneuverability.

To detect a no signal state of "black screen," "blue screen," and "snowflake screen" in surveillance traffic video, this paper proposes a novel detection method of video images with or without signal based on AdaBoost strengthening BP neural network. At first, this method analyzes the characteristics of traffic video images with or without signal and establishes seven characteristic indexes, such as "jump point density," "interval peak ratio," "entropy," "energy," "contrast," "correlation," and "homogeneity"; then, we confirmed the structure parameters of the BP neural network and trained the neural network. Finally, the strong classification detector is formed by utilizing the method of "AdaBoost strengthening BP." Experimental results show that the proposed method can dramatically improve the correct detection rate and reduce the false detection rate of traffic video images with or without signal, which has considerable significance in practical applications.

To solve ambiguous path and toll splitting problem raised by freeway constantly construction, this paper investigates a path identification toll system (PITS) and its architecture based on 5.8 GHz dedicated short range communication (DSRC) technology. Due to the OD coverage criterion and link independence criterion, a layout scheme of ambiguous path identification points which meet the requirement of precise segmentation of road links was proposed. Also, the layout model with the minimum number of points as the target was established. Then, all paths of road network were found with depth-first searching (DFS) algorithm and layout model was found based on greedy algorithm, to select the maximum coverage number of paths. A freeway network in Jiangsu Province was selected as a case study to clarify the model. In view of freeway networking toll collection condition of Jiangsu Province, this paper laid the ambiguous path identification points of the whole road network.

A better layout of fixed detectors on freeways can provide higher estimation precision with lower number of detectors. This paper introduce travel time estimation method using cellular handover positioning data along with fixed detector data. It also explored a fixed detector placement optimization method to improve estimation precision, including uniform placement strategy and blind area supplement strategy. BP neural network was applied for data fusion of cellular handover data and fixed detector data. VISSIM microsimulation experiments showed the availability of proposed freeway travel time estimation method and detector placement optimization method.

For solving the problem of difficulties acquisition and low accuracy for pedestrian traffic parameter, a pedestrian traffic parameter extraction system based on video identified with high precision and efficiency was developed. First, basing on moving target space-time consistency principle, the system uses the inter-second order differential and edge detection of image to cut up moving targets. Secondly, analyzing moving target characteristics, extracting feature vectors, and establishing BP neural network to train and classify the targets, the system can achieve identification and detection of moving targets. Finally, the system is developed and realized that can carry on the fast and accurate statistical processing by using MATLAB software. The test result shows that the detection system’s parameter extraction accuracy, efficiency, and data processing functions have achieve the desert results, the detection accuracy reaches more than 90 percent and has strong anti-environmental interference.

The accessibility of various traffic data archived by advanced data systems in intelligent transportation system (ITS) has been a critical issue to the efficient utilization of those valuable data. The constant growth in users and data systems brings more challenges in maintaining the performance of traditional data portals. This paper presents an approach in using representational state transfer web APIs for communication between data systems. It offers concise interfaces for data requests, efficient mechanisms for data transfers, and better management of data security and access restriction. The integration of transportation data to the cloud also makes it available to more end users and utilization. The components of a web-based traffic simulation platform demonstrated the consumption and production of web APIs. An ideal schema for transportation data services was suggested for transportation research, as well as an alternative schema due to lack of existing web APIs in the ITS community.

The weaving of main road and on-ramp traffic in urban expressway is known to lead to congestion under heavy traffic conditions. This is mainly due to the complete lane changing within a limited distance. The interference between vehicles on main road and ramp causes the traffic chaos and ultimately leads to congestion. This study proposed a novel control strategy based on cooperative vehicle infrastructure system (CVIS) to eliminate this interference and improve efficiency and safety. The strategy was divided into three phases. The first was destination oriented lane control which was to reduce conflicts in the weaving area. The second was main carriageway platoon control which was to reduce the interference caused by merging in main road. And, the third was weaving lane platoon control which was to reduce the interference caused by separating from main road. Simulation results showed that both traffic and energy indexes had a significant improvement in the ideal condition, and a test with real smart vehicles in Beijing Proving Ground has been planned.

We conducted a preliminary evaluation on the performance of Bluetooth/Wi-Fi-based smartphone sensing approach for estimating pedestrian walking characteristics. A fuzzy RSSI-geometric localization method was developed based on the wireless information derived from smartphone sensing. To test the effectiveness of the method, a series of controlled laboratory walking experiments with 50 participants carrying a smartphone accessible to Bluetooth/Wi-Fi were conducted. The experiments included uni-directional and bi-directional movements of participants in a narrow corridor and each experimental setup was repeated for three times. Three smartphone sensing routers equipped with Bluetooth and Wi-Fi scanners were deployed in the experimental sites. Meanwhile, video recordings of participants’ movement from two stereo-cameras and a UAV were used for the verification of ground truth data. Before the walking experiments, essential scanning performance indices of the routers and smartphones under different environmental complexities were tested. Pedestrian walking trajectories and the corresponding motion properties were calculated through using the proposed algorithm and validated by comparison with the ground truth data obtained from the video recordings. Results suggested that Bluetooth/Wi-Fi sensing approach has low detection rate. However, it is noteworthy that the velocity estimation has a relatively higher accuracy than density and flow estimation. Furthermore, flow, density, and speed estimation under normal walking conditions have a relatively higher accuracy than that under slow running conditions, suggesting that the walking speed affects the detection rate. Moreover, contrary to intuition, the estimation of macroscopic parameters revealed higher accuracy in bi-directional movement as compared to uni-directional movement. Findings of this study might be helpful for identifying the limitations and opportunities of Bluetooth/Wi-Fi sensing devices in terms of the automatic collection of pedestrian empirical data.

Based on the analysis of spacing and timing resources of intersection imported road, an optimization model named lane assignment and signal timing is established, the objective function of which is minimum delay of signalized intersection. Visual Basic is used to write program to solve the model. Then, the author chooses the typical intersection of Hohhot as an example and simulates the current situation of the intersection in order to analyze the existing traffic problem. The optimization result is calculated with the traffic flow of the crossing road using the LDSD program. At last, the optimization plan and output, the results of intersection performance evaluation, are simulated. Comparing the evaluation of current situation and the optimization, the spacing and timing resources of intersection can be reasonably reallocated, and the average delay of intersection represent a reduction of 43%, which shows the model built up is practicable.

This paper addresses the problem of receiving a timetable for a settled network with maximal synchronization. In this paper, a model has been developed to solve the problem which attaches importance to creating a timetable for maximizing their synchronization and enables the transfer of passengers from one route to another with least waiting times at transfer nodes. Next, this problem is formulated as a mixed integer linear programming problem, and a new algorithm based on the traditional particle swarm optimization (PSO) has been developed to solve the problem in a short time. Finally, the relationship of iterative time and the complexity of the given network is qualitatively established.

This paper focuses on presenting an integrated transportation information system standard framework for a typical Chinese city in the condition that the system is built based on the existed multiple independent transportation information systems covering all transportation modes of a city. The standard framework is an open architecture which aims to facilitate interoperability among the existing transportation systems under the jurisdiction of several different departments, to enhance transportation data sharing and interoperability of existing isolated transportation systems. There are three contributions in this paper. Firstly, the common standardization requirements of an integrated comprehensive transportation information system are put forward based on that of a typical Chinese city. Secondly, a detailed three-dimensional structure of an urban transportation information system standard model suitable for typical Chinese cities is proposed. This three-dimensional structure is determined by three elements: the object, the content, and the level. Among them, the object presents all the categories according to the classification of the urban transportation management departments in China. Finally, it is the first study to illustrate that a set of standards should be applied to construct an interoperable urban comprehensive transportation system in China.

According to the existing research on the highway traffic processing models based on FCD (floating car data), researchers mostly use the traditional methods based on travel speed, and there is few research on features of vehicle running on complex road section of highways. Because a large number of vehicles run near the toll gate and service area of highways, or temporarily pull over on the roadside, it will seriously affect the accuracy of real-time traffic identification if the vehicles cannot be cleaned and eliminated effectively. In this article, we analyze the differences of positioning data between BEIDOU and GPS and discuss the features of vehicles running on the complex road section and pulling over temporarily on the roadside, setting up a recognition model for the vehicles by means of cluster analysis, and implemented data cleaning and data eliminating of the vehicles effectively. On the basis of the above work, we set up the real-time traffic identification model based on BEIDOU positioning data for highways and try to improve the identification accuracy.

Traditional intelligent driving systems suffer from low efficiency in the dynamic traffic environment because of their rigid planning modules. On the other hand, experienced human drivers can deal with dynamic situations adaptively without a complex planning system. This study aims to develop a learning-based driving system that can learn from human drivers and realize the human-like longitudinal control. The proposed system contains two main parts: a reinforcement learning module for learning human driving strategies and a PID control module for converting the strategies learned to control actions for vehicles. Experiments based on simulation are carried out to test the performance of the proposed system. A driving simulator based on the software PreScan is used to collect the driving data from human drivers and build the test scenarios. Experimental results show that the learning-based system can duplicate human driving strategies with acceptable errors in several predefined cases.

The problem of matching measured flight position points to air routes is definitely important to improve data processing tools that can assist air traffic controllers in analyzing operational situations and making control decisions. This paper proposes a simple map matching algorithm that uses a hidden Markov model (HMM) to find the most likely flight path according to the flight trajectory. The algorithm is tested on real data collected from air traffic control systems. In addition, the correct matching percentage of different sampling-rate data is tested. The HMM works efficiently on conditions where a data preprocessing method is utilized and matches with an error less than 1%.

Air transportation network optimization through adding one or more flight routes to the existing network is an effective way to increase airspace capacity, alleviate flight delay, and improve network resilience. In this paper, we introduce the link prediction methods which are widely used in other fields to predict the future flight routes of air transportation network. A weighted local naive Bayes model is proposed taking both the network structural properties and traffic volume into consideration. A case study on Chinese airport network (CAN) is conducted using the proposed models and the prediction accuracy is compared with other link prediction models. The results show that the proposed method can not only optimize the network structure but also improve the network robustness which could provide some applicable advice for the decision-makers to optimize the structure of the existing air transportation network.

This paper presents a microeconomic analysis to explore the impacts of market force and government regulation on airlines’ flight fare and safety investment. Three specific mechanisms associated with fare and safety decisions are discussed, i.e., duopoly competition, social optimum, and safety regulation. Under all mechanisms, consumers are assumed to have heterogeneous safety tastes, and airlines are assumed to make optimal decisions in a two-stage game of first safety investment and then flight fare. It is found that the differentiation of safety can be beneficial to not only commercial airlines but also the society when there are symmetric and complete information of safety between consumers and airlines. Furthermore, safety regulation from the government can assure the higher social welfare at the cost of some loss of airlines’ profits.

Air transportation, as the dominant mode for long-distance passenger travel, plays a critical role for a sustainable and greener future. Therefore, the analysis and understanding of air transportation systems is crucial. Air transportation networks and their layers as induced by spatial aggregation were investigated. Existing studies analyze network layers individually, ignoring interactions or couplings between the layers. This research aims to provide a unified understanding of the transitions among network layers, based on the worldwide air transportation networks for the year 2015. Aggregated network instances at six different levels were derived. While few nodes are important at all levels of aggregation, others only become important for few aggregation levels. Moreover, analysis results of different layers’ robustness are reported.

The expansion of high-speed rail (HSR) service is often regarded as a potentially effective, lower-carbon substitute for intercity air travel. It is experiencing unprecedented development and growth in China; however, its impact on the air market is yet to be investigated. An ex post analysis of the influence of HSR and low-cost carriers on air pricing and demand is carried out in this study by considering 23 city pairs with different lengths using panel data, of which 14 city pairs face HSR competition and 7 city pairs have low cost airlines operating. Our findings reveals that inter and intra modal competition can induce airlines to reduce fares, and a higher on time rate can lead to higher air fares. We also find that higher HSR frequencies and longer air travel time can result in reduction in air travel demand, whereas low-cost carriers have led to a modest increase in air traffic. These results have significant implications on future transport policy.

Weather is an important factor that influences the decrease of airspace capacity. Specific to the influence of weather types on capacity, this study extracts the required weather factors according to the historical weather data of the terminal airspace and employs the discretization method to simplify the weather factors. K-means clustering algorithm is adopted to classify the historical weather data and the corresponding discrimination rules of various weather types are obtained. A scenario tree is established and a BP neural network is used to recognize the data of unknown categories. Combining these aspects with an analysis of Kunming airspace, this study presents the assessment results of terminal airspace dynamic capacity influenced by various weather conditions.

This paper studied the low-temperature performance of different original asphalts, polyphosphoric acid (PPA), modified asphalt, and composite modified asphalts based on PPA by ductility test (5°C), penetration test (25°C), and bending beam rheometer (BBR). Finally, a comprehensive evaluation of these modified asphalts was carried out by combining the index of the performance grade (PG) and the cost of raw materials. The results showed that the low-temperature performance of the Esso 70# asphalt was improved, while the low-temperature performance of Iran 70# asphalt declined with the increase of the amount of PPA which indicated that the modification effect of PPA modified asphalt had a great relation with the chemical composition of original asphalt. Based on the PG index, the low-temperature performance of PPA+SBS increased first and then decreased. From the 5°C ductility value, the increase of PPA on the-low temperature performance of PPA+SBS was unfavorable. The softening point value indicated that PPA could improve the high-temperature performance of TB+PPA+SBS. Overall, the influence of PPA on the low-temperature performance of TB+PPA+SBS and PPA+SBS was not obvious and the performance of asphalt binder was mainly decided by TB rubber powder and SBS; TB rubber powder could significantly enhance the low-temperature performance of asphalt binders and reduce the production cost in way.

The fluctuation of external water level affects the phreatic line within the seasonal soaking subgrade, which directly relates to subgrade performance. Based on the theory of saturated-unsaturated flow, finite element models of seasonal soaking subgrade were built. Three measures, including low-permeability backfill, slope protection, and incomplete blind drainage, were investigated for lowering the phreatic line under four phases of water level variations. Water fluxes within the subgrade were also analyzed. The results indicated that the combination of slope protection and blind drainage was an optimal method. The water table at subgrade centerline was under the ground surface all the time and the peak water table was reduced up to 42.8%. These findings are applicable for seasonal soaking subgrade to reduce water effects.

This paper addresses a practical problem of transport infrastructure maintenance and preservation. The bus rapid transit system (BRT) of Xiamen City is the sole viaduct-based BRT system in China. It consists of exclusive lanes with complete property right and uses continuous viaduct bridges. As a result, the maintenance of the BRT bridges for Xiamen BRT is essential for its safe operation. In this paper, the real-time bridge bearing capacity of a representative bridge of Xiamen BRT is simulated and calculated using finite element analysis model. The influence lines of the shear and bending moments, and the deflection influence lines of the detected nodes, are analyzed. These influence lines can not only calculate the design value of automobile loads and the internal force value of related sections of BRT but also provide the strain and deflection of the relevant points to ensure a safe BRT vehicle operation environment.

The research conducted on highway construction cost management has been primarily based on the perspective of owner. For example, the famous whole life cycle cost management theory is about how owners control the cost in the sections of estimation, budget estimation, budget, and final accounts. Construction cost was regarded as the duty of construction enterprises. This paper analyzed the construction cost control based on the internal data of construction enterprises, proposed a method of using the RBF neural net to predict the tendency of future construction cost, calculated the ratio between the predicted cost and the actual cost according to the field data, then analyzed the characteristics of the ratio above, such as means, variances, and confidence intervals. After that, ratios such as the ratios of next month were predicted by the use of exponential smoothing. If the predicting value was between the confidence interval, the construction cost was controllable, otherwise, the construction cost was uncontrollable. This method was verified to be feasible through its use in the Guizhou communications construction group.

Structural damage recognition has always been a hot topic in transportation and structural engineering. Under the action of environmental erosion, material aging, and daily load, as well as other factors, maglev track girders will inevitably incur some damages. In this paper, damage identification of the high-speed maglev track beam is carried out by using the finite element model to analyze the structure’s dynamic characteristics. We can then verify the sensitivity and feasibility of this method. The research results and analysis examples show that because the change of the natural frequency is small after damage occurs, it is difficult to use the natural frequency to judge structural damage. Additionally, using displacement modal curvature difference and modal curvature squared difference to recognize single damage and multiple-damage can be effective to some extent.

This research aims to investigate the fatigue performance of semi-rigid asphalt pavement influenced by subgrade soil modulus deterioration caused by rainfall infiltration. The researching process could be divided into seepage field simulation and fatigue life calculation. In the first part, ABAQUS was used to establish an integrated model of subgrade and pavement and to simulate the seepage field in the subgrade. Then, the resilient modulus variations of subgrade soil were determined by combining the humidity distribution acquired and the soil modulus deterioration provided in NCHRP1-37A. In the latter part, the mechanical responses of asphalt pavements with reflective cracks were calculated and imported into FE-SAFE; thus, fatigue life of different layers of the pavement was calculated. It discovered that the fatigue life of pavement is influenced by the subgrade soil types and rainfall patterns simultaneously. Among all the pavement layers, the fatigue life of subbase layer drops to a considerable extent. Considering the rainfall patterns, rainfall intensity plays a vital role in coarse-grained soil, while rainfall duration is more influential in fine-grained soil.

This study conducted a survey at the work zone located at Beigongdaximen Station of subway line 14 in Beijing. The survey included traffic flow, time headway, and speed. Comparing two types of road work zones (occupying and non-occupying opposite lanes), this paper analyzes the characteristics of traffic flow and traffic impact in these two types. The research results show that work zones have a certain influence on traffic flow and vehicle speed. The traffic organization for work zones which occupy an opposite lane has less influence on vehicle speed than the one which does not occupy an opposite lane. The traffic impact for providing variable message signs around road work zones was analyzed based on traffic planning and simulation software, TransCAD and DYNASMART. The results show obvious and positive effect for drivers given the variable message board, especially when displaying proposed speed and optional detours.

This article describes the guideway switch used in maglev Changsha airport express line in Hunan. The Changsha project was the first engineering application of the urban maglev guideway switch since it was designed and tested by NMTC a decade before. In order to meet the needs of the whole maglev system, NMTC made several optimizations to the original design and made the switch more reliable and stable.

Based on vehicle-track coupled dynamic theory, a heavy haul vehicle-ballasted track-subgrade dynamics model was developed, in which the coupling effect between the vehicle, the ballasted track, and the subgrade was taken into full consideration. The effect of the variation of the wavelength and the wave depth of rail weld irregularity on the dynamic response of heavy haul railway track-subgrade system was analyzed by a combined numerical simulation method of new rapid explicit-Newmark implicit integration method. The results show that for the weld irregularity, the more wave depth caused more vertical dynamic displacement and more vertical dynamic stress. The increase in the wave depth of the shortwave also has significant effect on the vibration acceleration of the system.

The PC void beam is the widely adopted bridge girder type in expressways. The bridge will be partly damaged under the conditions of developed traffic, over-service, and maintenance shortage. The main issues are resistance loss in single beam, connecting loss between beams, void loss in expansion joints, cracks on abutments, and failing in bearings. Among them, the connecting loss is one of the most serious issues. In this paper, both site test analysis and FEM mode analysis will be adopted on a temporarily constructed bridge with several recycle void beams. The results show that the strengthening method by extra steel beams or ex-dosed pre-stress strand mesh will both enhance the transverse connectivity between the beam and its neighbors.

The purpose of this study was to evaluate the pavement fatigue and rutting performance with an accelerated pavement testing (APT) test for the in situ asphalt pavement. On a selected expressway ramp in Shanghai, the full-scale accelerated pavement test was carried out during which the average air temperature was 30.5 °C. During the APT test, the portable seismic property analyzer (PSPA) was used to measure the seismic modulus, and the MLS profilometer driver-p2003 was used to measure the profile of the in situ pavement. The PSPA test results showed that the pavement modulus decreased with the increasing of accelerated loading passes and could be used to analyze the fatigue damage parameter of the in situ pavement. The pavement profile measurement results showed that the pavement deformation developed with the increasing of accelerated loading passes and was used to calculate the pavement rutting depth. The field measurement results during the APT test well illustrated the fatigue and rutting performance of the in situ asphalt pavement which are useful for making the maintenance and rehabilitation decisions for expressways.

When the rigid substratum is set in multilayered elastic system to back-calculate layer modulus, the matching degree of deflection basin can be improved. Therefore, the depth of rigid layer should be set reasonably. A regression model based on Bousinesq’s equation is developed, and the independent variable is the horizontal offset from load of the zero-deflection point. The rigid layer depth also can be treated as an unknown parameter and be directly obtained through modulus back-calculation. The measured deflection basin data on the top surface of subgrade were analyzed to simply the problem. The results show that when objective function in modulus back-calculation was set to absolute error, the prediction depths to rigid layer from the regression model were very similar to the results obtained through modulus back-calculation, and compared to objective function set to relative error, the back-calculated subgrade modulus had lower sensitivity to the depth to rigid layer.

Pavement roughness is an important performance indicator for urban roads. Existing roughness evaluation methods are adapted from the highway system, which do not consider the pavement partial unevenness of urban roads such as manhole covers or repair patches. Significant differences can easily be found between the actual driving experience and roughness evaluation results. A survey investigated 428 urban roads in central Shanghai was launched, and the results show that manhole covers account for the largest amount of pavement partial unevenness types. By using the original pavement elevation data measured from the laser roughness meter, characteristics of manhole covers are analyzed and the critical thresholds (i.e., 4 mm) are also calculated. The automatic identification, location, and classification of manhole covers in the pavement vertical profile is realized on this basis. Ultimately, an IRI correction formula considering manhole covers is proposed, while a new roughness evaluation method is also established to evaluate the influence of manhole covers.

Relying on the reconstruction project of Togo No. 1 Highway, the typical distress of asphalt pavement in long and steep sections was investigated and its mechanical characteristics were analyzed. The results indicated that rutting and interlaminar slip were the typical distresses of long and steep sections, and these distresses were mainly caused by asphalt layer shear stress and interlaminar shear stress. Then, an asphalt pavement design approach of long and steep sections was proposed from the aspects of asphalt mixture shear property and interlaminar slip property. Next, the adopted mechanical control parameters, the standard of these design parameters, and the designing process were given. The design principles and recommendations on asphalt pavement design of long and steep sections were provided. Finally, the example of the asphalt pavement structure design of the No. 1 Highway was given.

The main objective of this study was to evaluate the fatigue characteristics and remaining fatigue life of field emulsified asphalt cold recycled (EACR) mixture. In this study, the field EACR mixture specimens were cored from car and truck lanes that suffered different traffic loads. The specimens then underwent the indirect tensile strength test and indirect tensile fatigue test. Moreover, field hot mix asphalt (HMA) mixture which covered the EACR mixture was also tested for comparison. Test results indicate that difference in traffic loads has almost no effect on the indirect tensile strength of mixtures irrespective of mixture type. Fatigue test results indicate that more traffic loads will lead to greater reduction in fatigue life for both EACR mixture and HMA mixture. In addition, an analytical method based on laboratory and field correlations was developed to determine the shift factor for predicting the remaining fatigue life of field EACR mixture as well as HMA mixture. In this study, the shift factor values of EACR mixture and HMA mixture were calculated as 107.16 and 13.71, respectively.

With the rapid development of rail transit, dynamic detection technology of alignment survey is playing a more and more important role. It is not an exception to Shanghai Maglev line (SHA). For SHA, long-wave deviation is the main result measured obtained from dynamic detection equipment. In recent years, long-wave deviation result has been applied not only for formulating an alignment maintenance scheme but also for analyzing alignment improvement degree. Because application of alignment dynamic detection method is new to SHA, it is necessary to research the evaluation method of alignment maintenance effect. Research on effect evaluation method such as influence area around the adjusted pier, change ratio of long-wave deviation value before-and-after alignment maintenance, readjustment scheme of alignment maintenance has been made. According to the engineering practice of SHA in 2015 and 2016, it is proved that the three kind methods are feasible.

This paper proposes a new index named “bridge deterioration rate,” using recent 10 years bridge data in Shanghai. Significance and correlation analysis were conducted to finally derive the top three main influence factors: bridge age, location (road grade), and deck pavement type. Accordingly, the bridge age is considered as independent variable. The bridge deterioration models were constructed with respect to the bridge location and deck pavement type separately. The characteristics of diverse bridges and deck systems were analyzed based on the proposed model. Finally, the distribution of bridge state in next two years was predicted. The result shows that both the whole bridge and deck system deteriorate fast followed by slow (like a concave curve). The deterioration mode under distinct groups differs from each other, especially the case in diverse deck pavement types.

In the future, girder-on-bridge system with noise barrier will be an optional type for high speed maglev system. Noise barrier is the important characteristic of this system. Research on its aerodynamic effect in normal condition is necessary and has been developed. Extreme weather conditions, including snow covering the bridge and icing on the front of maglev train, should also be considered in the future project. In this paper, CFD technology and dynamic scaled modeling test are used to analyze the effect on the pressure of noise barrier. Summary of the results shows that compared with the normal condition, the extreme weather conditions have less additional influence on the aerodynamic effect except the snow thickness. All of the work will contribute to the further research about the load requirements for girder-on-bridge system.

Finite element models can be used to analyze the structural behavior of bridge deck pavement under heavy traffic conditions by analyzing the influence of the reinforcement ratio, the thickness of the asphalt layer and the concrete layer separately. The results show that the thickness of the asphalt layer has more impact on the tensile stress and shear stress in the bottom of both the asphalt layer and the bottom of concrete layer. The thickness of the concrete layer has more impact on the tensile stress and shear stress in the bottom of the concrete layer and the bridge deck. According to the results, a new type of 4 cm asphalt layer + 12 cm concrete layer is proposed.

Fiber reinforced concrete has been widely used in civil engineering; however, aramid fiber use is still relatively new. An experiment was conducted to study the uniaxial compressive constitutive relationship of aramid fiber reinforced concrete. The volume fraction of aramid fiber tested varied between 0%, 0.2%, 0.4%, 0.6%, 0.8%, 1.0%, and 1.2%. Based on the analysis and comparison of complete stress-strain curves, the constitutive equations are given and the associated fitted data curves consistent with the experimental results. The parameter B in these equations is related to the fiber volume fraction. The experimental results also indicate that the aramid fiber volume has some influence on the workability of AFRC but has little effect on the compressive strength. All of these experimental and theoretical results are significant to the use of fibers in concrete design.

Roads constructions adjacent to the high speed railway bridge will cause displacement of bridge piers and girders as well as deterioration of the geometry irregularity of the slab track. As a result, the operational safety of the high speed train and the ride comfort of passengers will be influenced. In order to determine the reasonable threshold of displacement for a bridge pier, a 32 m-span simply supported girder bridge with the CRTS (China Railway Track System) I-type and II-type slab track systems is simulated by finite element method (FEM). The roles of bridge support are considered to describe the interaction between girders and piers. The deformation and stress of CRTS are employed as the criteria. Results indicate that the practical threshold of pier displacement of the high-speed railway bridge should be controlled within 2 mm.

Road section division is a prerequisite for pavement management and maintenance. The existing research mostly divides road sections arbitrarily without considering the significance of each influencing element. This work presents an improved ordinal cluster division method. The weight of each maintenance factor including equivalent standard axle loadings (ESALs), pavement thickness, pavement roughness, and pavement damage condition is determined by the fuzzy analytic hierarchy process (FAHP) approach. Based on the weight system, the maintenance section is divided utilizing ordinal clustering method. It is solved by recursion formula of dynamic programming with objection to minimize the sum of internal variation of all sections. A numerical case study of the G6 highway was conducted, which demonstrates the effectiveness and efficiency of the proposed method.

A nonlinear emulation model of continuously reinforced concrete pavement (CRCP) was established by FEM to analyze mid-depth punch-out mechanism to provide a theoretical basis and reference for pavement design and punch-out control. Thermal sensitivity of concrete material and steel as well as bonding performance between the reinforced concrete and steel were accounted in the model. Influence factors of horizontal cracks, such as coefficient of thermal expansion, elasticity modulus, the bond-slip relationship between reinforced concrete and steel, and the location of steel bars, were researched to analyze how horizontal and longitudinal cracks appear and develop into punch-out in real projects. Results show that thermal stress is far greater than warping stress and the maximum main stresses of them were located in the same place; aggregate of low modulus and low expansion as well as reasonable steel location can help to control cracks; modulus of steel shows little effect to horizontal cracks.

The basic object of damage diagnosis for bridges is to ascertain the damage occurrence, location, and severity based on extracted modal parameters from the monitored vibration response. In practice, modal parameters are extensively affected by ambient environment, especially temperature. In this paper, a novel damage diagnosis approach was proposed to identify the structural damage under the influence of temperature. Firstly, temperature field in the concrete slab was numerically simulated through thermodynamic analysis. Secondly, mechanisms of temperature affecting modal frequencies were analyzed, and the reference finite element model was constructed considering stress stiffening effect and temperature-dependent characteristics of concrete. Finally, damage was simulated by adding concrete specimens onto the surface of concrete slab. And, damage of concrete slab was identified using probabilistic finite element model updating technique. The feasibility of the proposed approach was demonstrated using experimental study of concrete slab monitored outside the laboratory.

In order to meet the requirements of heavy load traffic and improve the performance of asphalt pavement, experiments are carried out to study the nano inorganic mineral modified asphalt in this paper. In addition, the optimum asphalt content is determined by optimizing the asphalt mixture ratio. Furthermore, in terms of the pavement with different dosage of nano montmorillonite mixtures, laboratory tests are conducted to study the performance, i.e., high temperature stability, water stability, and low temperature cracking resistance. By comparing the content of nano montmorillonite 0% and 6%, the results show that the DS value and TSR value increased by 67.8% and 11.6%, respectively, indicating that adding nano montmorillonite can effectively improve the high temperature stability and the water stability of the asphalt mixture. In addition, the bending stiffness modulus reduced by 2.4% and the flexural tensile strength (strain) increased by 6.3% (7.6%), which indicates that the content of nano montmorillonite has little effects on the low temperature cracking resistance.

This paper introduces an approach to determine the coefficient of thermal expansion (CTE) of road concrete using an optical lever system using CTE of 144 concrete specimens. The coarse aggregate in these specimens contains 8 kinds of stones from 19 origins. Measured data was satisfactorily verified by AASHTO TP60. The results showed that CTE of road concrete was between 4.76×10^-6/ °C and 12.1×10^-6/ °C and that coarse aggregate type had a great influence on CTE. When aggregate content was equivalent, the higher CTE of the aggregate, the higher CTE of concrete contained that aggregate. The sensitivity of concrete for different temperature ranges and the same temperature gradient had no significant difference, such as 20 °C to 30 °C and 30 °C to 40 °C. Meanwhile, the amount of concrete expanded when the temperature increment was equal with concrete shrinkage when temperature reduces.

As the main emergency rescue channel during a disaster, highway damage has important influence on rescue work. However, there is a complicated internal process of highway damage under major natural disasters, and its failure mechanism and estimate were difficult to describe accurately by a mathematical model. Considering the advantage of cloud models in solving uncertain problems, a cloud model was introduced into the digital characteristics research of road damage and application in this paper. First, the uncertainty of highway damage was analyzed, and the input format of a damage sample were given based on a backward cloud generator, then the algorithm steps of cloud model digital features were put forward. Based on the highway damage survey data in 2008’s Wenchuan earthquake and the water damage data of national and provincial roads in Yunnan Province, some cloud model digital characteristics of the road subgrade and road were generated according to a backward cloud generator algorithm. This highway damage cloud model figure was generated by MATLAB software, which implements the mathematical description and characterization of road damage uncertainty. The research provides an understanding of highway damage analysis and forecast, which could support disaster relief efficiency and reduce the losses of property.

To determine the internal deformation and damage mechanism of asphalt pavement materials under tension, a double punch test which can better reflect the tension state of asphalt mixture is introduced in this paper. The repetitive loading test of asphalt mixture was carried out by using computerized tomography (CT) and scanning technology synchronously. The movement law of the aggregate and mortar in the asphalt mixture, the movement trajectory of the internal structure particles, and the mortar in the asphalt mixture are collected by the CT technology. The displacement locus of the different loading stage is calculated. The results show that the displacement trajectory is small and almost negligible at the initial stage of loading. As the loading progresses, the displacement gradually increases, and the trajectory of the particles and mortar moves gradually. For the double punch test, the calculated results are in agreement with the measured data. The limit analysis of the cylinder double punch test is carried out, and the calculated results are in agreement with the measured data.

Serving an important role in the public transportation system, modern tram tracks developed very quickly in recent years in China. However, there is still no reasonable method to calculate the subgrade settlement of the tram track. In this paper, an explicit-implicit iterative algorithm is employed to predict the subgrade settlement of a newly constructed tram track in Shanghai. Results indicate that the subgrade settlement of the tram track can be well controlled by a recommended pile-slab structure. The consolidation settlement of the tram track subgrade is 8.2 mm, and the cumulative plastic deformation of the subgrade induced by tram track dynamic loads is 3.3 mm. The total settlement of the tram track subgrade is 11.5 mm, which can meet the requirement of the post-construction settlement control standards of the tram track.

An artificial neural network fatigue life prediction model with single hidden layer was built under Matlab environment. Three independent variables: asphalt content, air void, and strain level, and one outcome: fatigue life, were included in this model. Four point bending fatigue test was carried out under different conditions to obtain training data for neural network, and a total of 27 sets of samples were organized for training. The R² of the neural network fatigue life prediction model was 0.91. To make a comparison, the same data was used to fit the empirical fatigue life predicting model. Analysis exhibited that the neural network model prediction was better than the empirical fatigue life predicting. Meanwhile, the neural network model showed an ability to weigh the impact of variables.

Modification of polyphosphoric acid (PPA) can significantly improve the high temperature properties of asphalt binders. However, its influence on the low temperature property is not yet clear. In this study, the bending beam rheometer test (BBR) was conducted to evaluate the low temperature properties of PPA hybrid asphalt binders under different aging conditions. Furthermore, the FTIR test was used for chemical analysis of the asphalt binders. Finally, thermal stress restrained specimen tests (TSRST) were used to evaluate the low temperature performance of the asphalt mixture. The results indicated that PPA had a positive effect on the low temperature properties of asphalt and the optimum ratio was obtained. Meanwhile, PPA also showed effects on anti-aging both in the BBR test and chemical analysis. Finally, TSRST test results showed that addition of PPA can slightly improve the low temperature performance of base asphalt mixtures.

Buton rock asphalt (BRA) is a natural asphalt modifier that can improve different properties of mixture. The traditional mixing process for BAR mixture is to mix aggregates with BRA firstly, and then mix them with heated base asphalt and mineral powder in that order. However, it results in an unevenness of asphalt film thickness, which can influence the performance of mixtures. Therefore, the primary objective of this research is to develop an advanced mixing process for BRA mixture. Six mixing processes were attempted. Air void, dynamic stability (DS), indirect tensile strength, and tensile strength ratio (TSR) were selected as the evaluation parameters. By comparison, the optimum mixing process is changed by adding the BRA and base asphalt in different orders. Lastly, contact angle measurements and inverse gas chromatography (IGC) were conducted to analyze the adhesion work in different mixing processes, which can explain the improvements of the advanced mixing process.

This paper presents a decision model in a single-period two-stage supply chain for perishable products with long lead-times, short selling season, and high demand uncertainties, which can achieve the collaboration between the supplier and the buyer. This paper obtains the optimal initial order quantity and the number of the purchased options. The optimal option pricing decision of the supplier is confirmed via a numerical study of various parameters. This paper also derives that the profit of decentralized supply chains using the call option supply contracts is higher than that of the base model, i.e., newsvendor model, but lower than that of a centralized decision model. Thus, in the numerical study, it has been proven that the supply chain’s profit can be improved by using call options.

At times, drivers need to pass a designated area on the way to a destination. This paper established the area-through path planning model of the road network, ensuring path planning results pass the designated area. The area-through path planning is realized in two steps: First step is the virtual destination path planning where we set a virtual destination first, and then the A* algorithm is used to guide the path through the designated area. The second step is the direction-changing path planning that leads the path planning from the designated area to the destination through changing path planning heuristic costs. The experimental results confirmed that in urban areas where the algorithm helped users to pass the high-grade roads in the designated areas. This suggest the real-time path planning can meet the demand of embedded navigation equipment.

The ship berthing time is an important indicator to measure the operation efficiency at the port. Characterizing the distribution of the berthing time can quantitatively describe the patterns of ship operations. In this paper, large amounts of data of ship operations at the port are obtained from the automatic identification system (AIS). The data mining technology is applied to abstract the information of ship berthing time in an accurate way. Combined with the data of Ningbo-Zhoushan Port through 2015, we analyzed the berthing time of different types of ships (e.g. container, chemical and oil, crude oil, bulk and general cargo, iron ore, coal) and built their distribution model. The research result in this paper can provide the basis, and the reference, for the government to schedule the ship at the port and improve the port operation efficiency.

Quay crane scheduling and yard truck scheduling are two highly interrelated important problems in container terminals. Most studies have attempted to optimize each of these two problems independently. However, there exists delay in quay crane or yard truck that would decrease handling efficiency; so, this paper studies quay crane scheduling and yard truck scheduling in an integrated manner. To increase the utilization rate of resources and to reduce the rate of yard truck empty running, this problem uses the dual-cycling strategies. Due to the intractability of this problem, we design a genetic algorithm to acquire the near-optimal solutions.

To improve the operational efficiency of container terminals, optimization of the continuous berth allocation and quay crane assignment problem is studied in this paper. We investigate the combined problem as a bin-packing problem and propose an integer programming model for the problem. The proposed problem formulation includes important practical considerations such as the increase in the handling time if vessels are not moored at their desired positions. Factors causing the loading and unloading time loss due to berth position deviation are considered. The objective is to minimize the general time loss considering the time and space factors together. Due to the inherent computation complexity, a genetic algorithm is designed to obtain near optimal solutions for the problem. The results of numerical experiments show that the proposed approach is effective and efficient to solve the problem under study.

Aiming at the problem of the low capacity and high cost of the distribution mode of auto parts, low complexity time window, not meeting the demand of parts of JIT, this paper theoretically proposes a milk-run model based on 3PL, and per the characteristics of the model, adding dynamic supplier delivery time window conditions, optimization model is established using path search algorithm taboo. Combined with the data of 14 suppliers of FAW international logistics center, this algorithm is used to plan the optimal path, making cost savings of 67%. The application further verifies the milk-run model to balance the parts supplier transportation cost reduction and the demand of parts of automakers JIT, providing a way for China's automobile manufacturing enterprises to implement and improve the milk-run parts delivery project.

The artificial bee colony (ABC) algorithm is a kind of swarm intelligence optimization algorithm which has been developed rapidly in recent years. It is widely used in solving the vehicle routing problem, production scheduling, function optimization, etc. This paper proposes an algorithm that is based on an improved neighborhood search strategy, dynamic cross strategy, and an improved forced reservation elite tournament following the strategy of the original ABC algorithm (ABC-NDCT). Finally, the ABC algorithm, the ABC algorithm based on "Neighborhood Search Improvement Strategy" (ABC-N), and the ABC-NDCT algorithm are tested by using the first 61 points (including a base) in the Solomon RC206 instance. Compared with the results of a GA algorithm, the ABC-NDCT algorithm is more effective and applicable.

As traditional cold chain logistics (CCL) location models neglect customers’ satisfaction (CS), this study proposes a bi-objective location optimization model which considers the cost and CS service level of the CCL. The proposed model is to maximize CS and minimize the cost of CCL, which is solved by using epsilon-constraint to transfer the bi-objective model into a single objective. In addition to the traditional cost of handling, and transportation and storage, the refrigeration and damage cost is also taken into account; the CS is the probability of satisfying the requirements of the customers. In the end, this paper designs an example and verifies the accuracy and rationality of the model using LINGO11.0 software.

The main goal of emergency logistics is the time of rescue. Two key problems that need to be solved in emergency logistics are the location of depots and the routes of vehicles. However, many scholars have studied these problems separately. It is reasonable and necessary to study the two problems simultaneously. At present, many literatures on location-routing problem are concentrated in the general logistics. Relatively speaking, there are few scholars to study the LRP in emergency logistics. In this paper, location-allocation problem and vehicle routing problem will be studied simultaneously. Combined with the actual purpose of emergency rescue, we propose a model whose object is to minimize waiting time of recipients. We also considered the time window constraints and the uncertainty of routes. The feasibility and effectiveness of the model are proven.

For the purpose of risk equity during transportation of hazardous materials, a vehicle routing optimization model is proposed with consideration of risk equity, and an algorithm is developed for solving the model. An example is used for empirical analysis, which presents the vehicle routing optimal solution. In this way, the model is tested to be reasonable and applicable in giving certain guidance on decision-making of vehicle scheduling for hazmat transportation.

There is a price competition between product manufacturers and material suppliers (buyers and sellers) that has already provided a trading relationship during the procurement logistics. Game analysis of price negotiations is conducted by considering the market factors such as supply-demand relationship and product sales. Then, a complete information static game model is established and a mixed strategy Nash equilibrium is obtained. A conclusion can be drawn from the analysis that the probability distribution of selecting each strategy of manufacturers and suppliers is highly related with external market environment. This finding provides theoretical reference about how to choose the pricing strategy for both manufacturers and suppliers during the procurement activities.

Recently, stability under disruptions has become an important research area in supply chain systems. In this paper, stability of supply chains under disruptions is studied based on a predator-prey model. The paper attempts to build a dynamic model to capture the interdependent correlation between disruptions and stability of the supply chain. This model includes two nonlinear differential equations representing dynamics of the two independent states, namely, disruptions and supply chain systems. Then, the interdependent correlation of these two species are then analyzed and forecasted, and the practical significance of the equilibrium point is explained by differential equation stability theory. This stability of a supply chain under disruptions should be lost if the equilibrium point is unstable. The results demonstrate that the predator-prey model can be used effectively to formulate the interdependent correlation between disruptions and stability of a supply chain, and the stability of the supply chain under disruptions mainly depends on the equilibrium points. The research is meaningful to enrich the supply chain management system and to effectively prevent and respond to disruptions, as well as reduce losses.

Multimodal transportation can improve transportation efficiency and further reduce logistics cost. The reloading efficiency between different transportation modes becomes more and more important in multimodal transportation. The operational efficiency in seaport terminals is of paramount importance for international multimodal transportation, so it is necessary to research the berth allocation problem. This paper addresses the berth allocation problem under a cost-based objective. Its goal is to allocate the available berthing positions and schedule the berthing time for the container vessels in order to minimize the total operation cost in seaport terminals. Total operation cost is mainly embodied with additional travel cost and penalty cost due to non-optimal berth allocation. In previous papers, the authors have developed and presented the berth allocation problem based on time limitation, whose goal is to minimize the total waiting time of all vessels. In view of the steadily growing cost in seaport terminal operation, this paper proposes a cost-optimal-based model for the berth allocation problem. A numerical example is conducted by mixed integer programming and calculated by AMPL. The result shows that the optimization model is feasible.

With the rapid growth of the number of logistics parks in China, there rises a problem of logistics park’s vacancy, so it is necessary to determine the radiation’s scope of the logistics park. Based on the theory of breaking point and its field strength, this paper proposes a fusion breaking point model to identify that the radiation’s scope of one logistics park in logistics parks. Factors, such as the comprehensive capacity of logistics park, traffic location, and service type, etc., are considered. This paper uses the improved breaking point model to determine the radiation’s scope of logistics park from the qualitative perspective and quantitative perspective. By calculating the logistics park’s breaking points and field strength of the breaking point with SPSS, the radiation’s range can be determined. This paper provides a methodology to determine the radiation’s scope of logistics park and lays a foundation for logistics park planning and design.

In today’s environment of Internet+, people’s consumption patterns continue to upgrade, traditional logistics industry to adapt to the development of the times, is committed to the development of innovative logistics situation and the establishment of intelligent logistics. “The Urban freight taxis” is developed with the depth of the Internet out of the new form of short-distance logistics and distribution. “the Urban freight taxis” is committed to solving the logistics of the last mile of the operation, establishing information path between the driver and the customer through the network of vehicles. And, through the innovative starting price plus mileage innovation fee model, making the cost more transparent. Because this kind of form is newly developing, there are many deficiencies in the operation and management, and the development of “taxiing car” and some other car software has a great similarity with "the urban freight taxis.” And in this paper, we analysis matching of goods and car through people to match the car, so as to “the urban freight taxis” to make a reasonable planning development.

Double eleven online shopping day (DEOSD) made great achievements. However, a series of problems have arisen. Based on the supply and demand analysis, the key to the problems are identified due to the fact that the logistic supply cannot meet the sudden increase in logistic demands. This paper proposes to guide demands and integrates logistics supply to match demands based on the analysis and forecasting. From the standpoints of logistical demands, this paper proposes that the amount of logistics demands is just that of information processing. Consequently, the logistic information network becomes the key to solving the problem. From the standpoints of logistics supplies, we studied the case of the Cainiao network based on a data-driven logistic platform of social coordinated operations.

This paper proposes a bi-level decision making process for an urban freight routing problem considering the interaction between background traffic and freight trucks. The urban freight routing problem is formulated as a Stackelberg game, where the followers’ game is a static traffic equilibrium among private vehicles and the leader’s game is a minimum-cost freight routing problem subject to the lower-level equilibrium and additional constraints. The interactions are captured through the link cost function employed in the user equilibrium accommodating the truck use. User equilibrium is solved by the fixed-point algorithm, where candidate routes are generated by the sweep method and the k-shortest paths method. Compared with two traditional methods, this methodology shows a reduced cost of freight trucks in test networks. The reduction depend on factors including the demand pattern of private vehicles and the network. However, this methodology can be computationally more expensive, which can be improved by parallel computing techniques as demonstrated in this paper.

In recent years, many cities in China have introduced a brand-new way of freight transportation—urban freight taxi. Firstly, the paper elaborates running status of freight taxi since the Shanghai government carried out the opening price policy in 2015. Secondly, evaluation on price is analyzed from the two perspectives—the enterprise and the driver. Considering the expenditure and income of the two sides, this paper assesses the rationality of the existing price. Finally, some necessary suggestions are embedded into the paper on price management of urban freight taxis in Shanghai. This study is based on huge amount of statistical data collected in Shanghai in the past two years. The results have practical value on macro-control of the freight taxi market.

Compared with passenger transport, study on freight value of time (FVOT) has been largely neglected. In China, a large scale stated preference studies for freight transport to identify FVOT has not been done. Researches focus on inland leg of international containerized transport is even fewer. This paper intends to fill that gap by estimating the FVOT of railway for the inland leg of containerized transport in China, using a mixed logit (ML) model and stated preference techniques. The corridors Bengbu-Shanghai and Yiwu-Ningbo in China were selected for the survey. The FVOT obtained in this paper (€0.31 per ton and hour) is much lower compared with the empirical evidence identified for containerized shipments in the door-to-port leg in European country. The FVOT in Bengbu-Shanghai corridor (€0.32 per ton and hour) is equal to that in Yiwu-Ningbo corridor. And FVOT is quite different across different decision-makers.

In this paper, the observed trip chaining behaviors collected in the pilot survey on freight vehicle traveling characteristics in Northwest Logistics Park, Shanghai, are analyzed. The selected indicators include the number of trip chains, the number of stops, the distance of trip chains, and the distance between the stops and the base. The analysis for each indicator is performed based on the type of vehicle to reveal the underlying causes. Regression analyses are performed between the selected indicators and finally correlations are computed. Results show that a significant positive correlation between the number of trip chains generated and the two indicators (the number of employees and the storage area). The number of trip chains and the number of stops per trip chain have a negative correlation. In addition, with the increase in the number of stops, the distance of connecting trips decreases gradually and shows an approximate exponentially decreasing trend.

We investigate the stochastic and dynamic vehicle routing problem with crowdshipping (SDVRPC), in which temporary couriers are willing to make a single delivery in return for a certain reward. This study is motivated by the trend of crowdshipping, especially in the field of meal delivery service. Since the shipping requests and temporary couriers are revealed over time, a discrete event simulation model is used. It formulates a mixed integer program with the objective of minimizing customers’ average waiting time (i.e., to maximize service level) by optimizing the route of regular couriers. Numerical experiments were conducted with stochastic generated data and diverse parameter combinations. Computational results demonstrate the validity of the method and indicate crowdshipping’s potential to reduce delivery cost and enhance service level.

Highways, railways, and waterways transport containers between hinterland and coastal ports. This study used the Nest-Logit to model the system. Port route density, access facilitation, transportation time, transportation cost, punctuality, and flexibility were chose to be the variables of utility function on the basis of market surveys. Transportation chain in PHS of Ningbo Port and its competitors was studied to calibrate the parameters of the model and estimate the probability chosen by the shippers in each transport chain. Also, the model can be used to obtain the advantaged hinterlands of sea-railway transport in Ningbo Port by using the investigation data of the container shipper in the hinterland cities. Finally, probability of the improved performance of railway transport chain was analyzed to provide a quantitative analysis method for the sea-railway transport development of Ningbo Port.

A logistics network with cross-docks can simplify the logistics operation process, reduce inventory costs, and improve the efficiency of transportation. In the traditional location design of a cross-docks network, only the cost of transportation and fixed construction is being considered. The carbon emissions costs caused by environmental pollution, in the process of transportation, has been ignored. In order to make the location design meet the demand of economic sustainable development, a location selecting algorithm considering carbon emissions for cross-docks of multiple products is presented based on a particle swarm algorithm. Firstly, the facility location problem is analyzed, then a corresponding mathematical model is set up, and finally the problem is solved by using the particle swarm optimization (PSO) algorithm. The example shows that the proposed algorithm is effective and feasible. The model is stable when the parameters change, and the problem can be well solved by the PSO algorithm.

To study the developmental level of current environmental logistics, four indicators are considered; logistics policy, environmental logistics economy, resource conservation, and sustainable development. This paper combines the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation (FCE) to formulate the environment logistics development level evaluation model. The model measures the important factors which affect green logistics development levels. The research results reveal that green logistics economy is most significant, and the results may guide the green logistics development direction to provide a reference for evaluation of the development level of green logistics.

This paper firstly analyzes the trend in China’s cargo volume, cargo turnover, and their growth rate since the twelfth five-year plan. Then, it compares and analyzes the trend and features of the development of cargo transportation in China and in the United States with respect to economic development, index of total freight tonnages, and evolution of transportation structure. Finally, it puts forward countermeasures and suggestions to respond to the new trend for the development of cargo transportation under new normal from the perspectives such as cargo transportation demands and increase, transportation structure, transportation organization mode, freight representation, freight space distribution, etc. Thus, it hopefully provides references and basis and basis for gaining a full understanding of the development of cargo transportation in China under new normal.

With the rapid development of economy, port logistics has become a new economic growth point in Inner Mongolia. Firstly, the development status of the port logistics is analyzed from infrastructure, logistics location layout, logistics development, and other aspects in this paper; then, the scale of CEKE port logistics is obtained through establishment of Grey model GM (1,1) and the second moving average model. Finally, policy measures for the development of port logistics are proposed from four aspects on the basis of forecasting conclusion.

Containerization is the greatest revolution for transportation industry in the 21^st century, and container accumulation is based on national economic development; however, there appears no study focusing on the evolution rules between containerization and national economic development. The historical data of containerization and national economic growth for 10 high-income states and 10 middle-income states are selected in this study, and then after deductive analysis and inductive reasoning, the container accumulation pattern is summarized. Results show the container transportation is experiencing sustained growth in different countries; container generating density shows different growth trend in different income stages, and the density level is relatively low and stays stable for high-income states, while it shows relatively high level and fast increase for middle-income states; the relationship between the container generating density and per capita income is found to be an inverse U-shape curve, while systematic differences are also noticed between different types of states; the turning point for China’s container generating density has already appeared, and with the slower economic growth rate and economic restructuring in the near future, China’s container accumulation density will experience a medium-low growth.

Trains normally run as scheduled in a non-disrupted situation. However, due to the external and internal factors, trains may deviate from the original timetable during daily operations. To this end, the involved dispatchers are required to select the proper strategy to reschedule disrupted trains efficiently. We focus on the train rescheduling in a seriously disrupted situation where a track segment is completely blocked for a relatively long period of time, e.g., 2 hours. In this disrupted situation, trains cannot pass the blocked segment during the disruption. To properly handle the disrupted trains, mainly two train rescheduling strategies are applied: disrupted trains waiting strategy and disrupted trains turning back strategy. This research will analyze the feasibility of these two strategies and quantitatively compare their performance. Train rescheduling problems under the two strategies are formulated by two mixed integer programming models. Finally, we test our models on the Chinese high speed railway.

In the “13th Five-Year” Conference, National People’s Congress Wang Zhan suggested that maglev be incorporated into China’s high-speed rail technology strategic reservation. Maglev transport is an emerging, environmental technology and has improved research and industry from all respect. From the method of SWOT analysis, the paper analyzed the strength, weakness, opportunity, and threat of maglev transport from the aspect of the supply of maglev transport’s industry, infrastructure, technology application, and market, as well as the demand of maglev transport. Finally, some proposals for the development of maglev transport are made.

In order to analyze the mechanical factors of the elastic bar fracture of two kinds of fastening, single-toe fastening and Type III fastening, the finite element (FE) models of the elastic bars of above two kinds of fastenings were established. The simulation shows that the stress of the elastic bar by its vertical displacement and installation position influenced its fracture. For the single-toe elastic bar, the stress of the elastic bar in excessive installation was 2.1%-3.5%, 5.0%-7.7% larger than that in standard installation and in incomplete installation, respectively. For the type III elastic bar, the maximum stress of incomplete installation was close to that in standard installation. Comparing the elastic bars of two kinds of fastenings in normal installation, the stress of the elastic bar of Type III fastening was smaller, 46.3%-47.1% than that of single-toe fastening.

Travel time has a critical influence on the capacity of bus systems, as well as the travel efficiency of passengers. The signalized intersection is the critical location where delay occurs. To improve bus operational efficiency, the method of speed control between nodes under the CVIS environment has been proposed in this research. The main purpose is to reduce the delay of bus vehicles at intersections, with the restriction of bunching prevention and stop capacity by controlling the fleet size. This will improve service reliability and reduce the stop delay. The case study shows that the optimized model plays a part in reducing vehicle delays and promoting service reliability, which has important practical value for improving the quality of public transport services.

In recent years, China’s major cities began to enter the “era of subway” where new urban public transport systems were developed whose backbone is rail transit, mainstay is common transit, and supplement is individual traffic. This paper designs a non-standard assignment model to study the transfer condition between common transit and rail transit. With the help of this model, transfer condition between them will be improved, and public transport system can better play its role. In this model, the reasonable matching between bus stops and station entrances is a key design consideration related to the transfer time of passengers. The congestion of station entrance is another factor. Finally, the model is tested and verified on the Yongningmen station of Xi’an Metro line 2 and surrounding bus lines. The result of optimization tool Lingo proves the feasibility of the model.

With the rapid development of high speed railway, the transport capacity of railway system has increased substantially, in which case passengers’ requirements for travel quality are more demanding. Therefore, grasping passengers’ satisfaction with rail travel is of great significance. The majority of current researches about railway passengers’ satisfaction focus on a single railway line, while the comparison among different lines is neglected, as well as the effects of access and egress journeys. The paper takes three typical lines (high speed railway, intercity railway, and conventional railway) in Shanghai-Nanjing Corridor as research objects. Based on a satisfaction survey, three binary logit models are developed for each line to identify main factors of satisfactory degree. The results show that both train services and transfer journeys play important but different roles in three lines. The findings are helpful for improvements of railway service and promotions of passengers’ satisfaction.

Railway passenger comfort value is determined by a combination of physical and psychological factors. Comfortable passenger transport environment can be regarded as a resource for social consumption, and passenger travel comfort benefits can be regarded as a special generalized cost. The improved railway passenger comfort benefit (IRPCB) is a key evaluation index for newly-built railway projects, which can be regarded as the passenger comfort benefit difference before and after the newly-built railway project. However, the important value is neglected during the actual engineering practice, because the existing calculation approach is incomplete and the basic data is hard to obtain. In this paper, we propose a new IRPCB quantitative calculation method based on passenger candidate time value set. First, we establish the suitable railway passenger candidate time value set after analyzing and summarizing the railway passenger time value calculation methods and application conditions. Then, we multiply the passenger fatigue recovery time and the passenger time value selected from the candidate time value set. Next, the multiplication difference before and after the newly-built railway project will be calculated as the IRPCB value. A numerical experience is carried out by taking the Beijing-Shanghai passenger corridor in China as a background in order to test the performance of the approach presented in this paper.

Policy headway serves as a lower bound for scheduling service frequency and is usually used by transit operators in low-demand routes to ensure that passengers would be able to receive desired level of service. The main effect of policy headway on the level of service is reflected in passengers’ waiting time. Therefore, a method for determining the policy headway is proposed from the perspective of waiting time in this paper. Twenty bus routes categorized by location (urban or suburban) in Beijing are chosen as the research objects. Firstly, passengers’ perceived and acceptable waiting time in different seasons (winter or other seasons) and periods of a day (peak hours or off-peak hours) are collected, respectively, through questionnaire surveys. Secondly, by combining with the data of operating records, the relationship between perceived waiting time and actual headway (or time interval) is established. Then, the model for calculating the policy headway is developed. Finally, suitable values of the policy headway for different routes are estimated and compared. The results show that (1) for different route locations, the policy headway for urban routes is about 1.3 times larger than that for suburban routes; (2) for different seasons, passengers’ acceptable waiting time in winter is two minutes shorter than that in other seasons, and the policy headway in winter is 80% of that in other seasons; and (3) for different periods, the policy headway during peak hours is about 1.7 times larger than that during off-peak hours. The results also suggest the need of modifying the relationship between the average waiting time per passenger and the headway when the quality of bus service is evaluated from the perspective of passengers’ perception. The methods developed in this study can be used to determine the policy headway for a bus system to improve the level of service when considering passengers’ waiting time.

Based on cumulative prospect theory of behavioral economics, walking time, travel time, time sensitivity, and other factors are analyzed to quantify the effect factors of paths choice as building the generalized cost function. Then, according to the cumulative prospect theory, the paper discusses the fit area for pedestrians’ transfer to rail transit. Compared with the traditional way of determining experienced attraction value and travel intention data, the model is likely more effective and reasonable. It is aimed to provide references for the planning, set selection, and evaluation of urban rail transit stations.

Urban public transport and public life are closely related. Since routine bus system has the advantage over capacity and flexibility, its role is irreplaceable in urban public transportation system. Thus, the approaches that can take full advantage of routine bus system’s service ability is a key point to improve the level-of-service (LOS) of urban public transportation system. Based on previous research on flexible-route bus systems, this paper proposes a set of design principles, method, and processes which can be adapted to the majority of Chinese cities, especially medium and small-sized cities. This essay will focus on a flexible-route bus mode which can be adopted within Chinese cities and then propose a set of principles, method, and steps to transform existing bus routes in low traffic density areas into flexible bus routes.

Computational formulas for headway are proposed, and dynamic evolution of headway is further analyzed in this paper. Firstly, we divide the high-speed railway line into three areas according to the process of train operation. Secondly, different types of headway are computed respectively and the corresponding critical block sections are identified at the same time. Thirdly, a station with the typical layout is taken as an example. Considering length of block section, speed limitation, and train order mainly, dynamic evolution of headway and critical block section is obtained. Furthermore, sensitivity analysis is performed to identify vital factors of headway.

Intelligent shuttle bus scheduling is one of the main development trend of urban public transportation in the future. In the public transportation operation and management, we usually make shuttle buses’ scheduling plan of stations according to the traffic demand during the morning and evening peak periods. In this paper, firstly, we use the intelligent transportation system to predict the exact time of next bus to arrive at certain stations. Then, based on the interests of the passengers, the time cost of passengers and the bus operation cost are formulated as objective function and the scheduling model of the urban public transport shuttle bus is developed. Finally, a genetic-newton algorithm is used to solve the problem, and the model is validated by an example. The result provides a reference for the scheduling plan of urban public transportation shuttle bus.

With the increase of the residents’ commuting distance, the demand of the long-distance bus transit rises rapidly. The fluctuation of the passenger flow along the long-distance bus line inevitably leads to the waste of vehicle resource. The spatial analysis and quantitative measure of the long-distance bus are proposed on the demand. According to the route distance, the bus lines are classified and the passengers’ usage rate is calculated. Moreover, the bus service level is also evaluated by both the overall usage rate and the usage dispersion of the bus line. In this paper, 7 long-distance bus lines in Beijing with approximately 400,000 smart card data records are taken as the case study. This research can help to improve the efficiency of long-distance public transit and promote the allocation of the operation resource. It can be applied in other cities’ public transit utilization analysis based on the smart card system.

The study of rail transit assignment could not only help improve the operations of urban rail transits but also help study the fare clearing. So far, studies on rail transit assignment are mainly based on the random utility maximization (RUM) principle, which is not consistent with the actual route choices of passengers in some cases. To overcome these shortcomings, this paper analyzes passenger route choice behavior based on the random regret minimization (RRM) principle. The factors, such as travel time, transfer time, and in-vehicle congestion, are all taken into account. The RRM model is formulated based on multi-class passengers. It puts forward the transit assignment programs on the basis of LOGIT models. Through a numerical example, the result of the RRM model is compared with that of the RUM model. It shows that by avoiding situations where non-chosen alternatives perform better than the chosen one, travels aim to minimize anticipated regrets, rather than to maximize expected utility, which more realistically reflects the route choice behavior of passengers.

Metro disruptions due to unexpected events reduce the reliability of transit systems and result in significant productivity loss and long travel delays for passengers. Bus bridging strategy is often used to evacuate stranded passengers during metro disruptions. Most existing studies on bus bridging focus on designating buses to pre-designed routes, which restricts the flexibility of bridging plan and decreases the operational efficiency. This paper develops a plan-based flexible bus bridging strategy, under which a bus will serve an ordered list of disrupted stations instead of a fixed route. Two bus bridging models are formulated to optimize the bridging plans with the objective to minimize the bus bridging time and the total passenger delay, respectively. The proposed models are validated and evaluated through a case study. Results demonstrate that the proposed models are capable of minimizing the bus bridging time and the total passenger delay, as well as balancing bridging time over all disrupted stations.

The interfaces and external systems of operation control system (OCS) are investigated according to the communication modes and functions. The information flow diagrams of OCS are analyzed in terms of four perspectives, including information flow driven by operators, information flow driven by drivers, information flow driven by on-board controllers, and information flow driven by location components. Eventually, the functional architecture of OCS is derived from the information flow diagrams.

Based on the smart IC card data, the data regarding passengers’ loading stations are obtained. Besides, the passengers’ unloading stations can be induced by combining the bus travel chain and site attraction weight methods. After getting the passenger flow data, the maximum passenger flow station is determined by constructing the passenger flow curve of the stations. We further establish the expected passenger capacity model considering the bus operation cost and the passenger demand. The passenger flow at the maximum passenger flow stations is determined by the expected passenger capacity. Eventually, the bus timetable is established to solve the bus load problem.

This study investigates the phenomenon of bus stop failure due to the non-uniformity distribution of headways and proposes a method to adjust bus headways. In this study, a bus arriving time prediction method using generalized regression neural network (GRNN) was proposed to forecast bus operation data. According to the prediction network, the arriving time of buses at each stop could be predicted and further to determine whether to adjust the arriving time in advance. A case study was conducted to show the applicability of the proposed method with data collected from a bus route in Chengdu, China. The results showed that with the proposed method, the total travel time and stability of headway outperformed the current situation. Further research can be conducted to investigate the effects of different weather and locations on the performance of the proposed method.

This paper studied the application of virtual reality technology to the operation simulation of urban rail transit, aiming at enhancing the simulation effect and reducing the cost of development. Therefore, researchers can find the problems more accurately and quickly by concentrating on problem solving and operation optimization. A methodology for the simulation model construction was proposed based on multi-agent simulation technology and virtual reality system. Series of adaptive adjustments and improvements were made compared with the traditional method. A computer simulation model with a station as the scene was built. Experiment has been carried out to collect data for the simulation model validation. The results show that the methodology proposed in this paper is feasible and effective, which can provide support for the development of the virtual reality system of urban rail transit.

Modern trams have gradually been applied to the public transportation in China due to its features in humanity, efficiency, ecology, and environmental protection. In the initial stage of development, however, there are still many problems that occur in terms of function-orientation, route selection in initial implementation, construction standards, and project cost, because of the lack of guidance in the aspect of practical experience, technical standards, and criteria. Based on the analysis and thinking on modern trams that have been built or are being built in China, this paper provides discussion on key problems that have occurred in the development and application of trams and puts forward its development trend.

Public bicycle systems (PBS) are becoming popular in many cities for environmental friendliness, convenience, and other merits. Imbalance of bicycles between stations may hinder users to lend or return public bicycles and influence the service quality of a PBS. This paper analyses the temporal pattern of a PBS’s usage in Nanjing and then categorizes all stations into five clusters. A strategy to determine the optional stations is proposed based on the characteristics of different groups. Then, a model is established to obtain the optimal route for dispatching vehicles. Finally, a case study in Xuanwu District of Nanjing City is provided and presents that the strategy can narrow the scale of the redistribution scheme efficiently. The sensitivity analysis illustrates the moderate increase of initially dispatching bicycles in the depot can reduce the total operation cost.

Bus lanes have played an important role in improving the operation of urban public transportation. However, the operation of dedicated bus lanes is still far from satisfactory. Therefore, understanding the factors affecting bus lane operation can help to improve the efficiency of urban public transportation. Traditionally, problems affecting bus lane operations have been attributed to factors within the transportation system itself; however, other factors may affect the efficiency and operation as well. This paper used data related to the infrastructure within the transportation system as well as data from the urban built environment to evaluate bus lane operation in Shanghai, China. Key factors affecting operation were identified and then, the C5.0 decision tree algorithm was used to divide the bus lanes into 8 types according to the aforementioned indicators. Conclusions showed that the transportation infrastructure has a greater influence on bus lane operations than the surrounding urban built environment.

Existing research on the critical distance for pedestrian access to urban rail transit systems overlooks the effects of vital differences among urban rail transit stations. To overcome these challenges, this study included on-site questionnaires that were distributed at several urban rail transit stations in Nanjing. Analysis of the influential factors affecting pedestrian use of urban rail transit found that specific features of the rail transit station as well as walking distance have an effect on use. Due to these findings, distance was then used as an independent variable to develop a distance attenuation model. The critical distances for pedestrian access to urban rail transit were obtained and ArcGIS was used to analyze various station configurations; while many rail transit stations have a catchment area, or an area from which passengers originate from, that is, circular in nature, non-traditional catchment areas may also exist depending on lines serviced and rail schedule.

Automatic fare collection (AFC) systems record both time and location information when a passenger enters or leaves the urban subway system. Using this data, this paper analyzed the travel time components including access time, platform wait time, transfer time, egress time, and in-vehicle time, as well as their corresponding distribution functions. Using this data, a time-dependent graphical model was created in order to estimate passenger time-space paths within an urban subway network. Based on subway train timetables, the time-extended network model produces feasible time-space paths. A time-space path sheds light on traveler movements throughout activity network. The proposed methodology provides a way to analyze subway passenger behavior, which can serve as a reference for dynamic traffic flow assignment for urban subway networks.

The urban bus line has the characteristics of short interval of departure time and non-fixed arrival time. The rural bus lines have the characteristics of long interval of departure time and an accurate starting point. Research on holding control strategies for the rural bus lines at urban and rural public transit transfer stations can improve the reliability of the transferring between buses and reduce travel delays for passengers, especially for the passengers who transfer from the urban bus lines to the rural bus lines. This is an effective way to improve public transit service quality and use but also helps promote the equalization of urban and rural public transportation development. In this paper, urban and rural public transit transfer stations are proposed, and the holding control strategies for the rural bus lines are put forward. Then, by synthetically considering factors such as cost of enterprise, passengers, and capacity constraints, the model of holding control is established, which aims at the minimum total cost and to calculate the holding time threshold. A case study is given to analyze the effectiveness of holding control.

To achieve optimal transfer connections and shorten the waiting time of transfer passengers in urban rail transit, this paper proceeds from a perspective of network operation to analyze the target, characteristics, methods, and influence factors of network coordination. An optimal model of a train timetable coordination for transfer connections is proposed to minimize the average transfer waiting time by arranging the arrival and departure times of the trains at transfer stations. Particle swarm optimization (PSO) is used to solve the model, so the train timetable of the optimal transfer connections can be obtained as well. Then, the model and algorithm are used to coordinate the transfer connections in a three-line transfer topological structure by programming and simulating on MATLAB. The result indicates that this model and algorithm can effectively shorten the average transfer waiting time and optimize the transfer connections of the entire network.

Mass passenger flow in stations has become a frequent and serious problem for urban rail transit daily operation, especially at peak hours in many big cities of China. It is a main factor influencing the safety, reliability, and quality of the operations. According to the train running characteristics and ridership arrival regulations, this paper presents a mass passenger flow control model with the number of waiting passengers and dynamic capacity of the train as its subjects, aiming to minimize and balance the passengers’ waiting time in objected multi-stations of the whole line, by analyzing waiting delay mechanisms. The fmincon function provided by MATLAB is then designed and used to solve the model and find the optimal value. Finally, the analysis of a simulation case shows that the model and algorithm can be used to greatly optimize and balance the waiting time of each station for mass passenger control.

Skip-stop operation is a low-cost approach to improving urban rail transit operation speed and passengers’ travel experience in single track lines. This paper proposes a novel bi-objective model to optimize the skip-stop scheme for bi-directional urban rail lines so that passenger’s waiting time and in-vehicle travel time can be properly leveraged. Different from the conventional “A/B” scheme, the proposed flexible skip-stop scheme (FSSS) can provide flexible stops to accommodate spatially varied passenger demand and better utilize the line capacity. The train headway is constrained by three kinds of departure intervals: the minimum safety interval, minimum departure interval that meets the passenger demand, and the minimum turnaround interval at the terminal stations. A tabu search procedure is then developed to efficiently search for the optimal solution. A case study is conducted based on a real-world bi-directional urban rail line in Shenzhen, China, using the ridership information extracted from smartcard data. Compared to the all-stop strategy, it is found that the proposed skip-stop strategy yields a better performance. The results show that around 22.6%-58% of passengers benefit from the FSSS operation with shorter significant in-vehicle travel time, while the passenger waiting time only increases slightly. The operational time per train is reduced by 5%. Transit agencies may also benefit from this scheme due to the savings of energy and operational costs.

The metro system is a critical part of urban public transportation systems because of its high capacity and reliable operation status. Although metro systems have attracted a large number of passengers, information service and trip guidance is still extremely deficient. Lack of effective guidance information may lead to the loss of network capacity and is not conductive to improving the level of service for metro systems. Unfortunately, few studies have paid attention to trip guidance for metro users. In order to provide alternative pre-trip guidance for metro users, a method to generate dynamic path guidance is proposed. First, a valid path set was generated by conducting a K-shortest path searching algorithm to avoid path enumeration. Then, a passenger flow assignment strategy based on the metro schedule was designed to identify the real trip path of each metro user from the valid path set. Third, pre-prior sectional passenger flow was forecasted from historical travel information accumulated from individuals. Finally, alternative pre-trip suggestions were generated by combining trip time and trip feeling with variable weights. The proposed method was tested using the metro network in Nanjing city, China. The empirical results show that metro users without pre-trip guidance information may choose a path which is neither time-saving nor comfortable. Comparatively, the guidance information generated by the proposed algorithm can help metro users find the most suitable paths.

The passenger demand for urban rail transit is dynamic and different for working days, weekends, and holidays. Train timetables that adapt to different and dynamic passenger demand can improve the capacity and efficiency of urban rail transit and meet the needs of passengers. This paper first analyzes the characteristics of the passenger demand and then formulate a model with the objective to minimize the user cost and operating cost under constraints of train operations and the passenger boarding and alighting process. The proposed model is examined through a real-world case solved by an improved simulated annealing algorithm that adjusts train number, train running time, train dwell time, and departure and arrival times at each station. It then analyses characteristics of the train timetable under different passenger demand, such as train load factor in each section and remaining passenger number, which is useful to operational management in the subway system.

This paper aims at studying travelers’ public transport mode choice behavior on routes when faced with regular passenger congestion in transit corridors during the weekday peak hours. Initially, stated preference (SP) survey experiment under the assumption providing PTIS for passengers on certain routes was conducted and, then, discrete choice models were formulated by using cumulative prospective theory (CPT), multinomial logit (MNL), and mixed logit (ML), respectively, with 2970 observed samples, and accuracy comparisons between the three models was made. Also, by analyzing the level of service of information and public transit, this paper discusses the influence factors of traffic mode choice behaviors of on-route passengers. As comparison results revealed, CPT model is more accurate than MNL model and ML model. As sensitivity analysis results show, waiting time has a significant effect on public transit mode choice, and the threshold value of waiting time is 25.89 minutes according to model calculated results.

With urban development and increasing traffic volume, the expansion of urban rail transit has attracted more attention. Rail transit has entered a new era of networking operation in mega cities like Shanghai and Beijing. However, with the continuous increasing of network-scale and operating mileages, potential safety risks emerging from the operational procedures are also increasing. Thus, it is necessary to obtain the implications and impacts which make the rail transit network operation evolve from “passive type” to “active type.” This study examined 663 metro operation accidents that have occurred in the 14 lines of Shanghai for the 2011-2015 period. This data includes accident characteristics, date, location, delay time, and operation status. Temporal and spatial distributions of operational accidents were analyzed using statistical analysis and comparative analysis. The results revealed that vehicle failures and signal failures were the main types of accidents, together accounting for 60% of the total. Signal and power supply equipment were the significant factors influencing the delay time, namely the greatest impact on the users. Identification of the characteristics of metro operational accidents can provide valuable information for development of effective countermeasures to improve metro operational safety in China.

Large special events always generate additional traffic demands (i.e., event-related passengers), which usually stresses the urban rail transit with a much more severe test than normal conditions. In order to provide better operational measures and guarantee traffic security during the event, it is important to identify the impact and evaluate the influence of the event-related passengers (ERPs). Considering the China import and export fair (Canton Fair) as an example, this paper proposes a data mining method to identify the influenced stations and estimate the event-related passenger flow using data gathered by the automated fare collection (AFC) system of Guangzhou Metro. Meanwhile, based on the analysis of land-use and other attributes of urban rail transit stations, a multinomial logit (MNL) based destination choice model is introduced to account for the ERPs’ travel behaviors. The result shows that station attributes (e.g., attraction of the stations, land use type, land use intensity) and some level of service variables (such as travel time, transfer time, and station betweenness calculated by network analysis) play an important role in the destination choice behavior of the ERPs. Thus, the model is capable of accurately describing the ERPs’ choice behavior during large special event. The study is helpful to accurately estimate the passengers’ destination choice behavior and make reasonable operational plans during large special events for urban rail transit operator.

Considering the congestion in the metro, the operator usually uses congestion degree information to guide passenger flow which can alleviate the operation pressure. However, the current congestion degree classifications are mostly based on subjective judgment, without fully considering passengers’ psychological perceptions under different load factors. This paper aims to set up a reasonable classification of congestion degree for urban rail transit system. First, passengers’ perception differences of congestion degree are analyzed from the views of different properties (i.e., age, gender, travel purpose, and travel time), based on the questionnaire data collected in Beijing Metro. Then, the distributions of passengers’ congestion degree perception under different load factors are calculated, and the load factors are clustered based on hierarchical cluster analysis. Finally, the congestion degree for urban rail transit is classified into four levels (i.e., comfortable, slight crowded, crowded, and heavy crowded), corresponding to the load factor interval (0, 40%), (40%, 75%), (75%, 100%), and (100%, +∞).

The preparation of shunting plan of the station is an essential element of rail transport organization. This issue establishes the dynamic programming model of optimal shunting plan in intermediate station according to its requirements and characteristics. Through the analysis of cars’ positions and their moving costs, a cost function about most times of moves of cars is established. Meanwhile, an approximation algorithm of potential search is designed. By the combination of the dynamic programming algorithm and potential algorithm, an implicit enumeration algorithm of intermediate station is designed, which has been applied very well in practice.

Changsha Maglev Express Line, connecting Changsha South Railway Station with Changsha Huanghua International Airport, started trial operations in May 2016. We conducted an investigation on passengers departing from Changsha Huanghua Airport in 2015, while the maglev line was under construction. In the study, information of transport modes choices on how to access the airport and the time costs without maglev was collected. We also investigated passengers about their willingness to take maglev in the future. Based on the survey results, we estimated the possibility of passengers’ transferring to maglev for the landside transportation. Changsha Maglev has started its commercial operation since May 2016. By analyzing its operation and service condition, we evaluated the impact of Maglev Express Line on the land-side transportation of Changsha Huanghua International Airport. The outcome of the study can provide valuable reference to other urban maglev projects planned in the future.

This paper takes the valley city of Lanzhou as an example to study the evaluation method of urban rail transit network planning for valley cities. We start by defining a valley city and then propose a comprehensive evaluation method of line network based on set pair analysis. By constructing an identical degree decision matrix, we employ analytic hierarchy process to determine weights and use comprehensive evaluation method to get the optimal scheme. Finally, a case study on rail transit planning in Lanzhou demonstrates the proposed evaluation method. The numerical example indicates the effectiveness; thus, a decision basis is provided for urban rail transit network planning.

Mobile communication solutions for high speed railway scenarios aim at providing seamless coverage, high system capacity, and a high data rate for end users, which in turn face Doppler frequency shift, penetration loss, mobility management, and system capacity planning. Increasing attention has been focused on time-division duplexing-based technology. This paper introduces time-division synchronous code-division multiple access (TD-SCDMA) and time-division long-term evolution solutions (TD-LTE). We also present experimental simulation results of TD-SCDMA and TD-LTE solutions in a maglev high-speed network environment. To improve performance, we discuss an enhanced solution based on mobile relays, a comprehensive solution for high-speed railway scenarios.

This paper proposes a price sensitivity analysis method for high-speed rail passengers based on the ticket purchasing process. This analysis method defines the time when second-class seats are sold out as time of demarcation, divides passenger’s ticketing into free state and non-free state, and passengers are divided into deterministic and non-deterministic. Passengers’ sensitivity to different prices is analyzed after counting their ticketing behaviors before and after the demarcation moment. Based on analysis of passenger ticket data at different distance intervals, the conclusion is that passenger price sensitivity increases as the distance increases, and the price is adjusted to correspond.

The accurate prediction for the travel time can improve bus operation efficiency. The improvement of bus service level and enhancement of bus trip can relieve the urban traffic problems. To predict bus travel time between adjacent signalized intersections, BP neural network model was used. Factors which influence bus travel time were considered as the input of the network model, and bus travel time was used as the output. Bus route No. 3 in Nanjing was chosen as a case study. The results verify model’s feasibility and indicates that the presented model has certain practical values.

Modern tram, a green and sustainable public transportation, has been widely used in the world due to its capacity, comfort, convenience, and environmental protection. This paper presented the priority organization mode of modern tram. The core of the scheme was to use the Webster method and a mathematical model to optimize the signal timing scheme. Also, he active priority strategy was adopted for modern tram, namely the early green and green extension. In the last part of the paper Dalian Modern Tram No. 201 was studies and data was collected from this line and applied modern tramway priority organization scheme. The VisVAP in VISSIM was used to show the reasonableness and superiority of the priority organization scheme of modern tram.

The adjustment of train diagram is becoming more and more frequent with the change of passenger flow under the condition of network-operation of urban rail transit (URT). The manual crew pairing method is difficult to meet the needs of operation currently. In response to this situation, an optimization model for urban rail transit crew pairing problem based on workload equilibrium is established in this paper. The model is based on datum of train diagram. Then, an improved heuristic algorithm is designed to solve this model. Lagrangian relaxation technique and column generation algorithm are adopted to improve the efficiency of the algorithm. Finally, a numerical example is given using data from Line 4 of Chengdu Metro. Results show that the new algorithm can get the optimal crew pairing plan in a short time and can satisfy the workload equilibrium among duties to a certain extent.

Crew scheduling is an important task of urban rail transit (URT) crew management. The crew scheduling is not only related to the reasonable allocation of the crew’s workload but also directly affects the efficiency of urban rail transit operation organization and operating costs. The process of crew scheduling involves many factors which are interrelated and play decisive roles in determining the preparation direction and methods of the crew scheduling. Based on the process of crew planning, this paper firstly analyzed the types and the features of each type and the calculation method of work shift. Secondly, this paper proposes the method to ensure the number and location of crew rotation station. Lastly, it analyzed the way to set the locations of dining place and the setting down station and makes a case analysis. This research guides well the urban rail transit crew scheduling and improves the crew management efficiency.

As many passengers fail to board on trains in an overcrowded urban rail transit (URT) system during peak periods, a novel phenomenon appears that passengers are willing to spend more time and energy traveling backward to secure a seat or just a place to stand. “one-ticket-transfer” smartcards in China can only record spatiotemporal information of origin and destination (OD) without path, therefore, it is hard to identify whether the fail-to-board (FtB) passengers are traveling backward or just waiting at the platform. To identify the phenomenon of “traveling backward” and its share of passenger flow, we use affinity the propagation (AP) method under the between-within-proportion (BWP) index to cluster automatic fare collection (AFC) data with single-path OD; and carry on a field research to verify the proposed method. We also establish a virtual path to rationally assign these “traveling backward” passengers to the network, which will reflect more actual passenger flows distribution and give advice for overcrowded URT network management.

In terms of emergency solutions to sudden occasions occurred in the railway transportation, the multi-intelligent model is introduced to model the passengers, emergency vehicles, and railway terminals so as to make analysis and predictions on the passenger volumes in the railway. For this reason, the passenger volume matching is proposed in terms of the emergency vehicle organization. The paper conducts the visual simulation toward the emergency under ArcGIS software and makes the evaluations toward the effect metrics, including the mean delay times, the mean occupancy rates of emergency vehicles, and the total volumes of passengers. Finally, the paper provides corresponding technology and solutions to optimize the organization of the emergency vehicles.

This paper presents a simulation method to calculate the turn-back capacity at terminal stations based on the analysis of factors which influence turn-back capacity on urban rail transit stations. First, a traction calculation mode of metro trains is designed to obtain the velocity curve in turn-back operations. Then, a network topology structure method is built according to the actual interlocking block at turn-back stations, aiming to control route opening, and reversing train protection in railway sections. These two steps are therefore used to simulate the actual turn-back process and figure out train turn-back capacity. Finally, a numerical example is given to illustrate the application of the simulation algorithm, its error is about 0.98% compared with the actual traditional method. It is expected to provide decision support to operational departments of urban mass transit in calculating turn-back capacity.

The rapid development of urban rail transit system appeals to more and more people choosing metro as the main trip mode. Meanwhile, the increasing passenger volume brings great challenges and much pressure on metro operation, thus breeding a lot of passenger complaints. In dealing with passenger complaints, now the metro agencies use the artificial method to classify complaint data and make a simple summary which is inefficient and to be enhanced. Based on that, this paper aims to establish a complaint data analyzing system to address such concerns, and the system structure is designed. Furthermore, several analysis methods are integrated to deal with the comprehensive complaint data, including text mining, geostatistical analysis, classification algorithm, and data visualization. With multidimensional analysis, this paper focuses on the hot points and pain points in passenger complaints, ultimately provides decision support for operation managers.

First, the hierarchy and external interfaces of operation control system (OCS) are summarized. Then, information flow analysis method is introduced. Finally, the information flows of forced stop management of OCS are analyzed according to three perspectives, including forced stop management driven by external interface failures, forced stop management driven by internal failures, and forced stop management driven by staffs.

Telephone block method is the common downgrading operation mode under ATC failures. In order to reduce the effect of abnormal conditions on train organizing, this paper seeks to improve the telephone block train organizing method by the application of collision avoidance devices and verify our accomplishment with practical data. According to a case study, the efficiency of telephone block method can be improved by 13.33%.

This paper presents a non-linear bus scheduling optimization model in order to achieve coordinated scheduling for rail transit and bus transit at transfer stations. Based on the diversity of transfer time and the vehicle capacity, the author considers to establish a model, which can minimize the total waiting time of passengers. Then an orderly inter-code genetic algorithm is designed to solve the optimization model and minimize the waiting time of passengers. At last, the proposed model is applied to a metro station in the city of Xi’an, China. The results are obtained by using improved genetic algorithm for solving the non-uniform grid schedule, which show that the total waiting time is reduced by 28.29%. Consequently, it indicates that the optimization model is validated and scientific. It can help to decrease the waiting time and improve the transfer efficiency.

In the process of urban rail transit operations and organizations, passenger inflow control is the key method to solve contradictions between rail network capabilities and passenger demands. In this paper, we propose a mathematical programming model to calculate limited passenger inbound volume in each separate period of the peak hours. We apply our model to a typical metro station in China and formulate a quantified passenger inflow control strategy. The results show that the model proposed in the paper is effective and this quantified approach could optimize present passenger inflow control measures.

The paper designs an integrated positioning architecture by coded odometer and Doppler radar and analyzes their error sources. Confidence interval of train safety positioning is identified based on the analysis. Kalman filter is used to diagnose abnormal output from coded odometer and Doppler radar. Safety comparator is used to identify the reliability of confidence interval. The experimental results show that the proposed method can effectively keep the efficiency and safety of train positioning when the train is in slipping state.

Driven by vigorous development programs of urban green transportation, as a kind of green, low carbon, health, energy saving mode, public bicycle system has been rapidly developed in many cities across the country. Therefore, evaluating and analyzing the satisfaction of urban public bicycle becomes significantly important in practice. After establishing the satisfaction evaluation index system of public bicycle system, the analytic hierarchy process (AHP) is established in the study using the Delphi method. Based on the present status of public bicycle satisfaction in Xi’an, the current overall level of satisfaction of public bicycle in Xi’an is scored 82.9 points, a level between “general satisfaction” and “more satisfied.” Through the fuzzy comprehensive analysis of the evaluation results, we found out the weak links in the public bicycle service system and proposed the corresponding improvement suggestions.

Traditional sand cone method is time-consuming and laborious for the detection of compaction degree. However, PFWD has the advantages of rapid, non-destructive testing, which has been widely concerned by the railway and highway departments. The relationship between compaction degree and deformation modulus E_vd was established to put forward the corresponding E_vd values of different compaction degrees. A typical road section was chosen for testing to evaluate the compaction quality. The contour model of the deformation modulus was established on semi-rigid base by Matlab software, and the area ratio was introduced to evaluate the compaction qualification rate. Results indicate that the correlation between compaction degree and deformation modulus E_vd is good. PFWD can quickly and effectively evaluate the compaction state of semi-rigid base.

The two-step left-turn control model of ring-intersection (RI) has been widely applied to many cities in China. It is still unclear which control model is more suitable when the RI must be reformed due to being unsatisfied with the rapid development of traffic demand. This paper presents a method based on the characteristic of traffic flow in RI considering the efficient performance of the vehicles. Next, a more accurate capacity calculation formula of the left-turn lane is proposed using the stop line method. Then, this method is effectively tested in a typical RI, in the Hunan Province utilizing VISSIM platform.

This paper presents an evaluation framework for quantifying interrupted flow traffic states. Based on the essential “uninterrupted component” (UC), and “interrupted component” (IC), characteristics of the travel time bimodal distribution is at each 15-minute interval. The real-time delay vector consists of uninterrupted delays and interrupted delays and is drawn as a point in the delay coordinate system in which the origin of coordinate is a pair of expected travel times in the free-flow bimodal distribution. Then, a novel evaluation indicator, named as the traffic state value (TSV), was calculated as a vector inner product of a delay vector and the vehicle number of UC and IC. The numerical analyses indicate that TSV can quantitatively evaluate the traffic state, and the tendencies of TSV demonstrate the traffic state fluctuations. TSV curves on Thursday show significant bimodal features during morning and afternoon peak hours, respectively, while unique peak is demonstrated on Saturday afternoon due to commuting journeys.

This study investigated the congestion formation mechanism at intersection groups and used the system dynamics model to initially propose to account for the cause-and-effect relationships among the influencing factors. In order to express the oversaturated conditions at intersection groups, the average saturation degree was introduced. According to the congestion formation mechanism, the system dynamics model was built up and simulated to account for the cause-and-effect relationship among the influencing factors. Two variables, degree of order and roadway adhesion coefficient, were selected to perform the sensitivity analysis to which the degree of order was found to be sensitive to the average saturation degree, while the roadway adhesion coefficient was not.

This paper proposes the method of successive averages (MSA), and the method of successive weighted averages (MSWA), for determining the stochastic user equilibrium solution (SUE) for logit-based loadings. The deterministic user equilibrium (DUE) model which assumes that users get perfect information did not work well in practice. Thus, the stochastic approaches to traffic assignment was proposed and used. This paper addresses the conventional MSA and MSWA to solve this iterative problem. Finally, a numerical example is used to assess the proposed methodology. Results reveal that the convergence property of proposed MSWA is much better than that of conventional MSA.

This paper presents a schedule-driven signal priority control for modern trams using reinforcement learning. This self-learning adaptive system has the following three features: 1) In addition to getting the minimal delay of general vehicles, the schedule-driven concept was adopted by using reliability of the tram’s schedule as the primary control goal. 2) Model-free reinforcement learning was used to find the optimal signal timing strategy instead of traditional model-based methods. 3) A dynamic phasing sequence was used to extend the flexibility of signal priority instead of fixed sequence in cyclic timing. Compared with three other traffic signal control strategies, extensive VISSIM simulation experiments under different traffic conditions have demonstrated the excellent punctuality and delay reduction of schedule-driven TSP for modern trams.

This paper conducted research on multi-objective optimization models of intersection signal controls considering traffic delay, stops of vehicles, and vehicle emissions. Each optimal objective was processed by dimensionless normalization method, and a solving method based on weighted optimization and genetic algorithm was designed for the model. Based on this method, a case study was conducted for the application analysis of the proposed model. The biggest difference from other studies is that this model separates buses from the heavy duty vehicles when calculating emissions of different vehicle types. In addition, the optimization results of the proposed model was compared with other conventional models, which shows that the multi-objective optimization model proposed in this paper had the preponderance over the other models. In other words, the proposed model can effectively reduce intersection vehicle emissions, traffic delays, and improve the efficiency of intersection traffic.

Road construction and congestion charging are popular methods for alleviating traffic congestion. New road construction aims at meeting with people’s travel demands. Congestion charging is a necessary way to configure reasonably basic infrastructure. In order to test the usefulness of congestion charging on new roads, four compared experiments are built on the basis of different cases of road construction and congestion charging. Based on stated preference data collection and four compared experiments simulated with VISSIM software, the congestion condition and social benefits confirm that new roads in the congestion area could not solve traffic congestion problem of incorporating induced traffic. Results show that congestion charging incorporated into new road construction is the efficient way to alleviate road congestion incorporating induced traffic and satisfy people’s travel demands. Therefore, congestion charging on new roads should be a good solution from both a demand and supply view to solve traffic congestion problems.

Different forms of exit ramp cross-section of standard driving behavior impact on traffic operating characteristics are different; it will eventually affect the level of safety and capacity ramp. Through field research overpass typical cross section in the form of the exit ramp, the use of technology to build Uc-winRoad exit ramp to build three-dimensional road scene, the driver behavior data collected by the virtual drive to give the vehicle longitudinal and transverse traveling speed, acceleration, accelerator pedal, brake pedal curve select and change lanes, choose the original vehicle was a single lane exit ramp traffic flow double random distribution law for the development of safe driving presented ramp carts drive on the right, small cars can be disposed of with countermeasure from the ramp on the left for traffic safety research proposes a new way of thinking.

The primary objective of this study is to quantitatively evaluate the impacts of driver population heterogeneity on capacity and traffic performance at basic roadway segments. Drivers were classified as two categories including novice drivers and experienced drivers, and they mainly differs in driving experience. On-road experiments were conducted for measuring the car-following data associated with the test drivers. Based on a nonlinear optimization procedure, the improved intelligent driver model was calibrated by minimizing the deviation between the observed driving trajectories and the simulated dynamics. Capacity and traffic performance were analyzed with respect to different driver population compositions by using the calibrated simulation model. The simulation results revealed that the mixture rate of novice drivers can result in capacity reduction and poor traffic performance. Specifically, the reduction rate of capacity may reached up to 28%, and most of other key measures representing traffic performance characteristics also significantly deteriorated.

This paper focuses on the operation changes of taxi systems with the wide-usage of e-hailing apps using the calculated taxi OD pairs data of Shanghai in 2012 and 2015 and applied an evaluation process to analyze the possible impacts from three dimensions, namely, taxi system supply, operation efficiency, and service coverage. The results reveal that the e-hailing apps have partly improved the taxi operation. On the one hand, the supply of later taxi systems has experienced an obvious drop; however, on the other hand, the existing taxi system has overall higher efficiency owing to the decreased percentage of vacant trips. The comparison of taxi trips demonstrates that the taxi system can better satisfy larger-scale and longer-time service demands than before, especially in some low-density areas and during the non-peak hours. The findings provide another perspective for decision-makers to comprehensively understand and management the taxi system or urban transport after e-hailing service.

In order to study bus priority control based on downstream bus stops at intersections, bus priority application is established which considers whether bus queue at downstream stop is “overflow” and predicted arrival time is “early.” Improving on Webster’s method, the green time ratio is determined by passenger proportion and bus signal priority model is proposed with the objective of minimizing average passenger delay. Using an intersection in Xi’an, China, and its downstream bus stop as an example, signal cycle optimization and two priority control strategies for green advance and extension are compared and analyzed. The results show that the proposed signal priority model and the two strategies can reduce average passenger delay and average vehicle delay of target intersection, as well as shortening the deviation from bus schedule.

In order to strengthen the deterrent effect of unitary penalty of traffic violation in China, the block traffic penalty is proposed. By introducing the fines elasticity, the drivers’ response in the amount of times of traffic violation and total fines is analyzed after implementation of the block traffic penalty. Then, the multistep fines model is established based on the theory of supply and demand and marginal effect, and the pattern search method is used to solve the model. A case study was conducted with the illegal parking data in Mudanjiang. The study results show that block traffic penalty has very strong inhibition effect with high frequency traffic violations; for example, the total illegal parking times of the third- and fourth step drivers are reduced by 9.58% and 21.52%, respectively. By adjusting the coverage rate and increasing the fines difference each step, the total traffic violations decrease more obviously.

The primary objective of this paper is to analyze freeway operation efficiency and safety of integrated control strategy. The simulation parameters are calibrated and validated using collected data on southbound I-880 freeway in California, the United States. The integrated control is estimated using cascade feedback of mainstream traffic flow control (MTFC-VSL) via variable speed limit. And the application of ramp metering (RM) is connected using specifically developed split-range-like scheme. The integrated control strategy is implemented through VISSIM. The results show that the integrated control had a 41.5% improvement in total travel time (TTT) over the no-control case. Moreover, about 59.8% of total conflicts are reduced by the integrated control strategy than the no-control case. Thus, it is beneficial to well understanding the mechanism of integrated control strategy and using it in practical implementation.

Taking the chevron sign alignment in freeway ramp as the object, we conducted comparative tests in a driving simulation environment on the spacing of chevron alignment sign. Simulation scenarios of three different spacings (10 m, 15 m, 20 m) of freeway chevron alignment signs was established by modeling software Multigen Creator, etc. 60 drivers were invited to the driving simulation test, and then their vehicles’ dynamic data from the experiment was analyzed by the researchers. The research team studied the operational safety of different spacing of linear guide signs regarding the vehicles’ track offset.

Shared lane design has been used to relieve the hindrance of left-turn movement. The primary object of the study is to identify the benefit zone where shared lane design overwhelms conventional protected control strategies at signalized intersections. Analytical models were proposed to calculate the minimum green time duration required to fully discharge all the approaching vehicles, which is used as an indicator of the effectiveness of different strategies. Based on the proposed model, the benefit zone of the shared lane design was identified. With the benefit zone plot, engineers can decide whether to deploy shared lane design according to the traffic demand. A realistic method was also provided to increase the accuracy of the proposed models. The results showed that the shared lane design is suitable to intersections where traffic demand from opposing approaches is significantly unbalanced. The proposed framework is expected to be helpful for traffic engineers to improve the performance of signalized intersections effectively and efficiently without complex computing.

Bike-sharing system, as an emerging public transport system, can allow users to rent a bicycle at a rental station for a short trip and return them at another station. To ensure a good quality of service to users, a reasonable and efficient scheduling is essential. In this paper, based on the multi-objective of minimizing the routing length and maxing satisfaction of users, we studied the scheduling route optimization problem based on bike-sharing demand characteristics at different times of a day. As a result, bi-level models at off-peak and peak are developed. By comparing the models considering weight of stations, customer satisfaction is improved, which has been little reported in literatures. With models successfully solved by LINGO, this paper verifies the efficiency and feasibility of the model by numerical examples. The study’s results show beneficial effects of scheduling route optimization on solving problem in rental and return bikes in bike-sharing system.

Ramps are vital pieces of infrastructure connecting city traffic networks to freeways. The performance of a ramp is to some extent determined by the on-ramp lane arrangement. In this research, our primary aim is to evaluate the performance in terms of vehicle emissions for two on-ramp lane arrangements: added lane and zip merging. We estimate the CO₂ emissions on the basis of the speed, and acceleration of vehicles in accordance with the comprehensive modal emission model (CMEM), and then analyze the impacts of traffic volume and heavy goods vehicles (HGVs) on emissions. The effect of main road traffic flow on emissions for the two on-ramp lane arrangements are analyzed under scenarios with traffic volumes of 800, 1000, 1200, 1400, 1600, and 1800 vehs/hr/lane. Meanwhile, the relationships between CO₂ discharge and various proportions of HGVs (2%, 4%, 6%, 8%, and 10%) for both on-ramp lane arrangements are evaluated as well. We eventually present emission contour charts for the two on-ramp lane arrangements based on the possible combinations of traffic volumes and HGV percentages.

This paper classified the major factors of traffic block incidents on national highway, which is based on the recent 3 years of national highway block incident data. In addition, the major causes of highway block incidents in different areas and time periods were compared and analyzed. Finally, according to the spatial and time distribution of highway block incidents, some operation management strategies were proposed for highway network management agencies.

In this study, travel time reliability was incorporated into the route choice decision-making process based on the prospect theory. A mathematical model was proposed to reveal the impact of travel time reliability on the route choice behavior. Under the condition of given travel time reliability, we can calculate the value of reference point. The proposed methods involves analytical model and stochastic user equilibrium model based on the prospect theory. Finally, by studying the different reliability conditions, the change in the value of prospect and the flow on routes were found, and a reliability sensitivity analysis was conducted. Using a numerical example, we present the whole analytical process.

Microscopic traffic simulation is an important experimental tool for traffic system management and control. For most of the traffic simulation software, model parameters are determined based on the foreign situations and the result is actually inapplicable to China. Firstly, we developed a calibration method for VISSIM simulation model using the golden-ratio based genetic algorithm. Then, a VISSIM simulation example of Lianhua-Xinzhou signal intersection in Shenzhen was built to calibrate the model parameters and test the proposed validation method. The results indicate the effectiveness of the proposed algorithm by comparing the on-site investigations and simulation results.

The continuous flow intersection (CFI) mode is one of most important unconventional intersection design (UID) modes. This kind of intersection transfers left-turn movements to secondary intersections to obtain a 2-phase signal control scheme at the main intersection, so the traffic efficiency is greatly improved. Existing research on the optimal timing model for the CFI merely takes motorized traffic into account. This paper establishes a theoretical study on optimal timing of CFI with non-motorized traffic. We first introduce a pedestrian phase design to the signal timing scheme of the CFI and extend the delay and queue model to accommodate non-motorized traffic. The optimization model, constrained by queue length and pedestrian phase, is then built to minimize average per car delay of the CFI and guarantee pedestrians crossing the intersection. Finally, the case study proves that the optimization model can produce an effective timing scheme for the CFI with non-motorized traffic.

An essential factor that affects the travelers’ route choice behavior is travel time variability, as indicated by empirical studies. Travel time variability elaborates both reliability and unreliability aspects. The concept of mean-excess travel time can capture this characteristic of travel time variability, which is one of the most widely used measures for assessing transportation network performance. Moreover, the travelers’ route choice decisions are based on the perceived travel time rather than the actual travel time. Therefore, this paper firstly presents a traffic equilibrium model based on the mean-excess travel time with stochastic perception errors that explicitly considers road network capacity in the route choice process. Thus, an acceptable capacity exceedance probability will be considered in a degradable road network. The network traffic capacity model based on mean-excess travel time with perception error is presented. Finally, numerical studies on a small network are provided to illustrate performances of the network.

Recent studies and empirical data showed that a macroscopic fundamental diagram (MFD) can reflect the relationship between the network vehicle density and the network efficiency, as well as the relationship between vehicle numbers in the network and vehicle numbers leaving the network. This paper proposes a traffic flow optimization model among traffic zones using macroscopic fundamental diagrams. Firstly, the coarser graph is used to represent the road network. Secondly, the traffic flow optimization model is proposed. The objective of the model is to maximize the total number of vehicles leaving the network at a certain period, where decision variables are traffic flow among traffic zones. Finally, micro-simulation is used to test and verify the validity of the proposed model. The results show that when the network congestion occurs, the proposed model can alleviate traffic congestion in the congested area and improve the overall operational efficiency of the network, which has important application value for preventing and alleviating traffic congestion of urban road networks.

This paper points out the necessity for active detection of commercial vehicle fatigue driving violations. And then, it puts forward a method to use technology to actively discriminate instances of commercial vehicle fatigue driving violation. The method is based on highway passing-through information obtained through highway gate system devices. Through analyzing both the driving time and the break time of the commercial vehicle, this approach can identify whether the driver is under a fatigue driving condition or not. By this method, the traffic management authority can gain an advantage to obtain a commercial vehicle’s driving track and to regulate the commercial vehicles more effectively.

In order to improve upon the theory of mixed non-motorized traffic flow, a statistical distribution of vehicle speed and traffic flow models were used for the non-motorized vehicle lanes that had either no segregation, visual segregation, or physical segregation. The first step was using video observation methods to analyze the trend changes of vehicle speed distributions for each of the three kinds of non-motorized vehicle lanes. Secondly, a help file on the traffic flow models of motorized vehicle lanes was referenced to obtain traffic flow models for the three kinds of non-motorized vehicle lanes. The results showed that there was no relation between the speed, density, or flow of non-motorized vehicle lanes under low density and flow conditions, and a negative correlation existed between the speed, density, and flow of the non-motorized vehicle lane when density and flow were high.