CICTP 2022: Intelligent, Green, and Connected Transportation

Currently, prediction of electric bus energy consumption (EBEC) has become an important prerequisite for the deployment of large-scale electric bus fleets. Physics-based models have limitations concerning the external factors; data-driven methods have attracted more attention for EBEC estimation with machine learning algorithms. This study establishes four classic machine learning models, with backpropagation neural network (BP–NN), support vector regression (SVR), random forest (RF), and extreme gradient boosting (XGBoost), to compare their prediction performance of EBEC with comprehensive consideration of influencing factors as inputs. The impacts are investigated via a variable importance analysis. The built models are performed on data collected from 1024 EBs in Shenzhen, China, from January to May 2019. Results show that RF has the best performance, with accuracy in RMSE raised by 7.97%–21.6% compared with other models. According to the variable importance analysis, the trip length is the most influential, while the dry-wet condition is least important.

With the development of intelligent driving and communication technology, tasks in the vehicular crowdsensing (VCS) system develop in a large-scale and complex direction, requiring higher data dimensions and more data types to be collected. Due to the diversity of sensing capabilities of vehicle nodes and the uneven spatial and temporal distribution in the road network, it is difficult for a single vehicle to complete such complex data collection tasks individually. This paper proposes a task allocation method based on cooperative sensing of vehicle nodes for complex multidimensional tasks in the VCS system. The sensing ability of vehicle nodes and the spatio-temporal correlation between nodes and tasks are considered to allocate single-dimensional subtasks after dimensionality reduction of complex tasks. Simulation experiments are conducted to analyze and evaluate the algorithm in this paper. Results show that our algorithm can reduce node energy consumption and data collection cost while ensuring task completion rate.

Along with the trend towards an autonomous transportation system (ATS), the intelligence of personal mobility service (PMS) can be further lifted by sensing travelers’ statuses comprehensively, learning behavior patterns accurately, providing travel options appropriately, and giving service responses timely. Such a process relies on a seamless information flow, which shall address data silos caused by laws and regulations about privacy. This paper proposes a federated architecture for PMS, called FPMS, which adopts federated learning, to provide personalized multi-modal options by aggregating personal data in a privacy-preserving way, and utilizing idle resources of personal devices within the service cluster. In general, by analyzing the physical objects involved, functions required, and data processed, a reference architecture of FPMS is designed to guide its construction in ATS effectively and efficiently. Moreover, a performance evaluation between FPMS and conventional centralized PMS is also presented to reveal the advantages of FPMS in saving service costs.

The development of a new generation of autonomous transportation system (ATS) requires a complete system architecture to guide, and user needs’ elicitation is the basis for constructing such a system architecture. The “travel activity” was first identified, and several travel modes, and transfer for passengers and by truck for goods, were considered to subdivide the travel activities. For each travel activity, the activity theory is applied. The subject, community, object, tools, rules, and division of labor of travel activities were specified. Activities were then divided based on the travel process and actions, and user needs were elicited from the actions from the perspectives of safety, convenience, efficiency, green, and economy. Finally, the framework was applied to elicit user needs in the context of autonomous driving, and the contradictory concept of activity theory was used to explain the evolution of user needs under different autonomous levels.

This paper uses video from UAV to optimize the Connected-Automated Vehicles (CAVs) operation system, to ensure the merging and yielding behavior safety and efficiency at freeway on-ramp section in an autonomous environment. The vehicle trajectories are extracted by Yolo and Tracker. An algorithm is supposed to identify the key point of merging behavior. Next, this paper applies the framework of game theory to the autonomous environment, uses bi-level programming to minimize deviation and find the pure strategy Nash equilibrium solution. Using TTC to test the safety of the model, the results indicate that this framework can effectively choose the best action of freeway on-ramp merging and yielding situations while achieving more effective merging operation.

Emerging techniques like connected and automated vehicles (CAVs) and vehicle-to-everything (V2X) communication may alleviate bottlenecks on highway on-ramps. This paper proposes a method using CAVs to solve the merging control problem. An improved state transition algorithm is proposed to determine the optimal passing sequence in different merging scenarios. A repair transformation is implemented to make discrete heuristic optimization algorithms suitable for this kind of problem. An algorithm memory procedure is proposed to avoid repeated calculation and improve operation efficiency. Simulation studies were conducted using trajectory data exported from SUMO. The study showed reduced time consumption by 22% on average, while eliminating some unstable situations. Compared with the First-Come-First-Serve strategy, the proposed method can decrease time consumption from 5.16% to 30.05% while traffic flow varies from 200 veh/(lane∙h) to 1,600 veh/(lane∙h). The study showed that the proposed method is scientific and reasonable even under the high level of traffic demand.

The control problem for large-scale traffic networks is computationally complex. In order to obtain satisfactory control performance with tractable computational efforts, the controllers should operate in a well-designed structure. In this paper, we propose a multi-layer hierarchical control structure for signal optimization, which consists of a cycle optimization layer, a coordination layer and a local control layer. Under the control structure, the Webster’s method is applied to update cycle lengths while the augmented Lagrangian method is introduced to deal with the signal split optimization in a distributed way. To apply the architecture control structure in real time, the model-based predictive control approach is adopted so as to obtain the optimal solutions for all layers. Experimental results on the Nguyen-Dupuis network are presented, which illustrate the performance of the proposed control approach.

Investigating effective freeway service area inbound vehicles will help manage the local traffic flow and improve the layout and service capacity of the freeway service area facilities. For the freeway service area with no detecting devices, two new methods are proposed by using vehicles’ RFID (Radio Frequency Identification) data from freeway upstream and downstream roadside collecting units. The GMM-EM based method is proposed to obtain the probability distribution of the speed of passing vehicles. Using the Ant Colony Algorithm, the speed threshold for distinguishing vehicles entering or not entering the service area is gained. Then the variational inference Gaussian mixture clustering algorithm is established to determine whether the vehicle drives into the freeway service area. Using RFID data of vehicles from the Chongqing Freeway in China, the experimental results show that the average accuracy of distinguishing freeway service area inbound vehicles is more than 96%.

Uncoordinated merging of different directions vehicles at the ramp is one of the main reasons resulting in the traffic oscillations phenomenon, which directly influences travel time and fuel consumption of vehicles. Emerging connected autonomous vehicles (CAVs) and Vehicle to Everything (V2X) communication technologies provide the possibility for detailed vehicle motion control to reduce negative influences of traffic oscillations imposed by uncoordinated merging at the ramp. This paper presents a novel platoon-based vehicle cooperative driving strategy at a single-lane ramp to achieve smooth traffic flows near the capacity of a connected and automated highway system environment. At the same time, we set an MPC-based ramp signal at the ramp entrance to dynamically adjust the traffic volume entering the ramp according to the traffic congestion status downstream of the ramp. The simulation results show that the proposed strategy can reduce cumulative fuel consumption and travel time.

Timely and accurate traffic prediction has gained increasing importance for traffic management. This study proposes an improved k-nearest neighbor (KNN) model to enhance prediction accuracy with consideration of spatiotemporal correlation. This study tries to find more suitable nearest neighbors by adjusting the influence of time and space factors on the state matrix. Four different methods are tried in this study to weight the state matrix to improve distance measurement in KNN. The method using the Gaussian function to weight the time dimension and the correlation coefficient of the velocity series to weight the space dimension (KNN-GC) performs best. Compared to original KNN, the accuracy of KNN-GC increases by 8.21%. Besides, KNN-GC significantly improves the multi-step prediction accuracy and consistently outperforms the competing models when the prediction step is within 30 min. Consequently, the spatiotemporal weighted KNN method is promising in short-term traffic prediction.

Multiple vehicle-pedestrian collision accidents show that drivers and pedestrians are unaware of the danger, increasing risks. A potential solution is to alert the vehicles and the pedestrians with the aid of the Cooperative Vehicle Infrastructure System (CVIS). This study proposes a vehicle-pedestrian two-way collision alert system to alert drivers and pedestrians under a CVIS environment, where real-time status information of traffic participants can be obtained. The states of vehicles and pedestrians are first predicted using Kalman Filtering to offset the system time delay, followed by the collision point identification. The risk indicator Time to Collision (TTC) is then calculated with kinematic formulations. If the TTC is within the alert range, alerts will be sent. A system architecture is established to realize the algorithm. The proposed method can be applied dynamically to cater to time-varying conditions. Numerical studies and field tests validate the effectiveness of the proposed system.

The unmanned aerial vehicle (UAV) monitoring platform has the advantages of flexibility, traceability, convenient storage and wide-view, but the challenge is to extract the traffic parameters from the UAV-videos accurately. This paper proposes a YOLO-X deep learning approach to detect vehicles using the DJI Phantom 4RTK UAV and vehicle on-board unit. The UAV is used to monitor the traffic situation at various altitudes, and the vehicle on-board unit is used to record the accurate speed of the experiment vehicle. The YOLO-X deep learning framework and the video frame-based detection algorithm are incorporated and a field experiment was conducted. The model was trained based on collected and aerial data processed by the deep learning neural network, and compared with the ground vehicle data. Results show the accuracy of target detection is 92.58%, and the target vehicle speed detection is 94%. This demonstrates the proposed approach is promising for traffic target detection.

In air traffic flow management, air traffic flow forecasting is the most important part; it not only can improve security and effective utilization of airspace and airport resources but also can greatly improve airlines’ economy benefits and operational efficiency. In this paper, the Elman neural network prediction method, which achieved relatively good results in prediction of ground traffic, is applied in air traffic flow prediction, and a new air traffic flow prediction method based on multi-feature Elman neural network is proposed. First, the velocity characterized by the first derivative and the acceleration characterized by the second derivative are introduced as two new features into the structure of the single-feature Elman neural network, and a multi-feature Elman network is built. Further, the parameters of the network structure are studied by using the steepest descent method with the driving quantity items. The air traffic flow of 36 cities in East China are tested as the experimental data of the proposed method. Experimental results show that multi-feature Elman method compared with single-feature Elman method can obtain better prediction results.

Abnormal detection has become an essential method of security protection in automotive. In order to meet the compatibility between electronic control units (ECUs), the data transmitted in controller area network (CAN) bus need to obey different protocols and specific communication rules. Generally, these rules can be learnt through statistics and used in the abnormal detection of in-vehicle networks. However, satisfactory abnormal detection performance is hard to guarantee when the in-vehicle network communication rules is relatively simple. To improve the detection performance comprehensively, this paper chooses the classification algorithm to carry out anomaly detection. Considering the particularity of in-vehicle networks, this paper proposes a classification algorithm based on the feature vectors of CAN bus packets. Combining with the feature vectors, the convolutional neural network (CNN) algorithm is used to realize high detection performance for in-vehicle networks. Moreover, the real vehicle experiment has verified the satisfactory results for the proposed classification algorithm.

We develop an evolutionary model based on Petri network to describe the development process of the transportation system which is driven by internal requirements and external technologies. The relationships between the basic elements in the transportation system are thoroughly discussed to build the model framework, including the traffic services, components, functions, technologies, and requirements. A text mining method is applied to define and quantify the attributes of each participant in the evolution model. A new transition rule that keeps the original number of tokens in the place after the occurrence of its transition is developed to active the dynamic evolution of the transportation network, where the number of tokens in each place denotes the state of the participant in the development process of the transportation system. The development of a service about vehicle visual perception is taken as an example to show the effectiveness of the evolution model for the transportation system.

The adaptive Monte Carlo localization (AMCL) algorithm is commonly used for localization tasks for automated mobile robots (AMRs). However, when AMRs move to a feature-less environment, AMCL shows poor performance in localization. We propose an improved AMCL algorithm to improve the accuracy and robustness of the localization for AMR. We first matched the laser scanning points with the pre-built grid map according to the localization result generated from AMCL. Then, we designed a localization credibility estimation (LCE) to evaluate the localization performance, and the match with the higher LCE score was selected and injected into the particle swarm with an adaptive amount to optimize the next estimation process of AMCL. The particle swarm of AMCL gradually converged to the correct location through iterative optimization. Results demonstrate that the proposed AMCL algorithm is superior to the traditional AMCL algorithm and previous improved AMCL in terms of accuracy and robustness.

Along with the transformation of transportation system toward autonomous transportation system (ATS), which impels its ability on intelligence and automation consistently and gradually, how to reasonably analyze the evolution across its development stages is essential to support an orderly development. Since the data explodes in the information era, contexts encoded in related knowledge bases become even more complex to be analyzed for the extraction of evolution details. To tackle such a challenge, an evolution analysis mechanism based on the knowledge graph (KG) is proposed to support knowledge modeling, evolution analysis, and information extraction. Moreover, through an evaluation based on a common set of KGs designed to store information about technologies, functions, and services to support autonomous driving in different levels, the merits of proposed mechanism are revealed in detecting more accurate evaluations with a reduced number of operations to rebuild the evolutions across KGs.

With the rapid improvement of self-organized operation and autonomous service capabilities of the transportation system, China has proposed autonomous and safe driving in road transport as the next-generation transportation system. The construction of logical architecture for the next-generation transportation system makes the system safer, more efficient, more convenient, greener and more economical. This paper proposes three methods to analyze and evaluate the logical architecture for autonomous and safe driving in road transport. Cost-benefit analysis mainly evaluates the economic effects of the logical architecture, multi-criteria decision analysis mainly evaluates the logical architecture from the perspective of environment, convenience, and efficiency, complex network theory evaluates the safety and reliability of the logical architecture. Based on these three methods, this study provides theoretical reference and guidance on the construction of the logical architecture for autonomous and safe driving in road transport.

With the acceleration of urbanization and the continuous improvement of people’s living standards, people rely more and more on cars to travel, resulting in serious traffic congestion. We need to promote the development of regular transit to alleviate traffic congestion, and the use of public transport is mainly concerned about the punctuality of its operation. This paper takes the characteristics of the bus operation as the research object, it analyzes the punctuality rate of the bus operation by collecting the real-life traffic data of the Nanjing Bus Line 3, and analyzes the factors that affect the punctuality performance of the bus operation. The results show that, the time periods, whether to set up bus lanes, the distance of the adjacent stops have a greater impact on the punctuality rate of the bus operation. Finally, proposing suggestions to improve the punctuality rate of the bus operation.

To achieve the economy, green, and high efficiency, the platoon of autonomous vehicles on the highway is a key service for the autonomous transportation system (ATS). At present, there have been few works related to the service components and the cooperation relationship in service architecture of platoon. Therefore, this paper first abstracts the basic services and operational processes based on the literature analysis. On this basis, the platoon service architecture is constructed and then it is divided into three generations, where the evolution of service component subset, then the cooperation relationship is analyzed. Third, the subset of the required service components and their cooperation relationship are analyzed for the typical driving obstacle encounter scenario, according to whether the roadside can facilitate the out-of-team awareness through communication. In addition, an example is used to verify that the intergenerational improvement of service architecture can improve system performance brought by technological progress.

Uncertainties always lie in predicting travel time along a signalized corridor or dynamic urban network. Frequent interrupted traffic flows, varying signal timing schemes, crossing traffic, and dynamic route choices make arterial travel time estimations much more challenging than freeways. This paper presents a study of developing an Artificial Neural Network techniques and Bayes algorithms for signalized corridors travel time prediction. Two types of models are tested with the data source obtained along the US-27 corridor in Ohio, including hourly instantaneous travel time prediction model and realized route travel time prediction model. The study suggests that the proposed methods can effectively capture the travel time patterns by combining a base profile and an ANN-Bayes-trained dynamic profile, and make it more sensitive to a variation aroused by non-recurring congestion. The testing results indicate a good performance of the predicted travel time models that the most errors fall into the range of 0.608 to 0.485, much lower than the threshold standard deviation (2.24) and close to 6% of mean value (0.077).

As the core force of a new round of scientific and technological revolution and industrial transformation, artificial intelligence is promoting the upgrading of traditional industries, driving the rapid development of “unmanned economy,” and has a positive impact on people’s livelihood in intelligent transportation, intelligent home, intelligent medical care, and other fields. This article focuses on the artificial intelligence development status and application prospects in the traffic control field. By tracing the history of traffic signal control and artificial intelligence, analyzing the limitations of existing hardware and software architecture and the lag of the algorithm in the field of traffic control, to discuss how to apply artificial intelligence technology to traffic signal control systems better, including front roadside unit, the middle signal control machine, and the back end signal control server, so that traffic facilities can see, listen, say, calculate, think, make decisions, and dispatch, to establish an all-round, multi-level traffic control “brain,” make the road traffic control more accurate, more real-time, and more convenient. At the same time, the application of artificial intelligence in the field of transportation, such as driverless driving, intelligent dispatching, route planning, and other applications, is expanded. The issues related to privacy protection, traffic laws, ethics, and others that may be brought by artificial intelligence are considered, and the evasive measures and future development direction are discussed.

Electronic Toll Collection (ETC) gantries are beneficial on highways and feature non-stop charging, improve highway traffic efficiency, and alleviate congestion. ETC gantry data is highly accurate, however, abnormal data detection can occur in some adverse driving environments. To improve the accuracy of flow data, a multi-source data fusion model based on K-means RBF neural network is proposed. A multi-source data fusion is carried out using traffic survey and ETC gantry data. The fusion result is compared with the detection result of the ETC gantry and other fusion methods. Results show the model has adaptive learning characteristics and realizes the high-precision correction of the running state of traffic flow in abnormal environments by learning the historical law of multi-source data which overcomes the problem of data detection of a single detector due to a bad driving environment and demonstrates the potential of multi-source data fusion in the transportation field.

Technologies such as big data, Internet of Things (IOT), 5G communication and artificial intelligence have granted modern vehicles with advanced functions like auto-pilot and blind spot detection. However, individual human beings do not receive these benefits directly. Turning sweep area, or the so-called death crescent, of vehicles and especially of large trucks and buses has caused many deaths of pedestrians at home and abroad. This paper proposes an automatic traffic safety alert system (ATSAS) for pedestrians to be capable of directly seeing the moving trajectory area and the turning sweep area of a large vehicle, thus avoid stepping into the danger zone. Mathematical models to calculate instantaneous moving trajectories and sweeping areas of a vehicle are presented first, followed by derivation of LED beam lamp angle control to project the crescent turning area on the ground. Promising results are demonstrated to build a safer and smarter society.

The heterogeneous traffic flow will be composed of Human-driven Vehicles (HVs), Automated Vehicles, and Connected and Automated Vehicles (CAVs) on future roads. This paper investigated the stability control problem for heterogeneous traffic flow at intersections considering two important aspects: both safety and efficiency. Building upon driving behavior characteristics, we propose a calibrated Full Velocity Difference model to capture car following behavior of HVs. The car-following control models are also established for AVs and CAVs, which are collectively called intelligent and connected vehicles (ICVs). We use a transfer function to analyze the car-following stability regions in different car-following models and use safety indicators to analyze the safety of heterogeneous flows under different control strategies. Simulations verified the best strategies for each scenario under various penetration rates of ICVs.

The popularity of mobile Internet enables passengers to use ridesharing services easily. Automatic vehicle technology indicates that car will be driven by a control system of vehicles instead of the drivers themselves. However, the choice of the driver for participating passengers is currently determined by subjective judgment. Therefore, the current model of passenger selected driver does not apply to future automated vehicles. To solve this problem, this paper presents the pattern and system of ride-sharing with self-driving cars. We propose a carpool passengers selected model (CPSM) of automatic vehicles and uses the heuristic algorithm to solve it. Through simulation of an actual situation, it was found that the average travel time of passengers at CPSM was shorter than that of passengers chosen randomly with approximately equivalent passenger loads.

Shared cars provide an alternative service for traveling, which has attracted extensive attention from the operators. To this end, this study addresses the competitive location and size problem for car-sharing operators. A bi-level linear programming model is developed to optimize the location and size of stations for the new-coming operator under the condition that a car-sharing operator exists already. In the upper-level model, all operators’ total demands satisfied are maximized. In the lower-level model, all travelers’ costs are minimized by choosing one of the car-sharing operators. The logit model is applied to describe this kind of behavior of travelers. The Karush-Kuhn-Tucker (KKT) conditions and big-M method are jointly used to transform the bi-level model into a single-level linear programming model, and then it can be solved by CPLEX directly. Results indicate that the new-coming operator can occupy a certain market share by setting a lower price.

To reveal the impact of connected human-driven vehicles (CVs) on CF behaviors in a real traffic environment, this study constructs a car-following guidance (CFG) system test platform based on CVs, followed by multiple field tests verifications. Specifically, the CFG system is established by incorporating hardware, i.e., Global Navigation Satellite System (GNSS) facilities, Dedicated Short Range Communication devices, personal computers, and human-machine interfaces. In essence, a CFG software is developed with a series of operations, including GNSS data obtaining, tagging and sending, subject and surrounding CVs’ GNSS data processing, desired speed generation, and data storage. Field tests of the developed CFG system are conducted at the connected and autonomous vehicle test field of Chang’an University. Fifty participants’ driving behaviors based on given scenarios are collected and analyzed. The results demonstrate that the CFG system can significantly improve the capability and stability in tracking the preceding vehicle.

To avoid collisions and long waiting times on the junction from freeway off-ramp to urban roads in mixed traffic, the first-in and first-out (FIFO) algorithm and intelligent driving model (IDM) are present. Then the cooperative control strategy is proposed from the perspective of the cyber-physical system to guarantee that the mixed vehicle platoon with connected and autonomous vehicles (CAVs) and connected human-driven vehicles (CHVs) can consistently travel within the traffic signal light proximity. The fuel consumption is employed as an evaluation metric to verify the reliability of the developed control strategy. The experimental simulation results show that the proposed control method can make the mixed vehicle platoon consistently and stably pass freeway off-ramp to the Vicinity of Traffic Signal Light (VTSL) in a green phase and improve the fuel economy.

Traditional traffic signal control methods have not played an effective role in alleviating traffic congestion. In order to make better signal control decisions, this paper combines deep recurrent Q network (DRQN) to carry out in-depth research on the optimization of signal controls. The core idea of this approach is to set up a deep neural network to learn the Q-function of reinforcement learning from state inputs and performance output. Specific state, action and reward functions were defined in reinforcement learning, a long short-term memory (LSTM) network was used to fit the state, then the appropriate signal control strategy was proposed. SUMO software was used for simulation experiments to create a single intersection traffic signal control.

Public transportation scheduling problem is a complex problem considering various influencing factors (such as passenger flow, etc.). However, the effect is not ideal when using the traditional machine learning method. Therefore，this paper aims to propose a Deep Reinforcement Learning Proximal Policy Optimization (DRL-PPO) algorithm for public transport scheduling according to different actual passenger flow. At the end of the paper, we make bus arrival time records and passenger OD data of Guangzhou route 560, for three months from July 1, 2019, to September 30, 2019 for training. Compared with GA algorithm, PPO algorithm is better than GA in convergence speed and final effect, which verifies the effectiveness of PPO in bus scheduling with complex passenger flow characteristics.

This study develops a Surrogate-based Optimization algorithm with Dynamic Adaption of perturbation search of All Dimensions (SODA-AD) to calibrate model parameters for traffic simulators. SODA-AD was created on the basis of Dynamic coordinate search using a Response Surfaces (DYCORS) with Radial basis function (RBF), it incorporates two critical improvements to balance global and local search, which are generating four groups of candidate points with unequal perturbation evenly in decision space and performing dynamic adaptation of the perturbation radius. In this research, SUMO, a widely used microscopic traffic simulator, is applied to model a real experiment scenario, i.e., an artery consisting of four T-shape signalized intersections in Nanchang, China, and a SUMO-Python traffic simulation optimization platform is built for calibrating the simulation model parameters with the proposed algorithm. The experiment results indicate the suitability and efficiency of the SODA-AD algorithm for calibration of the microscopic traffic simulator in stochastic environment.

This paper proposes a network traffic control method to enhance model-based traffic control with data-driven online adaptive optimization. A macroscopic traffic flow model is first developed for the model prediction. Then, the model is further enhanced in real time based on the performance measurements using the adaptive optimization or learning algorithm. Integrating the data-driven optimization into the model-based predictive control, the proposed control method is able to identify the key model parameters and optimize the actual network performance. Experiments on a toy network are conducted to test the efficiency of the proposed control method. Simulation results show that the proposed method generates better control performance than the general model-based control method.

The autonomous transportation system (ATS) is the next development direction of intelligent transportation. To construct the functional architecture of ATS scientifically, an algorithm of functional architecture is presented. First, the road emergency scene is selected, and the functions are analyzed in this scene by the text frequency, the text vector space, and the Density Peak Clustering algorithm. Second, the Silhouette Coefficient is applied to evaluate the clustering results, which determines the definition of multi-layer structures for the functional architecture of ATS, namely “autonomous perception, autonomous learning, autonomous decision, and autonomous response.” Finally, the integrated analysis results of the case study are displayed. The mean of the clustering coefficient is 0.78. The results show that the proposed algorithm could promote the construction of functional architecture and improve the development of the autonomous transportation.

As autonomous vehicles are coming to roadways, the transportation system will gradually become automated and connected, i.e., the current intelligent transportation systems (ITS) will evolve into autonomous transportation systems (ATS). As the basic element of ATS, function should be classified into functional domains, in order to better represent the architecture of ATS. The existing functional domain of ITS is mostly classified according to the experience of experts or practitioners, which lacks methodology and criterion. This paper explores how to categorize functional text into functional domain. In view of a large number of functional texts, the study uses Latent Dirichlet Allocation (LDA) topic model to extract text features, and inputs the weight of topic distribution into clustering algorithm to get 12 categories. Experiments show that the clustering results are basically in line with engineering requirements. This study provides a reference for text processing and function classification in the architecture of ITS.

Based on the progression coordinated optimization principle, a graphical optimization method for bidirectional corridor progression orient to Ring-and-Barrier structure is proposed in this paper. The equivalent transformation methods of intersection distance and common signal cycle are given for unequal bidirectional progression design speed in the coordinated road section. According to the phase characteristics of the Ring-and-Barrier structure, the graphical optimization methods of phase sequence, common signal cycle, and offset are designed. The case study shows that the proposed method can achieve a significant coordinated effect. This method has the advantages of explicit principles, standardized processes, and simplified calculation, particularly suitable for engineering applications.

The mixing of multi-mode connected and automated vehicles (CAVs) and multi-type human driving vehicles (HV) will become increasingly common. This study aims to establish a fundamental diagram that can accurately describe the heterogeneous traffic flows with CAVs and HVs. First, the traffic flow fundamental diagram of multi-type HVs is modeled based on the field data, and traffic flow parameters are calibrated; secondly, considering the difference of driving behaviors, the influence of CAVs mixed with HVs on the traffic flow characteristics is analyzed to establish the heterogeneous fundamental diagram; then, the credibility of the proposed fundamental diagram are verified by microscopic traffic simulation; finally, the sensitivity analysis of the penetration rate and driving behavior of CAVs are conducted. The results show the driving behavior of CAVs has a significant impact on traffic flow characteristics, while the increase of CAV penetration rate improves the roadway capacity.

A vast amount of traffic flow information can be collected and stored with the large-scale deployment of traffic detectors in urban road networks. Identifying a limited number of common traffic flow patterns from a large amount of road section traffic flow information can enable traffic managers to classify and manage urban roads and greatly improve their management efficiency. Traditional classification of traffic flow patterns is based on temporal context factors. This paper will present a method for recognizing large-scale urban road traffic patterns that considers both temporal and spatial context factors. A traffic flow similarity measuring approach is constructed based on the pre-classification of historical traffic flow data, and unsupervised classification is used to recognize traffic flow patterns. This study’s findings provide a better understanding of the traffic flow patterns of large-scale urban roads.

This paper aims to use autonomous driving technology to solve traffic congestion and vehicle safety issues at the non-signalized crossing, improve safety and efficiency, and achieve emission reduction. First, this paper describes the driving scene of autonomous vehicle platoons above the L3 level (SAE J3016) at the non-signalized crossing. Secondly, this paper proposes a method of splitting vehicle platoons. Then, comprehensively considers the actual needs of different conflicting direction platoons and the dynamic characteristics of the cooperative control strategy during the driving process. This paper builds a cooperative control strategy of autonomous vehicle platoons in conflict directions based on dynamic game theory, which enables the optimal strategy for autonomous vehicle platoons in conflicting directions. Finally, with the help of the simulation platform, the results demonstrate that the strategy proposed in this paper could reduce the energy consumption of the autonomous vehicle platoons at the non-signalized crossing.

With the application of information technologies such as 5G and artificial intelligence in transportation engineering, autonomous driving system (ADS) has developed rapidly. As the basis in system planning and designing, multi-dimensional architecture is an important part of the top-level design of ADS. According to the users’ demands, we divide the user services in ADS and construct the system functional, logical, and physical architecture in the study. A multi-dimensional architecture mapping relationship of ADS is designed, and a corresponding optimization method is proposed. By calculating the correlation matrix between the functional elements and selecting different thresholds to intercept the correlation matrix elements, the optimal multi-dimensional architecture mapping relationship structure is obtained. The results obtained in this study provide suggestions for the optimal mapping relationship design in ADS.

Based on the Global System for Mobile Communications (GSM) data and Automatic Vehicle Identification (AVI) data, this paper sorts out the characteristics of these data and puts forward an estimation method of OD matrix of urban expressway network. In this paper, an estimation model of the route flow is established based on GSM data and AVI data. The model takes the minimum difference between the estimated cross-sectional flow and the actual monitored flow as the objective function, and uses the estimated route flow and the cellpath flow as constraints to accurately estimate the route flow of expressway network. Based on this, the OD matrix of the expressway network is obtained. Then a case is designed to verifies the model. This paper takes ShenZhen urban expressway sub-network as the research object, conducts case empirical research, and discusses the rationality and effectiveness of the model in a real case.

The precise perception of moving objects in a dynamic environment is an essential task for autonomous vehicles. Inspired by the excellent performance of the optical flow method in moving objects’ detection in the image domain, we adopt optical flow method for point cloud data and propose a new moving object detection method using point clouds. The proposed new method contains semantic constraints and a scene flow module. In detail, we first fuse semantic information to judge static objects, such as ground and buildings, and exclude these points of static objects; then we apply the scene flow module to estimate the motion vector field of two consecutive point clouds and predict moving or static labels to each point. At the end, exhaustive experiments with a public SemanticKITTI data set are conducted for validation and evaluation. The results demonstrate the competitive performance of the proposed method compares to the existing state-of-the-art methods.

Transportation technology innovation represented by autonomous driving and car networking has promoted the research on architecture of a new generation of transportation systems characterized by autonomous operation and service. Multi-modal transportation as a comprehensive transportation mode of road, sea, rail and air, plays an increasingly important role in the integrated transportation systems. Based on the development status of multi-modal transport and application of Autonomous Transportation System (ATS) technology, this paper comprehensively analyzes the service objects, vehicles, infrastructure and network platforms involved in the multi-modal transport process. The common service domains are identified that describes the multi-modal transport scenarios with ATS characteristics and the basic system functions within each service domain. The logic architecture of autonomous multi-modal transport is established based on functions as nodes, logical data flows as links, and service operation process. Its construction will promote the rationalization and systematic transformation of multi-modal transport system from “passive” to “active.”

To alleviate traffic congestion caused by excessive left-turning traffic at intersections during peak hours, this paper proposes a signal timing optimization method based on reverse variable lane intersections. First, the operation rules of the reverse variable lane are analyzed. Then a model of the intersection capacity, delay, and vehicle emission of the reverse variable lane is constructed. The intersection signal timing optimization model is developed and solved by genetic algorithm. The operational effect of setting up the reverse variable lane is analysed through the actual intersection. The results show that the intersection capacity is increased by 14.57%, the average vehicle delay and vehicle emissions are reduced by 21.13% and 13.43%, respectively. By setting up variable lanes in the reverse direction and optimizing the signal timing with multiple objectives, traffic congestion and pollutant emissions at the intersection can be largely alleviated and the traffic signal control effect is very good.

In this paper, we focus on the outdoor cognitive electric vehicle (EV) network by studying a novel anti-interference robust communication scheme over mixed visible light communication (VLC)/radio-frequency (RF) links. Considering underlay cognitive mode, one EV in the second network can help the primary user to communicate with another jammed EV through VLC/RF channels, which is under the interference constraint defined by the primary system. Moreover, the RF channels in the investigated network experience Nakagami-m fading while the VLC channels follow Rician fading. Considering relative channel estimation error, the closed-form system outage performance of this scheme is derived, which is perfectly match simulation results. Finally, the superiority of the mixed VLC/RF scheme to only RF scheme can be obviously observed through numerical results under different shape index of Rician fading, RF channel fading parameter and channel state information (CSI) conditions.

Traffic wave is an important object of traffic flow evaluation. The research of traffic wave reconstruction using traffic data has attracted extensive attention. However, there are still problems in China’s urban trunk roads, such as low track permeability and low detection frequency of fixed-point detectors. Therefore, based on trajectory data, combined with LPR data and fixed-point detector data, a traffic wave reconstruction method based on data fusion is proposed in this paper. Firstly, based on the trajectory data, the least square method is used to fit the aggregation wave and dissipation wave. Next, the LPR data is introduced and the length of queue is estimated by the Bayesian probability algorithm based on kernel density. The simulation results show that the proposed method can ensure certain accuracy even in the case of extremely low permeability, and overcomes the shortcomings of previous methods that are difficult to reconstruct trapezoidal traffic waves effectively.

Cooperative Vehicle-Infrastructure Systems (CVIS) are an important subsystem of intelligent transportation systems. This paper sorted out typical scenarios in CVIS, then used analytic network process (ANP) to build the network structure model for the different scenarios. An information interaction priority hierarchy model was established, which fully considered the relationship among factors and the objective information contained in the factors themselves, and determined the priority of the information interaction. This paper divided 4 typical scenarios for CVIS on highways, and each scenario corresponded to a series of services. The priority of different services was determined through the information interaction priority hierarchy model, and the ranking results were analyzed from intragroup and global perspectives. Results show that the priority of information interaction in the traffic safety and control scenario is significantly higher than in other scenarios, and there is an obvious order of priority for different information interactions in the same scenario.

Traffic state information is critical for drivers’ route choice and traffic management, numerous studies targeted on traffic state evaluation for expressways exist. Few studies have conducted traffic state prediction for expressways, and the congestion mechanism has not been explored comprehensively. This study aims to conduct traffic state prediction and comprehensively explore the congestion mechanism. Section volume and space mean speed were collected by Closed Circuit Television. Before modeling, a variables selection method including random forest and correlation analysis was proposed to solve the huge dimension scale in modeling. An ordered logit model was built, which considers the traffic states’ sequence properties, to quantitatively analyze the impact of congestion contributing factors and predict the appearance probabilities of traffic states. The Support Vector Machine model was applied in traffic state prediction for comparison. The results revealed the appearance probabilities of traffic states are significantly influenced by traffic conditions and spatial–temporal variations.

A regional bus timetable optimization model was developed to improve the public transit attraction by reasonably adjusting the headway of various periods to reduce the passengers’ waiting time and transfer time. First, topology network structure indices as node betweenness and node degree such as passenger flow at bus transfer stations are considered to evaluate the importance of nodes. Second, assessment of the system cost, the waiting time cost of transfer passengers, and the costs of schedule deviation, is conducted to evaluate the effect of the optimal method conducted in this paper. Then a regional bus timetable optimization model is developed and solved with applying a particle swarm optimization algorithm for the appropriate headway between the objective buses. To verify the validity of the model, three bus lines in Beijing were taken as an example. Results show that the total system cost can be decreased by 8.6% with this model.

Currently, the congestion of on-ramp regions and the frequent occurrence of traffic accidents easily lead to freeway congestion. The traditional ramp metering methods have hysteresis in traffic flow adjustment. Therefore, it is of great practical significance to research the intelligent control methods in on-ramp regions of the freeway. In this work, the mechanism of the active management strategies of the freeway is introduced and the theory of deep reinforcement learning is illustrated. Then, a cooperative control model using a multi-agent deep reinforcement learning algorithm VDN is designed. Simulations are conducted in SUMO compared with other methods. The simulation result shows that the deep reinforcement learning control methods, especially the cooperative control method with VDN algorithm, perform better in the downstream average flow, ramp queue length and total travel time and offers better predictability for traffic conditions. This tool can assist in improving traffic flow smoothness and traffic efficiency.

As more mobile monitoring devices are used in highway transportation monitoring, more data need to be transmitted. Many of them not only contain index data, but also contain time data and location information. These monitoring data often have a high update frequency and need to be transmitted quickly to data users. V2I communication systems are an important part of the intelligent transportation system. This paper proposes a regional spatio-temporal calibration method based on data sharing. The method can be used to realize the high-frequency transmission of monitoring data through the V2I communication system with limited transmission capacity. This paper first analyzes the influence of vehicle communication network topology on spatio-temporal calibration mechanism. Second, the working principle and implementation process of the regional spatio-temporal calibration method are proposed. Finally, the application effect of the method is analyzed and evaluated, and applicable conditions of the method in application are expounded.

With the improvement of technology and requirement, services, functions and components also change in accordance with the interrelations, which reflects as the modes changes like inhibition, replacement, activation and finally results in the evolution of Autonomous Transportation Systems (ATS). Based on the existing research of intelligent transportation systems (ITS), there are few studies that clearly describe the difference of transportation system development, which is called generation definition here. Therefore, research on period division method benefits of planning development route of the intelligent transportation industry, clarify the development stage of transportation system and set development stage goals. This paper explores a method for defining the ATS evolution process based on key components elements, which is through a systematic study of the period definition of various fields at China and abroad. Based on this, we explore and verify the core indicator of component elements as the key elements of ATS period division.

As a new trend in the development of intelligent transportation systems, autonomous transportation systems (ATS) need to carry out relevant theoretical research in advance. In this article, ATS is divided into three generations, and the reference design set is decoupled and reconstructed according to the basic components of ATS, which is used to express reference schemes of fixed scenes. Based on the changing trend of the penetration rate of autonomous vehicles and the changes in communication technology, the scene changes of autonomous driving are divided. According to the construction method of the reference design set, based on the operation logic of “perception-learning-decision-response,” analyze the main changes of autonomous driving in ATS, and explain the evolution process of ATS between generations. Finally, the change of traffic patterns at road intersections in autonomous driving scenarios is taken as an example to demonstrate the rationality and effectiveness of theoretical research.

With the increase of the highway tunnel length and the total mileage of tunnels under operation, the potential safe driving issues in tunnels are urgent problems to be solved. This paper proposes an intelligent traffic control strategy based on the tunnel control unit concept. The definition and division principle of the tunnel control unit is proposed and the model of tunnel traffic flow control strategy is established. The decision-making process of tunnel control units’ traffic flow control strategies under accident and congestion conditions is explained. Using the VISSIM simulation tool, a simulation model for the immersed tunnel of cross-river link between Shenzhen and Zhongshan (Shenzhong Link) is built and a traffic accident condition is used as a case study. Compared with the traffic flow status under controlled and uncontrolled conditions, the results showed that the traffic control strategy worked efficiently for improving traffic speed and reducing the delay time.

The main goal of this paper is to optimize the alarm disposal of traffic police and reduce the disposal time. First, we propose a method based on a combination of heuristics and multi-layer naive Bayes, which automatically sends the alarm to the traffic police units through the alarm address, and the accuracy rate of the method is close to 89% in actual use. Then, we build a model for calculating the arrival time of police officers, which can quickly calculate the police officer with the shortest arrival time and support decision-making in the process of dispatching traffic police. It is verified through experiments that the disposal time of the above two disposal segment can be reduced to less than one second.

An unreasonable signal timing scheme is the leading cause of traffic jams, reducing capacity and increasing vehicle emissions. This paper proposed a multi-objective signal timing optimization method to improve traffic operation efficiency and environmental benefits at intersections. It built a nonlinear model with the effective green time as the decision variable, considering average delay, capacity, and stopping emissions. Experiments using an intersection in Guangzhou University Town demonstrate that the proposed method with the genetic algorithm is better than the Webster model, which can effectively solve the current signal timing problem. Specifically, the average delay and stopping emissions, in this case, were reduced by 15% and 11.4%, respectively, and the capacity was increased by 17.4%. In the optimization effect of intersections, the VISSIM simulation results are consistent with the model theoretical calculation results. The proposed signal timing optimization method could alleviate congestion and pollutant emissions at intersections.

Autonomous transportation system (ATS) has become one of the important trends in the next generation of intelligent transportation system (ITS). However, there is still a lack of theoretical research on the functional architecture of ATS. Even for the existing functional framework of ITS in China and the United States, shortages such as unclear definition, cumbersome construction process, and weak correlation between functions and services still exist and need to be improved. In this paper, based on the structured analysis (SA) method, a standardized functional architecture paradigm is proposed. Then a scientific and reasonable model is established for the realization process of ATS service. Finally, the validity of the construction method of the functional architecture is demonstrated by the automatic lane change service. The proposed theoretical approach can provide ideas for the design of the ATS system framework and offer a reference for the development of autonomous transportation.

This paper investigates the application of dynamic managed lane on expressways in a vehicle-road cooperative environment as a response to traffic congestion. The serious imbalance between traffic demand and traffic supply is the most fundamental cause of traffic congestion in cities. To alleviate this problem, the paper analyzes successful cases of the foreign managed lane and aims to figure out whether dynamic managed lane can replace the traditional general lane through the technical scheme of the vehicle-road cooperative algorithm. This algorithm calculates the capacity of each lane based on the different lane flows given by the traffic flow devices. The number of lanes of each type is then adjusted to achieve improved traffic efficiency. This paper concludes that the dynamic managed lane effect is significant when traffic flow is between 900 and 1,300 veh/h, and the traffic congestion problem would be significantly improved.

In a mixed traffic environment featuring connected automated vehicles (CAVs) and human-driven vehicles (HDVs), the reserved lanes for CAVs are regarded as an improvement approach to traffic efficiency, which can reduce the conflicts between CAVs and HDVs. This paper investigates the traffic capacities under different road management strategies to seek a critical condition on which the reserved lanes for CAVs generate positive effects on traffic capacity. The estimation method of road capacity of CAVs reserved lane is proposed based on the Markov chain. Then, three right-of-way allocation strategies are designed, and their traffic efficiency is quantitatively analyzed. The experimental results show that the CAVs reserved lane would positively impact the increase of CAV penetration rate. Besides, the traffic efficiency can be improved by setting a higher speed limit for a CAVs reserved lane and opening lane restrictions for CAVs.

The architecture of the future autonomous transportation system (ATS) has more comprehensive and complex characteristics, but there is still a lack of a scientific and effective method for the expression of this complex architecture. This paper builds a new visualization platform based on the ATS architecture, and completes the clear and efficient expression of the architecture. This new visualization platform is driven by data flow between systems and information flow between physical elements, combined with the operation logic of perception-learning-decision-response, to realize the dynamic simulation of ATS architecture. Finally, the effectiveness of the theoretical method is demonstrated by taking the scenario of "passing pedestrians passing zebra crossings to avoid vehicles at intersections without signal lights" as an example.

To the better guidance of planning and construction of ATS, research on the physical architecture of the Autonomous Transportation System (ATS) is carried out. By the conclusion of Sino-American existing physical architecture framework, some problems are found such as weak theory, lack of connections and complex construction procedure. In the following manuscript, a theoretical model called physical object (PO) is proposed based on the transportation entity. Then considering the existence modes and connection characteristics from PO, intrinsic attribute and connection attribute are designed. The former explains what a PO is and what it does, and the latter describes the incidence relation between POs. Finally, the construction method of PO is introduced according to the class of PO. In the end, taking the traffic status information service as an example, the physical architecture of the service is constructed, which verifies the effectiveness of the theoretical method.

Time headway at roundabouts, as one of the important parameters to describe the traffic flow characteristics at roundabouts, is the key to study behavioral characteristics of drivers, delay analysis at roundabouts, safety analysis, traffic capacity calculation, etc. In this paper, according to the actual observation data at signalized roundabouts and unsignalized roundabouts, the change rules of traffic flow and time headway are obtained. Then a mixed headway distribution model is proposed. After chi-square test and data fitting, it is shown that the mixed distribution model can better describe the characteristics of headway distribution at roundabouts. This paper completes some related researches on time headway of roundabouts, which is helpful to study the operation characteristics of traffic flow at roundabouts, and is convenient for delay analysis at roundabouts and design of roundabouts.

We constructed a vehicular network architecture combining multi-access edge computing (MEC) to address the serious problem of delay and energy consumption increase and service quality degradation caused by complex network status and huge amounts of computing data in the scenario of vehicle-to-everything (V2X). MEC services the edge of the wireless network to compensate for the delay fluctuation caused by remote cloud computing. We derived the optimal offloading decision, communication, and computing resource allocation scheme by modeling the MEC-based V2X offloading and resource allocation in a typical four-leg unsignalized intersection. The waste of communication resources caused by the uneven distribution of vehicles on each entrance is obviously reduced. An experiment is conducted, and the results showed that, as compared to other mechanisms, the proposed mechanism can utilize MEC resources and effectively reduce system cost.

The use of communication and digital development in platoons is an approach that promises to realize real-time location awareness and motion information communication. Based on this, connected and automated platoons have been widely studied. This paper examines the dynamic safety clearance that occurs between vehicles in the platoon. Firstly, a platoon field is established for a single-vehicle based on potential field theory, which is able to characterize the risk distribution around the vehicle. In the second step, the safety distances of single-vehicle are analyzed under a typical conflict scenario. Based on the foregoing, this paper proposes a safety distances model for vehicles in the platoon. Further, we find that the safety distances for platoon lane changes are available clearance with upper- and lower-bounds, which differs from that for non-platoon vehicles. Lastly, the simulation and comparisons with other gap methods are also analyzed to demonstrate the advantages of the proposed method.

LiDAR-based simultaneous localization and mapping (SLAM) can provide reliable and accurate location information for most automated vehicles. Loop closure detection plays an important role to eliminate accumulation errors in the SLAM system. Most existing methods usually extract descriptors from low-level features such as coordinate, normal, or reflection intensity of raw point clouds to represent scenes. It is difficult for these methods to keep robustness and accuracy in the case of occlusion and viewpoint changes. In this paper, we introduce the object-level information, as semantics to loop closure detection. Benefitting from semantics, our approach can achieve better performance in complex situations. Since the number of points contained in raw point cloud data is huge and redundant, we sample the point cloud with semantic information to reduce the point cloud density without losing effective information. Exhaustive experiments on the KITTI data set show our approach achieves competitive performance compared with the state-of-the-art methods.

Short-term traffic forecasting has been a hot topic in the intelligent transportation systems field. The traditional traffic forecasting methods mostly fix traffic sensors. However, most sensors are subject to bad conditions, leading to noisy and insufficient raw data. Recent advances have provided new traffic prediction opportunities. For example, the transfer learning method takes advantage of data trained on one good dataset and transfers the knowledge to others with bad data. Existing applications do not consider the underlying data distributions sufficiently, limiting the prediction performance. We propose a transfer learning-based traffic flow prediction framework using the Balanced Domain Adaptation (BDA) method. Various regression models are fed into the framework to evaluate a good data source and predict bad target datasets. A case study using data from the Highways England is conducted. The results show that the proposed BDA-based framework can match the distributions between traffic flow datasets and significantly improve prediction accuracy.

This paper discusses the traffic flow characteristics of close intersection group and the influencing factors of traffic capacity. Combined with the design method of intersection traffic signal coordination control, by increasing pedestrian crossing time, an optimization method of average crossing time between intersections based on simulation platform output as the phase difference is proposed. The typical close intersection combination in Zhengzhou is selected as a case study. From the original signal timing scheme, the timing extension strategy and the timing shortening strategy are proposed, and the effectiveness of the optimization strategy is tested by VISSIM simulation. The results show that under the same traffic volume, the timing extension strategy and timing shortening strategy are significantly improved compared with the original control scheme. The average vehicle delay decreases by 18% and 21%, respectively, and the queue length decreases by 16.7% and 19%, respectively.

The existing vehicle safe driving warning system has some problems, such as limited sight distance, high implementation cost, complex technology, and so on. Aiming at these problems, this paper develops a vehicle driving early warning system based on smartphone high-precision positioning. The system can obtain high-precision positioning data through mobile RTK (real-time kinematic) positioning and real-time solution, determine the lane position of each vehicle, and release vehicle safe driving warning information in real time through the vehicle information of the whole road network on the cloud platform. Through the test of positioning accuracy of smartphone self-developed program, the results show that smartphone can realize sub-meter lane positioning under various observation conditions. As the terminal of vehicle driving early warning system, it releases early warning information timely and effectively. This system has strong practicability, low technical cost, and easy product upgrading. It also has practical application value.

Establishing the relationship between built environment and traffic congestion states provides a feasible direction to explore factors affecting the spatial distribution of road traffic states, and supports to meet needs of traffic management and control under the information environment. The study analyzes the spatial pattern of traffic state, establishes two regression models to identify factors affecting local traffic congestion, and visualizes the impact degree of spatial heterogeneity. Firstly, the real data are collected and processed, and congested grids are selected as study range. Then, independent and dependent variables are obtained, presented in grids by ArcGis. Third, the ordinary least squares (OLS) and the geographically weighted regression (GWR) are applied to fitting. Results show the influence and the influential degree are different in spatial distribution. Analysis of the traffic state by GWR based on grids is feasible to present the traffic state.

Due to the influence of dynamic objects and traffic rules, most of the existing methods on motion planning when applying to the dynamic urban environment show limited performance. Here, we propose a new motion planning method for autonomous vehicles by integrating the spatio-temporal map into motion planning. The proposed method contains three stages: spatio-temporal map, path planning, and trajectory optimization. First, a spatio-temporal map is constructed based on the static environments and predicted trajectories of dynamic objects. Next, an improved hybrid A star algorithm integrating the obstacle avoidance, vehicle kinematics, and traffic rules is developed to efficiently search collision-free feasible paths. Finally, a B-spline curve combined with vehicle dynamics is proposed to further optimize the trajectory, thus the autonomous vehicle can move safely and smoothly. Simulation experiments are conducted and compared with existing methods. Results show the proposed motion planning method achieves promising performance in urban environments for autonomous driving.

In the context of global carbon peaking and carbon neutrality, reducing transportation carbon emissions and increasing the proportion of green and low-carbon transportation modes are common goals pursued by all cities. To achieve this goal, transforming the travel mode structure from car oriented to public transport oriented is the most critical and difficult task. As new technologies develop, such as autonomous driving and Internet of Vehicles, the Autonomous Road Traffic System will provide the conditions to solve this problem. Based on the case of Chinese cities, this paper discusses the current dilemmas in the transformation of urban transportation travel structure, analyzes the changes of the relationship between travelers, vehicles, roads, and services in the Autonomous Road Traffic System. On this basis, the traveler-centered optimization goals of the transportation system are proposed, a new multi-level autonomous public transport-oriented mobility pattern is constructed, and its effectiveness and feasibility are presented.

Adaptive ramp metering is designated to release on-ramp flow based on the level of congestion on the mainline freeway. Under oversaturation condition, spillback on the on-ramp may occur and then the operation of the immediately upstream signalized intersection would be impeded. It is therefore necessary to address this issue through an integration of the on-ramp metering system with the signal at the upstream intersection of the on-ramp that connects with a local street. In this paper, an integrated synthetic approach is presented to investigate performance of three scenarios: traditional ramp metering system with no detector, and integrated ramp meter and the upstream signal with the support of multiple detectors and connected vehicle (CV) data source, respectively. The resulting diagrams of time-spacing vs. aggregated MOEs are visualized to intuitionally identify the deficiencies of the traditional detectors and benefits from multiple detector strategy as well as CV data utilization. The simulation results suggest positive benefits of improving traffic operation on ramp, the upstream signalized intersection and mainline freeway by using the CV data source, whereas multiple detector strategy performs better than traditional ramp metering system.

This paper presents an approach for developing a Mobility as a Service (MaaS) system architecture under Autonomous Transportation Systems (ATS), which is a new transition from Intelligent Transportation Systems (ITS). The concept of ATS is put forward, which aims to integrate all the components and users of the transportation system to achieve self-organizing system operation and autonomous mobility services. To provide a common framework to plan and implement MaaS applications in the context of ATS, it is essential to develop a MaaS architecture that captures the features of ATS. Through the general steps involved in a process in which travelers travel from the origin to the destination, we propose approaches to identify the user needs, the services and key elements of the system. With the basic components, three architecture elements are constructed: functional architecture, logical architecture and physical architecture.

The traffic flow characteristics of each grade road in the whole urban network are quite different. Based on the field data from Tianjin, China, this paper comparatively analyzes the macro traffic states of different grade roads on weekdays and weekends with the Macroscopic Fundamental Diagram (MFD). During weekday peak hours, the average flow and density curves are single-peaked in the morning and double-peaked in the evening. The inhomogeneous distribution of density results in hysteresis that makes the MFD scatters are dispersed on the expressways. The MFD is concentrated on arterials, while it appears double branches on sub-arterials. Based on the analysis above, a calibrated dynamic model is applied to investigate the traffic evolution among each grade road. The sensitivity analysis is used to uncover the interaction between different grade roads. Finally, some management strategies are proposed based on the model to relieve the congestion of expressways and arterials.

Traffic infrared target detection in inclement weather still needs further improvement. To improve the accuracy of thermal target detection algorithms in inclement weather, this paper introduced the YOLOv4 network as a detection model. By optimizing the activation function and batch size of network, we hope to observe lower loss, higher converge speed and good mean average precision (mAP). In our experiment, FLIR dataset and pre-trained YOLOv4 weights via MSCOCO was used to train the initial model, model with modified active function, and model with both modified activate function and network batch size. These models were then used to run the detection tests on the thermal data we collected. Our experiments show that the adjusted YOLOv4 model obtained a higher mAP (82.71%), lower avgLoss (0.2012%) and a higher accuracy of rainy target detection (84%) compared with other object detectors.

This paper presents a vehicle speed guidance method based on the cooperative vehicle infrastructure system (CVIS) at signalized intersections. Firstly, three speed guidance strategies are proposed to divide the vehicles at the signalized intersection into different fleets. Secondly, we establish an improved car-following model for each guidance strategy to guide the vehicle. On this basis, numerical simulation is carried out to verify the effectiveness of the proposed strategies and model. Results demonstrate that: (1) the proposed model can effectively guide vehicles in different states to accelerate/decelerate without stopping or braking smoothly; (2) compared with existing models, the proposed model can better eliminate sudden speed changes and ensure driving safety; (3) the guidance effect of this method becomes better with the increase of the length of the speed guidance zone; and (4) this method has certain superiority in controlling fuel consumption and pollutant emission.

Traffic congestion exists widely in the underground transportation of mines, which damages the mobility of the traffic system and brings safety hazards. Leveraging the advantages of connected and automated vehicle (CAV) technologies, this study proposes a novel approach for efficiently organizing vehicles in a scheduled manner. With the implementation of vehicle-infrastructure cooperation, CAVs are supposed to travel with the pace of the space-time passages that are designed on each link to fulfill their trips. In addition, vehicle trajectories are elaborately optimized to reduce vehicle delays in their trips under the premise of guaranteeing traffic safety. To validate the performance of the proposed methods, numerical experiments are tested under different traffic demand settings. The results verify that the proposed method could organize the traffic efficiently and in an orderly manner, reflected in eliminating extra delays of CAVs at intersections and concurrently guaranteeing collision avoidance.

Federated learning is a distributed learning paradigm that enables learning on multiple clients to solve the data privacy problem. This study proposes a multi-variable input traffic model based on federated learning, which predicts fuel consumption and emissions based on speed, acceleration, temperature, and seasons, and apply real-world vehicle history data for simulation. To avoid affecting predicting effect, auto-encoder is introduced to traffic model to remove outliers. The results show that the MSE after the outlier removal is at least 10% lower than the raw data, and this traffic model compared with the existing non-federal learning methods, the prediction loss value in MAE in winter only increase by about 0.2%–3.8%, the MSE in summer decrease by about 3%–17.8%. It is proved this model still have a favorable prediction performance while protecting privacy.

Urban traffic OD prediction has always been a hot research topic in the field of transportation. However, most of the existing OD prediction researches are under normal conditions, without considering the influence of holidays, temperatures, weather, and other factors. This paper proposes an urban traffic OD prediction model based on multi-source data. Firstly, traffic modes are divided based on travel trajectory, speed, acceleration and other factors, and OD data within a certain time granularity are extracted. The OD pairs integrating multiple factors is predicted based on long short-term memory (LSTM) networks. By comparing with the model without using multi-source data, the results show that the LSTM model with multiple factors has higher prediction accuracy and is a better prediction method.

With the rapid development of intelligent driving, object tracking has been widely used in many fields, such as intelligent transportation, video tracking, and human-computer interaction. As an important part of environment perception in intelligent driving, object tracking provides important information for vehicle control decisions. At the same time, as an important field of computer vision, the research of video object tracking is of great significance. With the rise of correlation filtering theory and deep learning method, single object tracking (SOT) algorithm has developed from traditional algorithms to algorithms based on correlation filters and algorithms based on deep learning. In this paper, we review the development of SOT and introduce the SOT algorithms in detail. Then we use the LaSOT data set to compare and analyze these SOT algorithms. Finally, we have made a summary and outlook on the SOT algorithm.

Traffic state perception is key to controlling autonomous vehicles, and predicting the future trajectories of agents in the context is critical in perception of traffic states. However, agents are not separated but interacting with other agents, as well as the surrounding physical scene. Currently, the best models are based on graph neural network, but prior works normally include only position and timestep without other available and valuable information. Our model incorporates more information, conforming to evolving V2I (Vehicle-to-Infrastructure) technology. We have proposed a novel relative-position-wise attention mechanism and constructed the graph not only based on distance but on yaw angle and velocity. In the main dataset used in the trajectory predictions, our model has achieved state-of-the-art performance. We have contributed a simulated dataset named Shougang Dataset upon the map of Beijing Shougang Industrial Park. We have applied our model to Shougang Dataset and obtained an excellent performance in trajectory prediction.

This paper focuses on two aspects, including the drivers’ subjective preference and the randomness of traffic lights, to construct a selection model of reasonable traffic shortest paths for drivers. Through a questionnaire, factors in people’s travel considerations are analyzed. Statistical results of the questionnaire are weighted to obtain the degree of preference for path choice. A virtual matrix is then generated to represent the traffic light color and duration of the random traffic light. Finally, an urban road network is constructed for model application. Using Monte Carlo simulation and Dijkstra algorithm, reasonable traffic shortest paths and corresponding proportions are obtained. Calculation results show that a reasonable traffic shortest path considering subjectivity and randomness is more in line with the actual situation of the driver. Due to turning differences in the intersection, reasonable traffic shortest paths and corresponding proportions for OD pairs in upstream and downstream directions are different.

The outbreak of COVID-19 and the unprecedented policies issued by the government have compelled citizens to reshape their daily travel behavior. Questionnaires were designed to obtain three main related indicators of socio-economic and demographic characteristics, primary travel purposes and patterns, and factors influencing mode choice to explore the impact of the pandemic on travel behavior and mode preferences of urban residents. By applying the online survey method with Snowball sampling techniques, a total of 591 valid questionnaire responses from Hong Kong residents were received in the designated time period. Then, the non-parametric test methods (e.g., McNemar-Bowker test and Mann-Whitney U test) were employed to implement the statistical analysis. Traffic administrators can use the survey findings to adjust current policies or delineate new policies to align with the passenger travel behavior. This will ensure more fluid and safer transportation.

Ridesplitting, as an emerging shared mobility, has gradually become one of the important travel modes for urban residents. With the spread of COVID-19, ridesplitting has been affected due to restrictions such as social distance and home office. However, few studies have analyzed the impact of COVID-19 on ridesplitting demand. This paper selects four periods before and after the pandemic as the research objects from the ridesplitting data in Ningxia of China, and compares the changes in ridesplitting demand in the four periods. On this basis, geographically and temporally weighted regression (GTWR) model has been used to explore the impact of COVID-19 on spatiotemporal factors affecting ridesplitting demand. The results show that the impact of some factors on ridesplitting demand has changed in different periods. In addition, we visualize the spatiotemporal coefficients of the model to deeply analyze the changing trends of factors affecting ridesplitting demand under the pandemic.

Air transportation, in particular, has faced unprecedented effects by COVID-19 in terms of flight cancellations and airline bailouts; some argue that the air transportation sector is probably among the hardest hit. In this study, we explore the impact of COVID-19 on air transportation as a networked system throughout the year 2020, while taking the unaffected year 2019 as reference. Exploiting recently developed techniques in data science and network science, we analyzed the temporal evolution of air transportation networks at several scales of fractality, including airports, countries, and continents. Our study provides a comprehensive, data-driven analysis, enhanced with pointers into the recent literature, dissecting the impact of the COVID-19 on aviation as a networked system. It is hoped that this work not only improves understanding of COVID-19, but also gives anchor points on how to better handle future pandemics.

With the effective control of Novel coronavirus pneumonia, the priority problem which all cities have to face is how to provide convenient transportation services for the resumption of production. Take Qingdao for example, It is proposed that the transformation of public transportation users to private transportation is the key to effectively control the spread of the epidemic before the NCP is completely resolved at the urban traffic level. On this basis, we put forward some suggestions on how to provide urban transportation services for commuting during the special period of the epidemic. Such as, we need to focus on the low-income groups who commute over long distances, the traffic environment of slow traffic commuter groups and traffic accessibility within the influential scope of core area.

Based on the GTWR model, factors affecting spatiotemporal characteristics of ridership in metro stations are analyzed. The temporal distance is converted into additional spatial proportional distance with a spatiotemporal proportional factor. Taking Xi’an Metro as an example, the building area of work, medical, and residential land is selected to influence the built environment, while betweenness centrality evaluates the station’s importance. The results show that the pandemic reduces the overall ridership, and all the factors had a greater impact during the peak hours. The ridership of working and residential land decreased most in February 2020, whose attraction has decreased to nearly 1/30. The large medical institutions around attracted 1/3 more passengers than other stations. The attraction of working land and BC was concentrated within the second ring road and the residential land in the southern city. In conclusion, early access to it can be helpful for pandemic prevention and control.

COVID-19 has had a large impact on tourism demands, preferences, and behaviors. Understanding tourist congestion perception during normal and abnormal events has become a key to formulate tourist flow control schemes and develop tourism demand management in the post-COVID period. This paper explores the correlation mechanism among tourist congestion perception, tourism information services, and tourists travel behaviors in normal and abnormal conditions. To acquire tourist travel behavior, congestion perception, and tourists’ socioeconomic attributes, a stated preference survey was done. The changes in travel information demand during different COVID-19 pandemic levels with different congestion perceptions were analyzed, and the effects of influencing factors such as information obtained from different tourism stages and tourist density were qualitatively analyzed. A tourist congestion perception model was established. The analysis provides basic data for establishment of tourist flow control schemes in tourist attractions and reasonable tourism information dissemination in the post-COVID-19 period.

Traffic operation has shown abnormal characteristics during COVID-19. This paper obtains traffic data from multiple fields in Beijing for the whole year of 2020, combines traffic operation data with the number of confirmed cases, and deeply explores the operating characteristics of road networks, public transportation, and intercity transportation at various stages during the major epidemic. The results showed that travel demand decreased significantly during the epidemic period. From the perspective of urban road network traffic pressure, the demand for rigid travel in peak hours during the epidemic recovery period is relatively large. Based on this research, it can provide decision support for the government to formulate relevant prevention and control measures and policies, thereby improving the ability of urban traffic to respond to public health emergencies.

With the further development of urbanization, the rail transit system with the integration of multiple modes is exerting more influence on regional development. “Four networks integration” will boost the sound and rapid development of regional economy. Wenzhou is an important city in Yangtze River Delta. In recent years, the government proposed to build a national central hub city. The promotion of “four networks integration” is particularly significant in this context. This paper analyzed the status of Wenzhou rail transit system, focusing on the comprehensive cohesion of region express rail and railway. Passenger flow characteristics, the distribution of lines and networks, and the train operation plan are researched at the beginning of the paper. In order to improve the connection of the first and last train of different standard rail transit systems, several optimization schemes for the first and last trains were proposed. We analyzed the necessity and feasibility of the interconnection between the national railway and the region express rail. The conclusion in this paper will be an important reference for the connection optimization of Wenzhou multi-system rail transit system. The evaluation index system is also valuable for the implementation of similar systems.

Public transport services are constantly transforming, which calls for innovative approaches to adequately model ridership in this complex environment. The increasing availability of shared connecting means of transportation transforms how passengers make travel choices in public transport systems and has solved the first-last-mile problem. This shared bicycle-on-transit travel pattern expands the catchment area of the stations and increases the coverage of rail transit ridership accordingly. However, there are current research gaps in appropriately defining catchment areas, while selecting built environment factors at station level in consideration of shared bicycle connection. To address this challenge, the expansion effect of shared bicycles on the catchment area of rail transit stations is presented and redefined by removing the overlaps. This study also deliberates the impact of shared bicycles on the range of built environment when conducting a spatial autocorrelation regression model.

In recent years, with the type of rail transit between the national railway and the urban rail transit, the city regional railway is gradually becoming an important rail transit mode from downtown to the new city. This paper analyzes the necessity of joint operations between the national railway and the regional railway. Taking the Line S1 of the Wenzhou Regional Railway as an example, this paper analyzes the through flow between the city regional railway and the national railway and comprehensively studies some technical issues on through operation such as the traction mode and power supply system, which enumerates some major difficulties for the transport organization and operation management of the city regional railway, such as the cross-line train running plan and ticket-selling. Through the research of this paper, the theoretical evidence is provided for promoting the integrated development of China’s rail transit system interconnection and interoperation.

The transportation industry is closely related to the macroeconomic development. The transportation services index (TSI) can be used to measure the relationship between them. The freight services index in Henan Province (HNFSI) is proposed based on the TSI. The HNFSI index is used to analyze the correlation, the periodicity, and the impact of COVID-19 between the development of economic and transportation in Henan Province. The results show that the correlation coefficient between the HNFSI index and the Growth Rate of Industrial Added Value which represents trends in economic development, is 0.745. The freight transportation demand increases with a certain periodicity and regularity. Due to the influence of COVID-19, the HNFSI in January 2020 decreased by 37.7% compared to the same period of 2019, and decreased by 56.3% in February. However, after March 2020, the freight service is greatly improved.

Efficiency evaluation of urban rail transit lines (URTLs) is essential in identifying operational performance. Given differences of line types in terms of layout and function, the effect of line types on operational efficiency is examined. The performance evaluation and analysis of three types of URTLs in different periods are performed using the data envelopment analysis-Tobit (DEA-Tobit) method. The typical URTLs in Beijing are selected for case study. The cross-efficiency evaluation results show that the efficiency of the connected line is the highest in peak weekday periods, while that of the connected line and loop line is higher in off-peak periods. Sensitivity analysis result shows that efficiency has the highest sensitivity to the input index of traction energy consumption and the output index of passenger turnover. Tobit regression results show the proportion of transfer stations have a significant positive impact on efficiency, whereas average trip length is negatively correlated with efficiency.

Building a new integration mode between security check and ticketing is a technical development trend of automatic fare collection (AFC) in the intelligent metro transportation systems, while the key link to achieve this integration is to establish a scientific passenger credit system. Referring to the payment credit indicator system in the field of commercial and business, this paper analyzes the influencing factors of passenger security credit from the aspects of passengers’ natural information, family situation, economic condition, and behavior performance based on the theory of behavioral science to establish the metro security credit indicator system. The analytic hierarchy process (AHP) method is applied to allocate indicator weights, and the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method is utilized to rank passenger credits. The proposed framework is helpful to provide method guidance for passenger credit management of metro.

The seat structure of passenger flow is the specific performance of passenger travel demand characteristics under certain transport supply conditions. In this paper, the influence of income level, transit time, and the relationship between supply and demand on passenger flow structure is considered as independent variables, and the proportion of seats at different classes is a dependent variable. The long and short-term memory neural network model (LSTM) based on deep learning is constructed, and the simulated annealing algorithm (SA) is used to optimize the hyperparameters of the model, forming an SA-LSTM model for passenger flow structure prediction of high-speed railway. Using passenger flow data of the Yangtze River Delta, the prediction accuracy of this model reaches 97.36%. Compared with other models, the proposed model has better prediction accuracy and generalization ability.

Emergencies in the railway network may disrupt high-speed train operations and cause delays. Therefore, it is critical to shorten the train delay time by adjusting the arrival-departure track operation scheme under emergency conditions. In order to ensure train safety, with the objective of minimizing the delay time and the adjustment complexity of arrival-departure track operation, this article establishes a real-time adjustment and optimization model of arrival-departure track operation under emergency conditions. In the algorithm design, the genetic simulated annealing algorithm is used to solve the problem quickly and with high quality. The experimental results show that the adjusted train delay time and arrival-departure track occupancy coefficient are low, which can meet the demand of real-time operation adjustment of arrival-departure track of high-speed railway station under emergency conditions.

The concept of virtual coupling train set (VCTS) is one of the latest research directions, and will fundamentally change station organization. Based on the given notion and operation features of VCTS, the requirements for its reception and departure are analyzed in combination with the station operation characteristics. The receiving route allocation model for VCTS is proposed and solved by an improved genetic algorithm with elitist and adaptive strategy. Finally, a simulation is carried out for the designed example based on a practice station. The result shows that the proposed model and algorithm can obtain an effective station receiving route allocation scheme under the requirements of safe operation and improve the efficiency of train reception and departure greatly in theory.

In view of lacking modern monitoring technology for safety management in complex transportation environments, this paper proposes to apply narrowband Internet of things technology to goods loading and securing safety status monitoring. A trinity monitoring system based on an embedded development system, including on-board, personal computer and cloud terminal, is designed. Among them, a sensor module is built at the lower computer, data collection was realized by a wireless communication module and a narrowband Internet of things module, and C# language is used at the upper computer to realize multiple functions such as data reading, storage, real-time display and threshold alarm. The system test results verify the effectiveness of the monitoring system. Its rich functions can help railway staff find the potential safety situation during railway freight transportation. Moreover, the historical data can also provide quantitative support for the optimization of loading and securing scheme.

In recent years, many scholars have studied the influence of wheel set torsion and bending deformation on wheel rail contact point. The vertical and transverse creep of wheel rail in the first-order bending natural frequency of wheel set has been analyzed, but the influence of elastic modal characteristics of the wheel set on wheel rail contact geometric parameters and wear has not been systematically studied. Most studies focus on the impact of line excitation on wheel wear and the impact on vehicle dynamic performance before and after wear, but less on the wheel rail contact characteristics caused by the structural characteristics of the wheel set. This paper systematically explores the influence of wheel set elastic mode on wheel rail contact geometric parameters and wear index and explores the influence degree of different modes on wheel contact geometric parameters and wear under different working conditions.

Railway switches are an important part of the rail infrastructure. Research has shown that the variation in some components of the turnout geometry has a great effect on the safety and reliability of the infrastructure. The purpose of this paper is to determine the risk/defects associated with switch rail, considering the switch panel (SP) and crossing panel (CP), because of the variation in the designed shape. An integrated analytical hierarchy process (AHP) with entropy objective weighting method is adopted to determine the weight of each risk/defect at different levels, based on the expert’s response. Analysis of the switch CP and SP demonstrates that wearing and plastic deformation of the switch SP are the most influenced risk-defects and can cause serious accidents at the railway switch. Two decision alternatives are presented, which include maintenance replacement of the system. The approach is a useful tool for optimal decision-making for infrastructure managers.

Short-term passenger flow prediction plays an important role in the guidance, control, and management of intelligent transportation systems. Aiming at the problems of unclear influencing factors and low prediction accuracy of the current short-term prediction methods of passenger flow, this paper proposes a passenger flow prediction method based on gradient boosting. Based on the spatio-temporal correlation passenger flow, the features that may affect passenger flow are extracted from the inbound AFC (automatic fare collection system) data of urban rail transit. The data set is aggregated by the passenger flow every 10 min. Finally, the LightGBM (light gradient boosting machine) model is established to realize efficient and accurate short-term passenger flow prediction. Experimental verification based on the AFC inbound data set of Nanjing rail transit shows that the prediction accuracy of the LightGBM model is higher than that of the Fbprophet model.

The sidings of Turpan modern tram Line 1 and Line 3 were studied in order to cope with the sudden increase of passenger flow pressure in tourist season and improve the operation efficiency of the tram line network. The characteristics of tram operation, such as flexible organization, moderate speed, and network effect, the design method of sidings were summarized by reviewing the relevant research literature published at home and abroad. In addition, the Turpan tourist tram was analyzed as an engineering case. On this basis, the setting method of the starting and ending stations and the siding form of the transfer station were focused on analysis through the study of the modern tram siding setting principles. Finally, two siding schemes were proposed to develop a comprehensive comparative analysis on the traffic organization of trams. The results show that one of the schemes has obvious advantages.

When evenly distributed overlong cargo weight is loaded on a flat wagon with one overhanging end, the allowable load of the overlong cargo is affected by the gravity center deviation of the cargo from the lateral centerline of the wagon and the vertical dynamic load generated by the cargo. At present, China’s “Railway Freight Loading and Reinforcement Rules” considers only one of the influencing factors of the allowable load of evenly distributed overlong cargo weight with one overhang end. The consideration is not comprehensive. Therefore, this paper comprehensively considers the two influencing factors to carry out simulation experiments. Finally, the leading factor affecting allowable load of evenly distributed overlong cargo weight with one overhang end is determined, and the suitable condition of each regulation is discussed. This paper can provide a scientific and theoretical basis for improving railway cargo loading theory.

To ensure commercial car transportation safety and avoid the contact of commercial cars with special railway vehicles, a method of restoring commercial car motion attitude was designed, which was motivated by the transportation needs of commercial cars loaded in special railway vehicles. This method can obtain the maximum longitudinal and transverse dynamic displacements of commercial cars in railway impact tests based on the data of displacement sensors. First, a model was established to describe the position relationship of points marked on commercial cars. Second, the numerical solution of the model was obtained by the Newton method. Finally, the translational and rotational transfer matrixes were deduced. The coordinates of points on the commercial car can be obtained by using the translational and rotational transfer matrixes. Furthermore, based on the commercial car danger points, the maximum longitudinal and transverse dynamic displacements of commercial car can be obtained.

As train running speed increases, wheel-rail excitation force increasingly influences stability of bogie frame structure. By analyzing frame load and wheel-rail force, the transfer function can be constructed. In the time and frequency domains, correlation features are used to obtain their transmission characteristics. Combined with rigid-flexible coupling multi-body dynamics, the whole vehicle dynamics model of a high-speed train is established. The vertical force of wheel and rail and the vertical load of the frame under different working conditions are obtained by simulation. Their correlation is analyzed by Pearson coefficient and correlation characteristics are extracted. The power spectrum of wheel-rail vertical force and frame vertical load under different working conditions is obtained by short-time Fourier transform method, and its transmission characteristics in frequency domain are analyzed. This study’s findings are of great significance for exploring the influencing factors of the structural stability of the bogie frame.

Passenger flow prediction in urban rail transit is a fundamental task in rail transit construction, and accurate passenger flow prediction can avoid the risk of waste or shortage of rail operation resources. The article proposes to apply XGBoost, an integration-based machine learning method, to rail passenger flow prediction. Taking Xi’an Metro Line 2 as the research object, the swipe card data from April to May 2017 for the whole line was used. The passenger flow characteristics of special holidays, weekdays and rest days were analysed, and the study time period was determined to be a weekday with more concentrated passenger flow distribution. The prediction results of XGBoost were compared and analysed with those of BP neural network model and ARMA model, and it was found that the XGBoost algorithm has higher prediction accuracy and can provide reference for rail transit construction prediction.

With the popularization of the Internet, technology is increasingly accepted, and the full support of the local governments to cross-border industrial, cross-border e-commerce has been developed rapidly. However, the development of cross-border e-commerce is restricted to a certain extent by the infrastructure of cross-border e-commerce logistics, distribution costs, and difficulties in returning and exchanging goods. This paper proposes four profit distribution strategies based on the idea of cooperative game: (1) The contribution distribution strategy based on Shapely value; (2) the profit distribution strategy based on the unequal power status of the alliance; (3) the profit distribution strategy under the structure change in the alliance; and (4) the profit allocation strategy based on risk-taking, through reasonable benefit distribution mechanism; the alliance can run stably for a long time.

With the improvement of metro networks, sustainability in influencing area of stations has been a focus in recent research. The gravitational model between cities from a geographical perspective has received great attention, but there is a lack of research effort in explaining the gravity form for metro stations. Based on Beijing’s points of interest and shared bicycle data, we propose a gravity model with five indicators. The model’s effectiveness is verified by the mixing proportion of shared bicycles in the paired stations. Results show that there is a correlation between the gravity and mixing proportion of shared bicycles, with a correlation coefficient of 0.989. Of the 13 pairs of selected metro station catchments, 84.62% of the functional mixing degree was above 1.7, and the average connection ratio of shared bicycle was 3.89%. Therefore, the gravity model proposed provides a basis for the sustainability of urban planning and management.

The land port logistics hub is a significant logistics infrastructure combining the intermodal transport, storage, consolidation, etc. Unlike general facility layout planning issues, internal and external traffic organization have an important influence on the layout of land port logistics hubs. This paper constructs a traffic network logistics hub considering truck travel time which distinguishes various types of intersections. A multi-objective model is proposed with the objectives of minimizing total truck travel time and maximizing internal correlations. Transit distance and time are both considered in total truck travel time. According to the characteristics of land port logistics hub, several constraints are added to the model, including the coverage of existing facilities, facilities’ size, direction, and entrances’ location. An improved algorithm based on NSGA-II is designed, which applies multiple mutation methods and slicing tree for chromosome code. The results suggest the designed algorithm is effective, and the layout scheme is reasonable.

Street oriented walking environment assessments are of great significance to the development of green transportation. 35 groups of eye movement data of pedestrians walking along the street were collected based on the eye movement instrument. The factors were identified and detected by using segnet and Yolo image recognition algorithms. The influencing factors of walking environment were screened based on physiological indicators, and the walking and riding environment index system for convenience, safety and comfort was established to construct the street scale walking index. Using the renewal of Ping’an Street in Beijing as an example, an empirical study was carried out on the 16km roads on both sides of Ping’an Street before and after the renewal, and the correlation analysis was carried out in combination with the residents’ travel survey data of 17 surrounding residential areas. We found that after street renewal, the walking environment of Ping’an Street in Beijing is improved. With the renewal of the street, the walking of the street shows a decreasing trend from the center to the periphery, in which the decreasing trend of the convenience index is the most obvious, followed by the safety index. There is a significant positive correlation between walking index and walking travel sharing rate of community residents. Finally, we propose policy suggestions for the improvement of walking environment in Beijing.

With the rapid growth of urban distribution, there are many problems that can culminate, including traffic congestion. Based on system dynamics theory, this paper constructs a dynamics model of urban distribution capacity system using systemic thinking, in which the model contains a social subsystem, vehicle subsystem, and environment subsystem. A cause-effect loop diagram is drawn and clarified by Vensim software to sort out the relationships and logical loops within the subsystems. Its purpose analyzes the factors affecting the change of distribution vehicle capacity and correlates it between the elements within the total system of capacity. For urban distribution in China, it is necessary to use the concept of sustainable development to establish the concept of green distribution into practice and implement it into the actual distribution work, so as to bring about the long-term development of the distribution economy.

This study addresses a widespread practical simultaneous pickup and delivery problem, where pickup and delivery requests for transporting are non-stackable rectangular items. To solve this problem, a branch-and-price approach is proposed based on a set partitioning formulation. In the approach, a strong labeling algorithm with an efficient packing heuristic are proposed to deal with the exploration and the packing procedure in pricing problem. Extensive experiments based on the widely used benchmarks are conducted. The results demonstrate the efficiency and superiority of the proposed approach for solving the 2L-SPD. Moreover, comparison results infer that splitting customers with both pickup and delivery into pickup-only customers and delivery-only customers increases the computational burden under the branch-and-price approach.

Transit Metropolis construction is the inevitable trend of large-sized city development in the future in China. Under the background of achieving national target of carbon peak and carbon neutrality, implementing the public transport priority development strategy, strengthening Transit Oriented Development, accelerating the construction of Transit Metropolis will comprehensively improve the development level of urban public transport and result a significant CO₂ emission reduction benefits simultaneously. Therefore, this paper firstly concludes the primary results have achieved in Transit Metropolis Construction, then analyzes the main impacting factors of Transit Metropolis Construction to promote the carbon emission reduction, including optimizing the urban space, adjusting the travel structure, promoting low carbon vehicles, changing the concept of travel. On this basis, the CO₂ emission evaluation model of Transit Metropolis Construction is put forward and six typical cities are selected to quantify the CO₂ emissions by the model. The typical cities can achieve about 800 thousand tons of CO₂ emission reduction benefits per year. If calculated according to this, constructing Transit Metropolises across the country can achieve CO₂ emission reduction benefits of more than 10 million tons per year.

In the 21st century, the issue of addressing global climate change has become more prominent. The European Union (EU) has made climate change a priority of governance for a long time. Meanwhile, China has attached great importance to green development and is trying to achieve the new long-term targets, to reach peak emissions before 2030 and achieve carbon neutrality by 2060. This paper systematically combs EU’s transport policy in the field of climate change, especially the sustainable and smart mobility strategy; closely focuses on the development needs of green transportation; analyzes the opportunities and challenges faced by China-EU green transportation cooperation; and finally puts forward relevant suggestions for deepening the China-EU strategic cooperation.

The metro is the main component of public transportation in China’s large cities. However, the existing Chinese standards for barrier-free facilities in metro stations have no uniform requirement for the number of barrier-free facilities. This paper uses automatic fare collection (AFC) data to study the usage of accessible elevators at entrances of Shenzhen metro stations and the characteristics associated with usage. An evaluation system was proposed from the perspectives of supply types and facility layout. It was found that the usage is distributed in a strip, which is concentrated in the stations along Metro Line 1 in Luohu District, Futian District, and Nanshan District. The usage on working days is small and variable, while that on non-working days is large but relatively stable and the usage changes periodically. The results of this study can serve as a reference for the construction and improvement of metro barrier-free facilities in other cities.

In order to solve the problem of untimely customer response and increased delivery costs caused by the dynamic changes in customer demand for express companies, a rolling cycle scheduling method for express vehicles that combines batch processing and emergency scheduling is proposed. And the determination method of dispatching time, dispatching scope and inserting position of dynamic demand for express delivery vehicles is expounded. Then, a novel m-TSP scheduling algorithm using hybrid ant colony algorithm is designed to tackle this model. We perform a series of experiments to validate the proposed model and algorithm. Experimental results clearly show the strength of handling dynamic customer needs and the superiority of the presented scheduling method over the original method.

National logistics hub planning is a strategic construction project currently being promoted in China. As one of the classic forms, land port logistics hub becomes the core infrastructure of China’s modern intermodal transport network. Facility layout design of a logistics hub depends heavily on reasonable facility configuration pattern decision. In this paper, we analyzed the core factors affecting the facility configuration pattern of land port logistics hubs and put forward the idea of planning and construction of land port logistics hubs according to the current construction situation and problems that occur. On the basis of analyzing development status of road-railway intermodal transport patterns in the world, this paper proposed four facility configuration patterns of a road-railway intermodal transport center of a land port logistics hub and gave the adaptability analysis of the four patterns.

Drivers with certain characteristics may produce high fuel consumption in the same driving environment. However, it is not clear which personal characteristics will decrease the fuel efficiency, limited by insufficient test data of drivers with full information in the same uncontrolled scenarios. Therefore, we introduce a simulation method for calculating fuel consumption. First, agent driver models were established to reproduce the real drivers’ behavioral trajectory of car-following (CF) from the Naturalistic Driving Research Project (NDRP). Second, those models were tested in 288 CF scenes extracted from the NDRP data and the fuel consumption was calculated. Finally, the decision tree algorithm CHAID is used to identify the potential relationship between driver characteristics and fuel consumption. The result confirms that fuel consumption is strongly affected by driving style, mileage, and driving age, also showing that novice and experienced drivers consumes less fuel than drivers with moderate experience.

Express delivery business volume and intra-city express service is growing significantly with the development of e-commerce. The appearance of the crowdsourced logistics mode seems an opportunity to ease the pressure of city express service. Using the three-party evolution game method, a three-party game model of crowdsourced platforms, retailers, and crowdsourced part-time deliverymen was established. The impacts of correlative parameters on tripartite selection were studied by performing numerical simulations using MATLAB. Results show that for all the tripartite choices to converge to the ideal state, the initial scale of retailers and part-time deliverymen should exceed the critical scale, the subsidies given to retailers and part-time deliverymen should be within a reasonable range, and an advanced effective advertising campaign should be made by platform to accumulate enough potential benefits. The research provides forward-looking suggestions for the strategy formulation of crowdsourced platform and helps to improve the efficiency of the intra-city logistics.

Due to its flexibility, dockless bike sharing (DBS) becomes a solution for the ‘first and last mile’ problem, especially in the central business district (CBD) of Beijing where daily commuting highly relies on DBS. This paper evaluates the impact of factors, such as built environment, public transportation facilities, and time of day, on the usage of DBS in CBD. First, temporal and spatial distributions of CBD trips are analyzed, showing that the usage of DBS is highly influenced by time of day. The daily regression model and the time-dependent regression model are estimated, indicating the mixed area, public transportation facilities, and the number of office buildings is significant positive predictors for trip production. Further, considering the interaction effect with time of day, results show a significant impact difference between morning and afternoon peak hours, discovering more aspects of interaction effect on the impact of DBS usage in afternoon peak hours.

Countries from all over the world, including Nigeria, source their industrial and consumer goods from China. Nigeria is a West African country with the continent’s largest population and significant demand for products, and China’s industry has supported the quest to satisfy this demand. With Long delivery times by sea freight and the high cost of air freight transportation as a significant challenge for the movement of commodities from China to Nigeria, there is a need for a relatively cheap and time-saving mode to satisfy the logistics demand of shippers. This paper develops a multimodal route selection model that minimizes transportation cost, time, political risk, and security threats for the selected transportation node between the East, Central, and Western Africa. The multimodal route selection model was developed as a multi-objective shortest path problem, an Np-problem, and was solved using the NSGA-III algorithm to obtain the Pareto-optimal solution.

Enhancing road infrastructure is a vital component of poverty reduction policies. Using panel data of 31 provincial administrative regions in China from 2013 to 2018, this study aims at applying a Gradient Boosted Decision Tree (GBDT) model to explore the non-linear relationship between road infrastructure variables and rural poverty incidence in two different time periods. Here, the rural poverty incidence was taken as a response variable, and length, density, investment and hardening of road infrastructure were selected as explanatory variables. Results show that the length of highway and village roads present threshold effects on rural poverty incidence, which means poverty alleviation occurs only when they reach a certain scale. The poverty alleviation effects will keep stable after the above indicators exceed their thresholds. These findings could provide helpful references for the planning and construction of rural roads during the transition period between comprehensive poverty alleviation and rural revitalization.

In recent years, annual domestic waste production in China has increased approximately 8% yearly, and positive progress has been made in waste classification. Reducing waste collection and transportation costs, which account for 60%–80% of the cost of the entire waste treatment system, is the key to the sustainable development of the waste treatment industry. This paper studies the optimization of municipal solid waste collection and transportation plan based on the status and future of front-end waste classification. Based on the idea of disruption management, a disruption recovery model is established for the temporary changes in waste volume and customer demand, and an adjustment plan with minimal disruption to the system is generated on the basis of the original waste collection and transportation plan to realize the dynamic response of the system to front-end changes. A specific optimization plan in Haidian District is proposed according to the above method.

Around the target of “peak carbon dioxide emissions” “carbon neutrality” for urban transportation and the target of deep emission reduction path, various cities have introduced corresponding policies to promote the low-carbon and green development of urban freight transportation. Based on the statistical data of transportation industry, this paper deeply analyzes the status quo of types, quantity, energy consumption and carbon emission of urban freight vehicles. Consider combining industry development and achieve carbon up to the peak, in the framework of scenario analysis method, based on the regression analysis method, fitting the data and predicting the future energy consumption and carbon emission reduction targets of urban freight in 2035, under the natural situation, the general policy, the strong policy such three situations. And put forward work priorities and policy recommendations for different scenarios. Promote the development of green freight transport in cities, accelerate the green and low-carbon transformation of the urban transport industry, and promote the sustainable and high-quality development of the industry.

The new mobility of urban transportation has an impact on the CO₂ emission of the transportation industry, which is reflected in the increased introduction of non-motorized transportation modes such as sharing bicycles, and the popularity of online car hailing has increased the amount of motorized trips. In addition, the effect of traffic efficiency due to the change of motorized traffic in urban transport has brought indirect CO₂ emission. This study obtains the migration of new mobility from other modes of transportation through big data and survey and the fuel consumption due to the new mobility is calculated. Furthermore, the CO₂ emission through the analysis of the change of traffic efficiency and migration of urban transport modes by new mobility forms is studied. This research can provide a reference for quantifying CO₂ emission reduction of new urban mobility.

In this paper, through the analysis of the traditional coal logistics and transport process of railway-port to port-shipping connection operation process, “freight train-port-ship” coal logistics mode under the coal logistics port collaborative operation optimization scheme is put forward. A loading operation planning model under uncertain arrival time and loading time is established, and the influence of different fluctuation degree of uncertain factors on the optimal goal of planning decision is analyzed.

Customized bus demand identification can help address the issues of single service mode, low service quality, and low attractiveness of traditional public transportation systems. Moreover, it can help meet the comfort, personalization and quality needs of urban public transportation. Mobile communication terminals promote mobile phone signaling data as an important data basis for studying users’ travel behavior. This study proposes a two-step method for demand identification of customized bus based on mobile phone signaling data. A framework is proposed to identify potential demand areas based on mobile phone signaling data. This study constructs an ordered Logit model to predict transition probability from different travel modes to customized bus to estimate potential transition demand of customized bus. Kunshan, Jiangsu Province was selected as the research area. The proposed method can identify potential bus demand areas and transit demands for buses. Future research may consider planning the layout of traffic routes.

The high density of express distribution stations ensures the distribution efficiency and good customer experience; they bring high rental costs, potential safety hazards of street operations, and other problems, especially in central business districts (CBD). This paper explores delivery modes without distribution station (DMWDS) by making express distribution vehicles into mobile distribution stations instead of the traditional fixed distribution stations. A vehicle routing optimization model based on three-dimensional container loading constraints is established, and a heuristic algorithm based on Tabu search algorithm is designed. Using the vehicle routing optimization model based on three-dimensional container loading constraints can effectively improve the vehicle load utilization and reduce the distribution cost. The use of small delivery vehicles can better allocate transport capacity resources and improve the optimization level of delivery schemes. The research can provide an effective solution for express enterprises to optimize CBD distribution schemes and control operating costs.

One of the increasingly serious problems facing public transit is the mismatch between metro service capacity and travel demand. In some sections with large passenger flow, this has created a bottleneck, restricting the development of the metro. This paper studies the bottleneck section of the Nanjing metro service. The following work is mainly done: (1) Proposing a method of using the express bus line to relieve metro passenger demand. (2) Extracting passenger flow from passengers’ card-brush data and assigning it to the metro network to obtain bottleneck sections. (3) Establishing an average saturation model to calculate the most suitable number of express bus stations. The results show that the method can effectively reduce the saturation of cross-section passenger flow in the bottleneck section of the metro and improve passenger travel satisfaction on the basis of ensuring the rational use of metro capacity.

This study aimed to investigate the technical properties of Aspha-min zeolite foaming asphalt mortar. Firstly, the Brookfield viscometer test and the temperature scanning test were conducted to investigate the viscoelastic properties of Aspha-min zeolite foaming asphalt mortar. Then, the high-temperature performance of Aspha-min zeolite foaming asphalt mortar was studied through the multi-stress repetitive creep recovery test. The study shows that the Brookfield viscosity and complex modulus of Aspha-min zeolite foaming asphalt mortar show a gradual increase with the increase of Aspha-min dosage. On the contrary, the irrecoverable creep flexibility Jnr3.2 shows a gradually decreasing trend. However, the increase of Aspha-min dosage did not significantly affect the phase angle.

Appointments at container terminals are an effective way to relieve congestion. In order to adjust the truck’s arrival time at the gate, concerning the cooperation and the time window adjustment mode among external trucking companies and terminal operators in the truck appointment system (TAS), the centralized and decentralized dynamic collaborative TAS model are established respectively by mixed integer programming (MIP). Numerical experiments are performed to validate the approach in this study. The results show that the dynamic collaborative TAS can significantly reduce the waiting and congestion costs of the external trucking companies, and better stagger the peak time windows, so as to maximize the synergy between terminal operators and trucking companies. Finally, optimization measures for the scheduled arrival of external trucks outside the container terminal are put forward.

Under the explosive growth of e-commerce orders, making the accurate prediction of third-party logistics demand is a key research direction. Existing traditional demand forecasting models are generally based on time series models, which only considers a single impact factor and there is less prediction of third-party logistics orders. Therefore, this paper presents a method for combination forecasting third party logistics demand based on multi-factor fusion. First, a multi-factor feature system is built integrating Baidu search index and e-commerce order information. We use MLP multi-layer neural network, AdaBoost decision tree, and support vector machine (SVM) as the prediction models and fuse them into a combined forecasting model by Shapley method. The proposed method is applied to predict the third-party warehousing demand of an e-commerce platform between 2017 to 2018. Results show that the prediction error of this method is below 10% and considerably improves precision.

Vulnerability is a good indicator to reflect the road capacity and previous researches mainly concentrate on the vulnerability modeling in critical link and analyzing with simulated topological road network. This study proposes a new methodology for evaluating vulnerability of urban road network based on the concept of macroscopic fundamental diagram (MFD) which is derived by comparing the behavior of the MFD under actual conditions with that under theoretical conditions, while the Real-time Vulnerability Index (RVI) that enables real-time measurement and monitoring of road vulnerability is calculated as a sliding average per hour. Using empirical data from Shanghai expressway, we demonstrate applicability and validate the proposed method in real traffic scenarios. Results show the RVI is effective in reflecting spatial characteristics like bottlenecks or traffic congestion on the road in real time once they occur on the index. The index is well placed to provide management advice to road management authorities.

Identifying the mechanisms by which aggressive behavior and road conditions affect congestion is essential for operational efficiency improvement. To address the frequent congestion in the exit area of expressways, this paper proposes an interpretable machine learning based approach to congestion causation analysis. The congestion index provided by the navigation software is used to identify the congestion status of the road. The XGBoost model is used to construct a congestion model, and the SHAP is used to analyze the impact of various influencing factors on the occurrence of congestion. The results show that the XGBoost can accurately identify the congestion of expressway exits with a precision of 82.5%, a recall of 82.1%, and an F1-score of 82%; the sharp acceleration and sharp deceleration behaviors are important influencing factors leading to congestion, and more sharp deceleration and sharp acceleration aggravate the induced occurrence of congestion.

Longitudinal driving scenarios can be roughly divided into free flow and car-following scenarios. This paper analyzed the performances of 7 representative car-following models in free flow scenarios, based on comparing with empirical data. Extracted trajectories show that in free flow scenarios, the vehicle acceleration value is between –0.5 and 2 m/s², and the acceleration value reaches the maximum within 3 s. Compared with the actual trajectories, no model has the best performance in all dimensions. Specifically, from the perspective of acceleration perspective, FVD and Gipps models are closer to the real trajectories, while S-K model can better represent the uncertainty in human driving behaviors. From the perspective of speed trend, both FVD, Gipps and S-K models compared with the actual situation, have good performance.

This paper proposes a cooperative lane-changing approach that considers the motion similarity between vehicles and conducts the numerical simulation experiment to verify its effectiveness. First, we take advantage of the real-time information interaction under connected and automated vehicle environment to design the lane-changing process. Second, the motion similarity between vehicles is used to improve the Intelligent Driver Model and the Minimizing Overall Braking Induced by Lane Changes model. Additionally, we consider the preceding vehicles’ influence on the subject vehicle lane-changing intention and analyze the benefits of the subject vehicle lane-changing behavior on the traffic flows. Finally, we conduct a numerical simulation experiment based on MATLAB. The cooperative lane-changing approach is evaluated from the homogeneity of traffic flow, efficiency, safety, and comfort. The simulation results show that the cooperative lane-changing approach we proposed can effectively reduce delays, reduce the probability of vehicle collisions and reduce the braking amplitude of traffic flow.

Promoting the regional dispatching mode of public transportation involves the division of dispatching regions, route combination, joint preparation of timetables, and reconstruction of work tasks of regional dispatchers. Based on the classification of individual experience, questionnaires are used to investigate the impact of regional bus lines on the workload intensity of the dispatchers on duty. Using multiple linear regression model, the number of buses with fuels other than electric, the total number of daily departures, the number of maintenance and inspection vehicles, and the number of manual interventions per hour at peak are four factors affecting workload intensity. Through cluster analysis, workload levels are divided, and the index combinations for dispatchers’ workload intensity are appropriate. Results show this method can effectively evaluate the workload of dispatchers and is positive for optimizing the efficiency and reform of transit regional dispatching strategies, and optimizing the implementation plan of transit regional dispatching.

Illegal use of vehicle lanes by e-bikes is common in county-level cities of China. To explore the characteristics and internal reasons of this behavior, the driving behavior data of e-bike drivers has been collected on seven typical road sections in Rong County, Zigong City, Sichuan Province, by using synchronous video recording. Using a binary logistic regression model, this paper studied the influencing factors of illegal occupation of vehicle lanes by e-bikes, including gender, whether umbrella devices installed, whether there were passengers, weekday or weekend, time of day, weather, types of isolation area, whether close to the cross, and whether couriers were involved. The study shows that the incidence of illegal road occupying behavior of electric bicycles in Rong County was 47.0%. E-bikes equipped with umbrellas are more likely to occupy vehicle lanes illegally. Illegal occupation rate in rush hours is higher than normal hours, and couriers are more likely to occupy vehicle lanes illegally. The study also shows that the probability of illegal use of vehicle lanes on weekdays is higher. Illegal behavior is more likely to happen on sunny and cloudy days than rainy days. Male drivers are more likely to illegally occupy vehicle lanes than female drivers. Our research results will help develop measures for more effective e-bike management in county-level cities, thus maintaining better traffic order and reducing e-bike related accidents. Road traffic safety in county-level cities can be improved accordingly.

With rapid development of China’s economy and urbanization, the fast growth of private cars has led to excessive burden on urban roads. As an important way to control traffic congestion, traffic demand management has to be based on traffic demand prediction. This paper establishes a set of activity-travel demand models based on multi-agent simulation, which is suitable for the characteristics of Chinese travelers. We built a training set based on the historical data of people’s travel in Jiangning District, Nanjing, which was used for model parameter estimation and a case study. By importing the travel demand data obtained from the model into the open source software MATSim for traffic allocation, the complete process of activity-based model is realized. By comparing the prediction results with the measured travel data, we can find that the activity-based model can achieve higher accuracy in travel demand prediction.

Existing studies focus on surveying cyberattacks and evaluating their impacts on vehicular longitudinal behaviors on a single lane. This paper investigates the effects of cyberattacks on vehicular lateral behaviors on a two-lane highway, i.e., the lane-changing behaviors under cyberattacks. The lane change is a complex maneuver, which requires integrating neighbor vehicles’ dynamical parameters such as velocity, position and/or acceleration. This study introduces a car-following model to derive vehicles’ dynamical parameters. Specifically, this study adopts the Intelligent Driver Model (IDM) to characterize the car-following movements. Then, incorporating a classical lane-changing model, i.e., minimizing overall braking induced by lane changes (MOBIL) model, this study proposes an extended lane-changing model with cyberattacks to obtain the decision-making conditions. Finally, simulations are conducted to illustrate the impact of different malicious attacks such as falsified acceleration, speed, and position on vehicles’ movements. These results demonstrate that cyberattacks could cause traffic congestion, low traffic efficiency, and collisions.

Based on expert knowledge and data fusion methods, a Bayesian network model for accident severity influence factor analysis is constructed. The K2 algorithm and correlation analysis are used to learn the structure of the model, and the Bayesian method is used to learn the parameters of the model. The data base of road traffic serious accidents occurred from 2011 to 2018 is used to analyze the influence of human, vehicle, road and environment factors on accident severity. The results show that the accident severity does not increase with the increase of the total number of people carried by the vehicle, but is closely related to the driver’s operation behavior and vehicle safety condition, and when the vehicle technical condition is poor or the driver’s operation is improper, once the accident occurs, it is more likely to cause mass death and mass injury. The model can better identify the factors that have a significant impact on the severity of the accident. The results of the study can provide a basis for the industry management to have a deeper understanding of the predisposing factors of serious accidents and improve the safety level of the road traffic system.

Using a traveler’s travel mode choice behavior as the research subject, we assumed that there are two travel modes, car and subway. Considering the randomness of the traffic state at different times and the limitation of traveler information processing capacity, this paper internalizes the traffic state and information cost, and establishes a travel mode choice model based on rational inattention theory. We begin with a single traveler and then extend the model to a system composed of multiple travelers and study the travelers’ choice of travel mode at different departure times. It is found that when a traveler acquires more information, the traveler will change their preferences along with the change of traffic state and choose a travel mode with lower travel costs. When there is less information available, the traveler’s choice tends to initial preference and fails to pay attention to travel mode with lower travel costs.

With the development of society, autonomous driving technology has developed rapidly. It provides a new way of travel for travelers. The shared Autonomous Vehicles are the product of the combination of autonomous vehicles and sharing economy. It can provide travelers with travel services on demand. This study constructs a model of travelers’ willingness to accept a carpooling ride under the condition of autonomous driving and shared car travel and discusses travelers’ carpooling preference in the travel process. The results show that the travelers’ willingness to accept a carpooling ride under the automatic driving mode is closely related to the characteristics of daily commuting travel, carpooling characteristics and personal attributes. This paper provides a theory basis for the study of carpooling choice intention of private car commuters and non-private car commuters, and provides a corresponding basis for the formulation of ride-sharing policy, traffic resources and the alleviation of traffic congestion.

When a vehicle is moving, its speed is affected by other vehicles in its lateral position. To explore the influence of the lateral vehicle on the speed of the moving vehicle, 30 video cases were extracted and indicators like speed, acceleration, distance were obtained. A decision tree model was used to analyze the importance of factors which affect the speed change of the moving vehicle. The results show that lateral distance at the moment of decision-making is the most important factor. When the lateral distance is not greater than 0.77 m, the speed of moving vehicle would decrease. Otherwise, the speed of the moving vehicle would increase or remain unchanged. According to the frequency distribution histogram, the speed variation of the decelerated samples and accelerated samples are both 1–3 m/s. This research characterizes the speed variation of the moving vehicle because of the influence of the lateral vehicle and contributes to new ideas for studying microscopic driving behaviors.

Skid resistance largely results from the interactive and collective behavior of microscopic friction behavior occurring at the rubber-pavement contact interface. This paper presents a detailed three dimensional (3D) and high resolution pavement surface re-construction methodology based on B-rep algorithm. High resolution pavement surface data is collected by a portable laser scanning device. Linear elastic model, Mooney-Rivlin model, and Yeoh model are considered to investigate the rubber material’s influence on skid resistance. The mechanism of interface friction parameters and rubber material influencing skid resistance is discussed by means of the rubber pavement interaction finite element simulation. The influence of interface friction and rubber material on pavement skid resistance is better presented by adopting 3D re-constructed pavement roughness model. Exponential decay friction mainly affects the ability of the rubber material model influencing skid resistance. Factors that influence skid resistance are in order as kinetic friction coefficient, material type, and static friction coefficient.

As a newly emerged eco-friendly public transport mode, the community bus is widely regarded as one of the effective means of solving the “last mile” problem for urban rail service. To inform policies that promote the development of the community bus, understanding the determinants that influence the choice of the community bus is essential. Taking Dalian, China as a case, this study conducts a questionnaire survey combining stated preference and revealed preference data to investigate the determinants of the community bus travel mode choice behaviors. This study demonstrates that high-frequency travelers with fixed travel needs are more likely to choose community buses for travel. They are more sensitive to the frequency of community buses’ departures, and the walking time for transfers when changing to subway. The findings of this paper provide guidance for optimizing the network layout of community bus schemes.

The purpose of this study is to analyze the influence of the ideal and tolerable commuting time on commuting satisfaction. From the perspective of the individual subjective expectation and tolerance level toward commuting time, this study introduced the ideal commuting time (ICT) and the commuting time tolerance threshold (TTCT), and compared them with the actual commuting time (ACT). The results of the empirical study on Kunming by the ordered logistic regression model show that the satisfaction of the commuters by active commute mode is the best, while the satisfaction of the commuters by public transit is the worst. The highlight of this study is to find that the subjective commute time boundary points have a significant effect on commuting satisfaction. Concretely speaking, the respondents have the highest commuting satisfaction when their ACT is within the range of their ICT. When their ACT between their ICT and TTCT, commuting satisfaction decrease moderately. While their ACT equal to or greater than their TTCT, commuting satisfaction decline sharply. These findings could serve as a reference for achieving a more sustainable urban commute system.

Forest fires are recurrent threats to many parts of the world and often require large-scale mass evacuations. The frequency and severe consequences of wildfires makes it necessary to develop evacuation plans for vulnerable communities. Phased evacuation strategies have been adopted and demonstrated to be effective in certain emergency situations. By dividing the population into zones and evacuating them through an optimized phased sequence, the network congestion levels can be reduced. Research is scarce regarding its effectiveness under diverse fire progression scenarios. In this study, a fire evacuation model based on cell transmission model is applied to assess the effectiveness of phased evacuation strategies under slow and fast fire scenarios. Using the town of Paradise, California, as a case study where a major wildfire occurred, results show the proposed model has the ability of incorporating fire threat levels into the optimization process and finding the optimal evacuation order time and routes for each zone.

This paper summarizes the research status and progress of big data correlation analysis of road traffic accidents. We discuss the development process of correlation analysis of road traffic accidents in three stages: traffic accident investigation and statistics method, accident cause analysis method and big data correlation analysis method. Then we introduce the concept of big data association analysis and the development status of association rule mining, study the current situation of traffic accident cause analysis based on big data association analysis, and finally discuss the future development trend of big data association analysis of road traffic.

This paper aims to make an analysis of traffic congestion by truck GPS data from Gaodong toll station in Shanghai, China. Firstly, the GPS data is processed and matched to the selected study zone. Thereafter, the finite mixture model (FMM) is applied to fit the speed distribution and establish a congestion evaluation index. Finally, the working status of the tolling station can be described. From the results of the statistics, it can be concluded that the influence distance of the charging behavior is a constant related to the design of the tolling station. The truck trip characteristics do not coincide with that of commuting trips in this area, and the two forms of the traffic will affect each other.

With the rapid development of intelligent transportation systems, injury severity prediction has been paid more attention. This study intended to predict the injury severity with massive data provided. To achieve this goal, the data set was firstly collected from Traffic Accident Database maintained by Nevada Department of Transportation from 2015 to 2017, and then Getis-Ord Gi* was selected as the hot spot analysis. Based on the hot spot analysis, deep neural network was considered to predict the injury severity. By training the data set and comparing the results with deep neural network, the results showed beneficial performance from the goodness-of-fit. Findings revealed that the proposed model can be considered as an alternative to predict injury severity. The results may provide potential insights for reducing the injury severity.

Mountain highways are an important part of road traffic in many countries, but accidents occurr frequently due to complex environmental conditions and driving speeds. Therefore, long-term skid resistance evaluation and driving safety warnings of mountain highways is of great significance. In this study, three types of pavement were studied, namely the Asphalt Concrete (AC-13), the Open-graded Friction Course (OGFC-13), and Semi-flexible Pavement (SFP). Based on a novel accelerated polishing equipment independently invented by Harbin Institute of Technology (HIT), the dynamic friction coefficient deterioration of the pavement was analyzed under the coupling influence of different speeds, pavement type, and service life. The mountain highway safety evaluation system was established based on “CarSim” software under the different road slopes and line types. The results show that the SFP has better skid resistance properties than other pavement types. Driving speed should be adjusted in real-time according to different road conditions.

It is of great significance to conduct dynamic risk assessments on the operating situation of dangerous goods transportation vehicles to effectively prevent dangerous goods transport accidents and improve the safety level of dangerous goods road transport. In order to construct a dynamic risk map of dangerous goods transportation vehicles, considering the availability of safety monitoring data for dangerous goods transportation vehicles, the dynamic risk assessment index system of dangerous goods transportation vehicles was constructed from four perspectives: human factors, dangerous goods related factors, environmental factors, and management factors. At the same time, we established a dynamic risk assessment method for road transportation of dangerous goods considering the time-varying characteristics of the number of road transportation accidents of dangerous goods, and the impact factors of accident consequences based on population distribution. Using this method, a risk map can be developed to track the safety status of dangerous goods transportation vehicles in transit in real time to provide support for the safety supervision of dangerous goods transportation vehicles.

We analyzed the influence of ice melting materials on pavement icing time and reducing adhesion force of ice to study the effect of different ice-melting materials on pavement and examined the effects of ice-melting materials at low doses. The results showed that the effect of reducing the freezing adhesion force was prominent under the condition of 5 g/m² lower spreading amount, and the freezing adhesion force was only about 1/10 of that obtained from the test without using industrial salt. Under the condition of small spreading amount, different ice melting materials could achieve better ice suppression effect. Under the condition of 5 g/m² low distribution, different ice melting materials had good effects in reducing the ice adhesion force, and the adhesion force obtained by using snow melting agent is only 7.18% of that obtained by the test without ice melting materials.

In conditionally autonomous vehicles, drivers should take over vehicle control when the automated system fails, which is a challenge to perform safely in urgent situations due to critical time constraints and great stress increment. To identify how driver stress changes during urgent takeovers and the association with driver performance, this study conducted a human-in-the-loop driving simulator experiment. The urgent takeover request was designed with a time budget of 3s. Stress levels were evaluated through galvanic skin responses (GSR). A moderating effect model was then built between stress levels, minimum time-to-collision and brake reaction time. Results showed driver stress was significantly high and peaked from 5 s to 10 s after the takeover was triggered. Besides, too much stress indicated by GSR data (0.059–0.096 mμS) showed significant moderating effects on extending brake reaction time. This study could contribute to improving the takeover strategy with stress traits considered.

Autonomous vehicles (AVs) will coexist with human-driven vehicles (HVs) for a long transition period. During this period, mixed traffic flow will pose a potential threat to traffic safety by affecting human drivers’ hazard perception and decisions when interacting with different styles of AVs. This study operationalized a human-in-the-loop driving simulation experiment to investigate the effect of different styles of AVs on drivers’ perception in an overtaking situation. A trajectory planning algorithm using the risk field as the cost function was established. The results extracted from gaze behavior indicated that the drivers’ attention at AV is significantly higher for an aggressive AV than others. Furthermore, through a causal reasoning assessment based on ensemble learning and SHapley Additive exPlanations (SHAP), overtaking duration during the second lane change process was found to be the most important factor affecting driver’s gaze behavior when overtaken by AVs, which provides a direction to improve the overtaking algorithm.

As an emerging approach, driving simulator plus VR has been gradually applied in the field of driving training. There is still a lack of scientific and comprehensive methods to evaluate the effectiveness of this. Through integrating the training data of 119 driving school students, the research adopts a hierarchical evaluation method to assess the training effect of driving simulator plus VR at levels of subjective reaction, knowledge mastery, driving behavior, and the exam passing result. In terms of subjective and objective scores, the research results demonstrated that the fusion technology performs the driver training task admirably. This research provides a solution for quantitatively evaluating the training effect of driving simulator plus VR, and manifests the effectiveness of driving simulator plus VR, promoting the application of this newly emerging technology. At the same time, the discussion on training effects provides a direction for the development of differentiated and personalized training programs.

With crowd logistics becoming an essential part of the last-mile delivery challenge in many countries and regions, encouraging public participation in crowd logistics has become an important affair affecting the development of crowd logistics platforms. Learning how people are motivated to participate in crowd logistics can shed light on the industry’s policies and efforts to support its growth. We developed a research model based on the Push–Pull–Mooring (PPM) theory to illustrate the factors influencing the participation of people in Hong Kong. Survey data from 78 Hong Kong residents were analyzed. The empirical results reveal that monetary rewards, entry barriers for work, and trust are significant factors for locals to participate in crowd logistics. Meanwhile, previous job enjoyment has a moderately negative correlation with the willingness of the general public to participate in crowd logistics. However, previous job flexibility and cost of participation show the least significant impact on the decision to participate in crowd logistics. According to the findings of this study, we recommend that crowd logistics platforms provide justifiable monetary incentives and constantly review these incentives. In the meantime, they should also build a high degree of trust between the platform, the general public and the workers.

Based on the historical travel data provided by commercial driver monitoring systems, this study presents the analysis of contributing factors affecting hazardous materials (HAZMAT) truck safety. A comprehensive data set of HAZMAT truck travel records (n = 4,487), including driver characteristics, roadway characteristics, environmental characteristics, and temporal characteristics, was developed for the study period of July 1st to July 31st, 2020. A random parameter ordered probit model was established for investigating the HAZMAT truck travel risk levels, which were measured by three warning frequency categories: high risk (more than one warning), moderate risk (one warning), and low risk (zero warning). The results show that speed, acceleration, lighting conditions (dusk/dawn, dark-lighted, and dark-unlighted), visibility (less than 500 m and between 500 m and 1000 m), weather (rain and fog), and time of day (0:00–6:00) have statistically significant effects on the increase of high-risk probability.

Road rage-provoking situations are closely related to external road environment and physical and psychological conditions of the drivers themselves, which may easily cause aggressive and risky driving behaviors. Common factors such as external road environment of complex topography on mountainous highways, road traffic risk factors like sharp turns, continuous long downhills, multi-tunnel sections and dangerous roadside environment are closely related to road rage-provoking situations. Considering the complex geological environment of mountainous highways, this paper aims to investigate road rage-provoking situations on mountainous highways, and compare them to those on non-mountainous highways. On mountainous highways, the majority of the drivers (70%) are easily enraged when flashed by high beams of cars traveling in the opposite direction. Moreover, female drivers generally have better psychological states and higher endurance than male drivers in most of the situations.

Understanding travel behavior is conducive to urban planning and traffic management. In a sustainable city, green traffic modes, such as bus, metro, and bike-sharing, are encouraged to reduce greenhouse gas emissions. Recently, metro and bike-sharing have experienced fast development. Urban residents have more travel choices than ever before with the rapid development of urban economies and the acceleration of urbanization. It is imperative to understand residents’ travel behavior. Large-scale travel surveys of residents could provide various information to analyze travel behaviors. In this paper, five common travel modes are selected to analyze the characteristics of residents’ travel choice. An MNL model was constructed based on individual, family, and travel attributes. The prediction accuracy of the model is 71.0%, which indicates that the model is suitable for predicting residents’ travel choice. This study could give help to understand urban residents’ behavior and travel choices, which promotes the development of urban transportation.

Understanding tourist travel patterns is essential for tourist management and planning. This study proposes a three-step framework to extract popular tourist routes from large-scale GPS data. In this paper, tourist trajectories are identified from original GPS data based on the distance between GPS trajectory points and the point of interest. We measure the spatial and temporal distance by dynamic time warping and earth mover’s distance, respectively. Then a similarity index is used to combine them and measure the spatial-temporal similarity between tourist trajectories. Finally, the popular tourist routes are obtained by clustering tourist trajectories with high similarity index. We validated the proposed framework with a real-world GPS data set collected in Beijing, China, and conducted the sensitivity analysis of important parameters. Results of the proposed three-step framework consistent with the reality, suggesting that our framework effectively extract meaningful tourist route information from large-scale GPS data.

With the complex topography of mountainous highways, temperature, humidity, precipitation and wind are greatly influenced by the ever-changing elevations, and adverse climate conditions like rainy, foggy, snowy and gale conditions are very common. Based on a self-completion questionnaire survey, we obtained the values of subjective/perceived risk. Then, we established the objective/statistical risk for adverse climate conditions and obtained the values by employing 9,533 cases of mountainous highway crashes in Yunnan Province. The results show that there are weak correlations between statistical and perceived risk of hazardous degrees of different adverse weather conditions. The overall perception competence of the hazardous degrees needs to be further enhanced in future. The elderly over 65 years old have much less insufficient perception compared with objective risk. Both the crash experience and the injuries in the crashes that occurred under the four adverse weather conditions have enhanced perceived risk values of the corresponding drivers.

To objectively describe the lane-changing behavior of the mixed flow of cars and buses in the inlet lane of urban signal intersections, an improved full velocity difference model (FVD) model was constructed by introducing a FVD car-following model in the specific process of vehicle lane-changing, considering the effects of vehicle type and driver attributes, defining the vehicle safety spacing as a function of velocity, and classifying the lane-changing behavior into four specific forms. Afterward, the lane-changing video clips taken at the approach road at the intersection were analyzed in tracker software to obtain data, and a matlab elite genetic algorithm was applied to calibrate the model parameters. The results show that, compared with the traditional FVD lane change model, the improved model is significantly improved and more consistent with the lane-changing behavior of heterogeneous traffic in the intersection inlet lane, while it provides a theoretical basis for modeling microscopic traffic flow.

To improve the operational safety level of intersections, a method to quantify the collision risk at intersections is proposed, under the condition that traffic participants are fully aware of the movement status. Based on the concept of Post-Encroachment Time (PET), a function that can quantify the risk value at each point in the intersection is established. Considering the actual vehicle shape, the intersection risk quantification method with different time dimensions is proposed, including a real-time monitoring method of intersection collision risk and intersection risk diagnosis method. Finally, an intersection case analysis is carried out. Results show that the real-time monitoring method of intersection risks can dynamically quantify the collision risk at the intersection according to the actual movement of vehicles and provide suggestions for vehicle driving strategies. The intersection risk diagnosis method can provide a theoretical basis for signal optimization and intersection channelization design.

Unlike a traditional fixed parameters logit model, the computation of marginal effects (MEs) for the random parameter’s logit model is much more complex. In this study, a random parameters logit with heterogeneity in means and variances model was estimated using crash-severity data. Three computing methods for MEs based on global means, individual estimates, and Monte Carlo simulation of random parameters were proposed. The results, along with the software-reported MEs, were comprehensively compared. Results indicate that (1) enormous bias was detected in software-reported MEs; (2) simply using means of random parameters also produces bias; (3) the Monte Carlo simulation was most likely an effective way to computing MEs; and (4) the individual estimates method may also be reliable as the random parameters distribution has been well captured. Methods provided by this study can prompt the proper application of random parameters approach for road safety and other traffic scenarios.

Shared parking is rapidly promoted as an essential way to alleviate parking problems. However, a challenge lies in how to improve the acceptance of shared parking. In this paper, travelers are allowed to choose between destination parking and shared parking. Firstly, a Cumulative Prospect Theory-based decision model for travelers’ parking choice is proposed. Then, we develop an optimization model for shared parking space allocation considering parking choice behavior, in which rejection of allocation results is considered. Numerical experiments demonstrate that (1) the proposed optimal allocation model has good performance in improving the utilization rate, especially under large requests quantities. (2) under the optimal model, the allocation results are more likely to be accepted by travelers. Significantly less travelers reject the allocation results. (3) the parking rate has great impact in the allocation results. It is suggested that managers can encourage travelers to accept shared parking by reducing parking rates appropriately.

Autonomous vehicles can provide automatic cruise and parking services for travelers and the parking location is not limited by distance from the destination. Exploring key factors affecting commuters’ choice intention of autonomous parking, a questionnaire survey was conducted on commuters, including daily travel characteristics, willingness to park during autonomous driving, etc. A binomial logit model is used to establish a parking choice model for autonomous driving trips, and the commuters’ preference for parking choices was studied. Results show that most commuters will choose to park at the workplace or at an external parking facility within 10 km of the workplace. The total parking cost of autonomous driving and the distance between parking facilities and the workplace have a significant impact on commuters’ choice of parking location. This research can provide a theoretical reference for the construction of parking facilities and urban land policies in the future autonomous driving environment.

To systematically measure safety impact, especially reveal how the Connected Vehicles (CVs)’ operation in traffic flows improve safety, a surrogate safety measurement can be used to disclose possible frequency and intensity of the traffic conflicts, as well as the likelihood of a conflict developing into a collision. Since a real crash could be difficultly estimated, the surrogate safety index can be measured in detailed safety-related events which can be realistically reproduced from the microscopic traffic flow models through the behavior adaptation of individual CV-equipped drivers. In this study, the CV-affected driving behavior parameters are integrated into the state-of-the-art traffic simulation model. The interactions between the CV functional attributes (e.g., CV function type, message type, message timing and drivers’ compliance level to CV messages) and drivers’ behavior attributes (e.g., perception-reaction time, desired speed and desired following distance) can be virtually observed and determined. The presented method is examined in a case study of a freeway site in the Cincinnati area, Ohio. It is found out that the CV-affected reaction time, desired headway and speed contribute to the reduced conflicting frequency and intensity of vehicles in traffic flows. This finding implies positive impact of the CV safety technologies on highway systems in the CAV environment.

Adverse weather conditions can threaten driving safety. This study focuses on exploring rear-end collision risks of car-following vehicles under adverse weather conditions. From the observations of car following behaviors and weather conditions on rural highways, some indicators are designed to measure rear-end collision risk. After the trails of each candidate indicator based on collected data, the indictor, namely the following risk, is selected to evaluate rear-end collision risk. A generalized linear model is established to assess the impact of different weather conditions on the risk. Results show that in terms of the effect of a single factor, the light condition has the greatest influence on the risk, followed by weather or road conditions. Meanwhile, interactive influence of the two factors is determined by specific car-following states. The interaction between each factor and light condition always has the greatest impact on rear-end collision risk.

In order to improve the passenger distribution efficiency of large passenger stations, this paper puts forward the carpool travel mode of passenger stations. Considering meeting the needs and requirements of passengers and drivers for real-time carpooling service at the same time and improving the attraction of carpooling travel, a billing model and path planning model with time, cost and detour as constraints are established. At the same time, the algorithm is improved to improve efficiency, and the real-time link traffic volume is used to calculate the link travel traffic cost, so as to minimize the overall travel cost.

The static topological structure of urban rail transit network (URTN) was established, based on which, the passenger flow was defined as the weight to establish the weighed network. Station connection strength and weighed node betweenness was defined to reflect the structural characteristics of dynamic networks; Considering at the spatiotemporal characteristics of the passenger flow and combining the passenger flow demand characteristics, R-passenger valid path subgraph and OD loss ratio S(R) was given to evaluate the vulnerability of URTN. With Xi’an as an example, the characteristics and vulnerabilities of the URTN were analysed. The analysis has expressed that the passenger demand tolerance threshold R greatly affects the importance order of the station, network vulnerability decreases with increasing R. It can be seen that the calculation indicators and methods proposed in this paper can better reflect the vulnerability and functional characteristics of URTN.

Restricted by urban layout, land use, and road network planning, the urban expressway toll station is often directly connected with urban roads. The traffic operation of a toll station is not only affected by the traffic flow of expressways, but also by the adjacent urban roads. Therefore, in the study of the capacity of urban expressway toll stations, the capacity of toll lanes, expressway ramps and adjacent urban roads should be considered comprehensively. This paper presents a method to analyze the comprehensive capacity of the urban expressway toll station and takes a toll station as an example to analyze. Results indicate that the capacity of toll lanes is the highest, and the capacity of intersection entrance lanes of adjacent urban road or expressway ramps is the lowest, which is the bottleneck of the road network. The research results can provide a new idea for the capacity assessment of urban expressway toll station.

Drivers’ perception of risk could better predict the potential danger in complex traffic environments, which could help drivers avoid collisions. To explore the predictors of drivers’ perception of risk, hazardous scenarios were selected and a risk driving behavior test platform was built based on the driving simulator. A total of 35 drivers were recruited for driving simulations. The prediction model of the driver’s perception of risk was constructed based on the Bayesian linear discriminant method, the support vector machine (SVM), and the multinomial logit model. The models were compared based on the precision rate and F1-score. Results show that the support vector machine best predicted performance for the perception of risk. This has significance for predicting drivers’ ability to perceive risk in order to improve traffic safety in future mixed traffic conditions.

To explore the driver’s needs of design elements of signs at short-distance continuous ramp exits of the three-dimensional composite expressway. The study road section including the advance area and three decision points, that the 3km advance notice signs, the destination separation information, and the lane guidance information are studied. A demand questionnaire was conducted through the network, and 271 valid questionnaires were collected. The KANO model shows that 3km advance notice signs, the destination separation information at decision points 1 and 2, and the lane guidance signs at decision points 2 and 3 belong to A (charm attributes). The destination separation information at the advance area and continuous setting the lane guidance signs belong to I (undifferentiated attributes). The research results provide guiding suggestions for optimizing the design elements of signs at short-distance continuous ramp exits of the complex interchange.

Research has looked at spatial separation between residence and workplace and the associated traffic problems; less research has investigated the relationship between spatial separation and commuting overall satisfaction through an empirical analysis. Using a revealed preference survey, this paper focuses on commuting overall satisfaction of residents who live in different size cities in China. The formation process of commuting overall satisfaction were explored based on expectation–disconfirmation model. Satisfaction indexes consisting of safety, convenience, comfort, reliability, and economy and an overall satisfaction question were established to measure and evaluate commuting satisfaction. Survey results show commuters living in megacity and megacity behemoth have higher commuting overall satisfaction than those living in medium-sized and big cities. The empirical analysis includes a data set of coefficient variables, applying an ordered choice model in order to explain the differentiation of the commuting overall satisfaction. The result helps formulate a well-directed traffic demand management strategy.

This article investigates long-distance intercity travel and studies inter-city travel under different factors. Travel intention selection mechanism. The questionnaire survey includes factors such as personal social attributes, travel characteristics, transportation mode, and service characteristics, and also incorporates potential variables such as reliability, comfort, convenient, and economy. Based on the maximum utility theory and behavior decision theory, the structural equation model and the SEM-logit integrated model of the three modes of travel of high-speed rail/motor train, ordinary train, and aircraft were established. The relationship between different latent variables was analyzed. The research results show that the SEM-logit model with latent variables added has a good model interpretation result by improving the traditional multiple logit model. At the same time, it also shows that travelers are not only the result of some single factors in the intercity travel decision process, but many cross-influencing factors also have an important influence on travel decisions.

To improve the safety of the exit-ramp area of three-dimensional composite expressways, this paper studies the impact of different traffic signs on drivers’ decision-making response and behavior choice from the perspective of drivers based on driving simulator. A total of 32 drivers participated in this experiment. The speed, standard deviation of speed, acceleration, acceleration standard deviation, lateral position and steering wheel angle were used as indicators of the response characteristics of drivers. By paired t-Test, this paper explored the significance of various indicators, and then get the impact of different traffic sign information on driver behavior. The results show that different traffic signs have different effects on driving behavior. Setting direction separation information at the exit of interchange is helpful to promote drivers to change lanes in advance. The research results provide guiding suggestions for improving the overall level of road traffic safety and optimizing the information giving mode of complex interchanges.

This paper uses an unmanned aerial vehicle (UAV) to video the traffic flow of an urban intersection entrance road to explore vehicle movement characteristics in the marking vacuum zone (MVZ). A methodological framework is developed, mainly involving the following five stages: image registration, background elimination and image preprocessing, contour fitting and vehicle determination, detector setup and vehicle detection, and vehicle tracking. Various rules of vehicle trajectory, velocity, and acceleration in the MVZ are analyzed in detail. Results show that the running speed of the vehicle offers a slow downward trend in the MVZ, and the entrance speed is higher than the exit speed. Straight vehicles are more inclined to choose Lane III to enter the intersection. The number of track intersections is positively correlated with traffic risk. This research can support traffic management and traffic simulation model calibration.

Infrastructure disruptions such as lane closures present traffic efficiency and road safety challenges, especially on freeways. Such impacts become more significant when combined with compulsory merging sections. This research sheds light on understanding the characteristics of compulsory merging at freeway on-ramps where a construction zone disrupts the mainstream. An integrated simulation platform consisting of virtual reality (VR) and traffic micro-simulation was utilized to collect behavioral data, which features a naturalistic environment for driving behavior analysis. Experiments were conducted, and participants’ perceptions and reactions in various scenarios were analyzed. Results showed that mainstream disruptions on the freeway could contribute to a heavier driving workload at the merging section; interestingly, fewer crashes occurred under adverse weather conditions and nighttime, although the probability of conflicts was higher than fair weather conditions and daytime. This study takes advantage of state-of-the-art technologies, contributing to a positive alternative for driving behavior study.

The relationship between traffic volume and traffic safety has been one of the key issues in traffic safety research. This paper tries to re-examine this relationship by looking at 3-arm and 4-arm intersections in Hong Kong. For each intersection, the sum, product, and variance of traffic volumes from each arm are calculated. Three sets of variables are extracted considering all arms, major two arms, and two arms with the most and the least traffic volume. A linear model and a negative-binomial model are built to model the relationship between crash count and indicators. The results show that the negative binomial is better than the linear or the classic Poisson model. The best indicators for both 3-arm and 4-arm intersections are the sum and variance of traffic volume in all arms. Our results could be insightful to select traffic volume-related indicators when building crash count models at intersections in the future.

The automotive industry is experiencing a paradigm shift towards automated and connected vehicles. With the increasing usage and complexity of electrical and electronic systems, this study introduces a new security risk assessment. In detail, this study first designs a tailored risk assessment framework, which begins with a threat analysis to identify the assets, and threats to those assets. The following risk assessment process consists of a threat estimation and an impact estimation. We introduce the security level conceptually similar to the Automotive standard safety integrity level (ASIL) in ISO26262. The strong alignment with existing standards and processes should make this framework well-suited for the automotive industry’s needs.

Based on a comprehensive traffic congestion analysis database, this paper conducts a causative analysis for urban expressway congestion. Firstly, based on the statistical analysis method of nonparametric test, the influence of each variable on traffic congestion is analyzed under the factors such as traffic control devices, road conditions, aggressive driving behavior, and environment. After that, based on structural equation modeling, the overall influence differences of various factors are derived, the important variables of various factors are inferred, the comparison of safety risk and traffic congestion influencing factors is initially derived, and the solution measures of traffic congestion on urban expressways are proposed accordingly.

In China, the problem of an aging population is becoming increasingly serious. Meanwhile, the age limit for obtaining a driving license has been abolished. The traffic safety problem of elderly drivers has become increasingly prominent. To study the regularity of traffic accidents in the elderly, nationwide traffic accident data of elderly driver deaths from 2011 to 2019 were analyzed by using a grey forecast model. At the same time, national accident data from 2014 to 2019 were used to compare accident types among elderly drivers and drivers of other age groups. The findings were as follows: (1) the number of elderly driver fatalities in recent years across China showed a trend of increasing; and (2) elderly drivers were more prone to vehicle-to-vehicle accidents and single-vehicle accidents. The results are helpful in identifying characteristics of elderly drivers in traffic accidents and providing a theoretical basis for improving traffic safety management of elderly drivers.

In the absence of signal controls for pedestrians crossing the street, there are conflicts between vehicles and pedestrians, seriously influencing pedestrian and motor vehicles traffic safety and efficiency. To master the regular movement of pedestrian-vehicle interaction when pedestrians cross the street and provide theoretical support for the implementation of the traffic safety measures, we investigated appropriate road crossing locations and constructed a pedestrian crossing decision model. Secondly, we constructed a simulation model of pedestrian-motor vehicle interaction on signal-free controlled road sections by combining the improved NaSch vehicle meta-automata model and the social force pedestrian meta-automata model. We described the evolution rules of the model. On the basis of the simulation model, the simulation data of pedestrians are compared and analyzed with the measured data to verify the reliability and rationality of the model.

With the advantage of large volume capacity, heavy goods vehicles have become an important part of road freight transport. However, the proportion of heavy goods vehicles involved in traffic accidents is very high, especially in traffic accidents with serious casualties. These traffic accidents have brought great harm to lives and property, so it is necessary to study the factors influencing the traffic accident severity of heavy goods vehicles and corresponding countermeasures. Based on traffic accident data of heavy goods vehicles in Shenzhen, the statistical analysis of heavy goods vehicle traffic accidents had been examined from the aspects of driver, vehicle, road, and environment. The characteristics of heavy goods vehicle traffic accidents have also been identified, and the influence factors of heavy goods vehicle traffic accident severity have been singled out. With the number of fatal accidents in traffic accident data as a reference, the main factors strongly related to the heavy goods vehicle traffic accident severity were determined through Grey correlation analysis. We then built the data set and the test set based on the selected main factors. We established a Bayesian network model to analyze the relationship between the heavy goods vehicle traffic accident severity and influence factors. The structure of Bayesian network was constructed. The parameter estimation of Bayesian network was conducted by EM algorithm, and the validity of the model has been verified. Then the reasoning analysis was carried out by applying this model. Based on the results of Bayesian network reasoning analysis, both the hidden dangers of heavy goods vehicle traffic accident and problems in the supervision and management of heavy goods vehicles were analyzed, and we proposed countermeasures of strengthening supervision and management and improving road hardware conditions to improve the traffic safety level of heavy goods vehicles.

To effectively shorten the taxi trajectory sequence and remove redundant trajectory point of interest information, in practical applications, a taxi travel network is often constructed based on complex network modeling theory to explore the user’s daily travel rules and characteristics. This paper takes the taxi GPS trajectory data in Beijing as the research object, constructs a spatially embedded complex network that takes into account temporal factors, and analyzes the small-world attributes of it. Finally, different from previous methods of directly clustering trajectory data, this paper conducts a Louvain network clustering algorithm to divide the constructed travel network. Corresponding with the distribution of point of interests of interest in different communities, the division result taking into account the temporal factor will be used as another important data source for analyzing spatial-temporal characteristics of urban travel. This method has good usability in different cities and regions.

This paper studies the influence of motor vehicle emergency avoidance behavior on traffic flow based on the improved NaSch model, focusing on the problem of human-vehicle conflicts caused by the implementation of compulsory yielding measures in a bad right-of-way environment. We use George2.1 software to extract running data of motor vehicles under the pedestrian crossing without traffic lights in urban areas and analyze the three movement conditions of motor vehicles: directly passing, slowing down and avoiding, and emergency avoiding. Based on the MATLAB software, an improved NaSch model was constructed to simulate the emergency avoidance behavior of motor vehicles on a single lane, and the impact of the emergency avoidance behavior of motor vehicles on traffic flow under the variables of pedestrian crossing probability, maximum speed of motor vehicles, and motor vehicle density was studied. The results show that the emergency avoidance behavior of motor vehicles has a greater impact on the safety of passengers and pedestrians and the order of road traffic. The increase in the probability of pedestrians crossing the street and the limitation of the maximum speed of the road will both have an effect on the traffic flow.

In urban areas, to enter/exit a parking lot often requires a vehicle to travel across the non-motorized lane at the roadside, which affects non-motorized riding. This study proposes a method to evaluate the impact degree of motor vehicle on non-motorized vehicle at access connected to urban road segment. From the video shot at the access, information of 523 non-motor vehicle avoidance scenarios is extracted, including the trajectory and behavior characteristics of non-motor vehicle, conflict vehicles and other traffic flows around. The researchers scored the degree of influence on non-motor vehicle riding conditions in 41 non-motor vehicle hedging scenarios. Furthermore, based on the support vector machine model, the correlation model between the influence degree of non-motor vehicle and independent variables such as riding behavior and conflicting vehicle behavior was established.

Exploring the tolerance threshold of commuting time (TTCT) for car users helps us to better understand their commuting travel behavior. Using the travel survey data in Kunming, the influence mechanism of the TTCT for car users is analyzed based on the structural equation model. The result indicates that the instrumental attitudes and car use frequency have a significant positive effect on the TTCT. The affective attitudes have a significant negative effect on the TTCT and the symbolic attitudes have a significant positive direct effect on TTCT. Socio-demographic attributes, such as gender, education level and driving experience, also have a significant effect on the TTCT for car users.

As a kind of infrastructure with the functions of commuting, recreation, and ecology, urban greenway plays an important role in connecting work, commercial, and residential areas. In this paper, the comprehensive slow lane in the urban greenway system is taken as the research object, and the traffic conflict model is established under the mixed traffic state of walking and riding. This paper selects Yuanbo greenway of Fengtai District in Beijing as the investigation site, collects data on the spot, and analyzes the data by SPSS. The results show that the two-way pedestrian flow rate, the two-way bicycle flow rate, the effective width of the road, the average speed, and other parameters have an impact on the traffic conflict.

Lane-changing is a common phenomenon in the driving process. It has a key impact on traffic flow and traffic safety due to its complex interaction with surrounding vehicles. Thus, this study aims to explore the influence of abnormal lane-changing behaviors on traffic oscillation by analyzing vehicular trajectory data. The video captured by Unmanned Aerial Vehicle (UAV) is used to extract vehicular trajectory data by deep learning target detection and tracking algorithm. In the whole switching-process from free flow to congestion, lane-changing behaviors increased significantly. Among the 621 vehicles observed, 448 vehicles had lane-changing behaviors, 69 vehicles were greater than one lane-change, in which most are abnormal. The study found that abnormal lane-changing behaviors worsened traffic oscillation.

The brake hub temperature of the truck at the toe of the slope, and the distances of the truck to the top of the slope at a brake hub temperature of 200°C and 260°C were studied separately using the brake hub temperature rising model proposed by PIARC. We used two profile design schemes of long and longitudinal downgrade section in expressways: a gradient slope scheme and a unique slope scheme. Also, the heating rate of brake hub, the correlation between the temperature rising and the truck weight, and the temperature rising characteristics at different velocities were analyzed under the two different schemes. Results obtained are as follows: the brake hub temperature of the truck on the slope with a unique slope is lower than that with gradient slopes, and the heating rate is linearly dependent on the travel distance; the temperature difference of the brake hub temperatures at two different downgrade sections decreases with increasing truck weight at a same velocity; and the temperature difference under two different sections changes little with the variation of velocity at a same truck weight. Both the distances of the truck to the top of the slope at a brake hub temperature of 200°C and 260°C change with different slope length, and two threshold lengths are obtained for the two temperatures: 15km and 20km. The distance of the truck to the top of the slope at unique slope scheme is greater than that at gradient slope scheme when the slope length is less than threshold length. When the slope length is greater than the threshold length, the gradient slope scheme is superior to the unique slope scheme.

There are a large number of highway tunnels in China. As an important factor affecting road traffic safety and driver comfort, tunnel lighting still has many problems. Based on a questionnaire survey for drivers and passengers, tunnel lighting designers and tunnel maintenance personnel, this paper analyzes the problems existing in tunnel lighting in China and puts forward corresponding improvement strategies and suggestions.

In order to study the influences of trajectory characteristics of vehicles at night in freeway expansion construction areas, day and night observation experiments are conducted to obtain the trajectory offset characteristics based on natural traffic flow. The trajectory characteristics are studied from three aspects: running speed, near road brightness and visual recognition of delineators. It is found that the vehicle trajectory is shifted to the right side and the change in trajectory offset of the passing lane is significantly higher than the traffic lane. It was found that the visual recognition of the delineators affected the vehicle trajectory, the smaller the visual recognition brightness ratio (right side/left side), the larger the lateral shift change of the vehicle. Based on the above research, this study can provide some reference for the improvement and enhancement of safety in the nighttime freeway expansion construction area.

Currently, BIM technology is widely used in building construction, but its application in bridge engineering, especially in swivel construction of continuous beam bridge, is still in its infancy. This paper takes swivel construction of super major bridge for Tongliao passenger dedicated railway crossing 304 National Road and Dazheng railway as an example. BIM core software Revit and Navisworks Manage is used to create a parametric model of the bridge by data file driving method. Research was then carried out from the aspects of 4D construction simulation, collision checking, auxiliary measurement, 3D technical disclosure, calculation of engineering quantity, and carbon emissions. The results show that BIM technology can visually display the bridge swivel construction stage and provide an intuitive and accurate reference for surveying personnel. This study provides a reference for similar projects by demonstrating the advantages of BIM technology, which can effectively speed up construction, save costs, and ensure quality.

Resilience plays a crucial role in management of large-scale road networks. This research studies the relationship between resilience and road changes under the influence of incidents and explores the resilient changing of road sections. The road incident is matched through the incident log data based on vehicle speeds on the road segments provided by the Amap data platform. The remaining resilience (RR) of the road is calculated according to the definition of resilience of road sections. Besides, the topological model of incident influence spread is proposed to analyze the spread influence of incidents occurring on roads. A case study with the accident data on roads near Beijing Olympic Park from October to December 2019 is analyzed to determine the relationship of resilience between different road sections. The results show a certain linear relationship in the remaining resilience between the affected road sections after an incident.

In order to evaluate the operating conditions of vehicles at urban underground road entrances and exits and investigate the influencing factors of their poor operating conditions, a three-stage DEA model of vehicle operating conditions evaluation method was proposed. The research results show that after stripping the environmental factors and random influence, the operation efficiency of urban underground road entrances and exits generally improves, and the environmental factors have a significant influence on the operation efficiency of vehicles. Different environmental factors have different significant degrees of influence on each urban underground road entrance and exit. In addition, the environmental factors that influence the operation condition of entrance and exits are different. The entrance should focus on controlling the distance between the hole and the confluence nose end, the length of the interweaving area, the lane, and the slope, while the safety factors at the exit should be taken into account.

With the enhancement of people’s awareness of environmental protection and the improvement of people’s requirements for driving comfort, the new types of noise reduction expansion joint for highway bridges came into being. In addition to noise reduction performance, deformation and fatigue performance are also the main performance indexes. Sufficient deformation and fatigue resistance performance can prevent the expansion joint from damage during its working life to ensure product quality. The deformation and fatigue performance of the noise reduction expansion joint were experimentally studied in combination with relevant test specifications at home and abroad, and then the rules were summarized. It is supposed to provide reference for academic researches on noise reduction expansion joints.

There is interweaving activity that takes place on the rung ramps of alfalfa leaf interchanges. The existence of these weaving sections will greatly affect the traffic capacity of the interchange. Therefore, it is necessary to set up collecting and distributing roads to increase traffic capacity and improve driving efficiency. In order to study the length of collecting and distributing lanes, this paper analyzes the factors such as marking, recognition sight distance, and the interleaving length, determining the applicable range of the length of the collector road. Finally, using VISSIM software and taking vehicle delay as the evaluation index, we verify the rationality of the conclusion by comparing vehicle delay under different simulation lengths.

A path planning method of the car to return to the original lane needs to be established when the self-driving car fails to change lanes in an autonomous driving environment. The study for vehicles to return to the original lane when the vehicle fails to change lanes in the unmanned environment is necessary. In the study, we used the quintic polynomial path planning method to establish a path optimization model of returning to the original lane based on the requirements of vehicle kinematics. Considering the safety and efficiency of lane-changing return, we set up a comprehensive benefit function to evaluate the model. Simulations were used to test the models and compared paths. The results indicate that the proposed lane changing return model has high accuracy and can guarantee the safety of the return path to actual driving.

Roads deteriorate, hence periodic maintenance and rehabilitation (M&R) is necessary to improve the road condition. Different M&R treatments result in different outcomes, and decision-making should design appropriate M&R strategies to the segments of road networks. Hence Multi-Objective Optimization (MOO) is increasingly applied. To better assist with M&R decision-making, a set of sufficient and well distributed Pareto solutions are needed, while this can be different without an effective algorithm. This paper developed a MOO method that optimizes the multi-objective M&R decision making of road networks and generates both supported and non-supported solutions for M&R decisions. It also introduces different algorithms for comparison purposes. A case based on practical data was studied. According to the result, the developed method can effectively handle multiple objectives and constraints, generate a set of sufficient and well distributed solutions including, and therefore assist with the M&R decision-making of road networks.

Parking difficulty in urban cities, especially in commercial areas, has attracted wide attention. With the help of remote parking of automated vehicles (AVs), people do not waste time cruising for parking, and the utilization of distant parking facilities becomes possible. This study presents a bi-level model to determine the optimal planning of parking facilities under coexistence of AVs and human-driving vehicles (HVs), where the locations and capacities of parking lots is obtained in upper and lower-level models to determine the traffic assignments. Simulated annealing (SA) is adopted in upper level for large-scale problems. Numerical experiments are conducted under different penetration rates of AVs and construction budgets. It is indicated that the total commuting costs can be reduced by 37% without increasing construction cost when there are 50% AVs. Meanwhile, the increase in total commuting costs caused by inadequate budgets can be well controlled by use of AVs.

Intersections play an important role in the road network. The traffic organization of the intersections may affect the safety and efficiency of the entire road network. This paper takes the intersection of Tanshan Road-Longmai Road in Zhenjiang City as an example to study the optimization design of intersection traffic organization. First, traffic survey is conducted to analyze the current situation of the chosen intersection. The traffic organization issues of the intersection are summarized based on the traffic survey results. Second, by combing several specific methods such as intersection channelization, a complete optimization scheme is proposed to improve the traffic service of the intersection. Finally, VISSIM is used to simulate the traffic flow of the intersection before and after the optimization. The simulation results show that the optimization scheme reduces the average vehicle delay and maximum queue length. Besides, the delay time of the optimized intersection is shortened by 7 s.

Bus detector location is significant for transit signal priority (TSP) systems. A location either too close or too far will reduce the effectiveness of TSP due to failure of triggering TSP or erroneous prediction of bus arrival time. To address this issue, this paper proposes a method to optimize the detector locations of multi-step detection system. The forecast detector location is optimized by maximizing bus passing rate considering different arrival conditions, and confirmation detector location is optimized by reducing the factors that lead to bus speed fluctuation. To validate the proposed method, VISSIM experiments were done. Results show that the optimal locations of forecast and confirmation detector are around 450 m and 100 m from the intersection. This method yields values of 460 m and 92 m, which matches the simulation results. This method could guarantee both timely trigger of TSP and accurate prediction of bus arrival time, thus improving service effectiveness of TSP.

The performance of asphalt pavement shows different forms of decay with increased service life of expressways in China, which affect driving safety and comfort. Actively taking preventive maintenance treatments can further improve the pavement performance. However, there are relatively few studies on the improvement effectiveness of various preventive maintenance treatments in China. The data foundation of this paper is the pavement performance, maintenance history, and other multivariate data of expressways in Beijing. This paper analyzes the performance distribution characteristics of Beijing expressway pavement from a macro perspective and uses mathematical evaluation method to evaluate the effectiveness of three preventive maintenance treatments (thin overlay, micro surface, and fog seal) on the PCI performance index. The results show that the improvement effectiveness of treatment is the best, followed by micro surface. However, the fog seal treatment is poor; it can only delay the decay rate of PCI. We recommend using the thin overlay treatment when PCI is at a good level or below (PCI < 92). When PCI is at an excellent level (PCI ≥ 92), we recommend the micro-surface treatment. This study will be important for pavement maintenance management and provides a reference for decision-making and implementation of preventive maintenance treatments.

The urban traffic situation has become increasingly complex in recent years, and the current performance evaluation code for urban road asphalt pavement is no longer applicable to the actual situation of roads, which led to a lack of comprehensive analysis to the proposed schemes of pavement maintenance. Firstly, this paper combines the attribute hierarchical model and the entropy weight method. Secondly, the pavement structure strength index and skidding resistance index are incorporated into the comprehensive evaluation to reconstruct the weights of each evaluation index. Thus, a new evaluation model of asphalt pavement is established. As the example shows, the current evaluation model is hard to correctly evaluate the road sections with the structural strength of the subgrade and does not meet the standard. The new evaluation model is effective and valuable for evaluating the current performance of urban road asphalt pavement.

Because of the great mechanical properties and advantages of environmental protection, basalt fiber has gradually become a common fiber modifier in asphalt mixture. Due to the difference in reinforcing mechanism, the reinforcing effect of basalt fiber and lignin fiber on asphalt mixture is different. In this study, the performance of the SMA-13L mixture with basalt fiber, lignin fiber, and composite fiber was explored through experimental methods. The rutting test, the immersion Marshall test, the freeze-thaw indirect tensile test, the bending test, and the sand patch test were carried out. Compared with the mixture with more commonly used lignin fiber, the mixture with basalt fiber has better crack resistance and skid-resistant property but worse high-temperature stability and moisture susceptibility. The composite fiber composed of lignin fiber and basalt fiber can integrate the advantages of the two, reflecting a more comprehensive performance.

Ramp metering (RM) is to deploy traffic signals on the ramp to dynamically control the ramp vehicles into the mainline of the expressway, which can smooth the mainline traffic flow and utilize the existing expressway capacity effectively. According to the control range, RM can be divided into local RM and multi-ramp RM. Local RM can solve the local congestion problem to a certain extent, but it does not consider the regional or system problems, which may cause congestion shifting and lead to congestion on other roads in the region. ALINEA algorithm is a typical local RM algorithm, but ALINEA algorithm has no prediction mechanism and thus cannot handle the sudden change of traffic flow smoothly. Besides, few studies to develop multi-ramp RM algorithms based on congestion risk prediction. Therefore, this study improved the ALINEA algorithm and established a system-optimal dynamic RM algorithm based on real-time congestion risk prediction.

Maintenance period is a significant indicator influencing decision-making on highway maintenance. The study of highway maintenance period at present is almost experience-based or focus more on theoretical analysis both in China and abroad, which lacks high accuracy and practical guidance for making decisions on highway maintenance. Based on large volumes of historical data of several typical provinces in five climate regions of China, this research analyzes the corrective maintenance period of asphalt pavement and its contributory factors. An entire national distribution spectrum of corrective maintenance period of asphalt pavement is established to provide meaningful guidance for highway maintenance decision-making.

The method of molecular dynamics simulation was used to study the adsorption mechanism of corn stalk fiber on asphalt and the factors that affect its absorption ability on asphalt in this study. By establishing molecular models of four different asphalt and corn stalk fibers, setting the molecular force field and calculation parameters, constructing a molecular dynamics model of interface, the absorption regularity of corn stalk fiber on different components of asphalt was obtained. The research results show that temperature and content of each component of asphalt are important factors that affect the absorption ability of corn stalk fiber on asphalt, and corn stalk fiber can increase the viscosity of asphalt by adsorbing light components of asphalt.

It is necessary to carry out maintenance work to ensure the quality and service level of the expressway, and a reasonable quota is the basis for efficient maintenance work. At present, due to different maintenance units and expressways, the labor, machinery, and material consumption of the same sub-purpose are different, which complicates the compilation of maintenance quotas. This paper takes “replacement of 140 columns” in the routine maintenance of Beijing expressway as an example. The entropy-TOPSIS method was used to comprehensively evaluate the consumption of labor, materials, and machinery under different maintenance conditions, and the pros and cons of each maintenance condition were judged by sorting circumstances to formulate reference standards for maintenance quotas. The results of this paper will be significant for the compilation of expressway routine maintenance quotas.

Noise control has become one of the major concerns for modern urban viaducts. The noise sources of highway bridge expansion joints are briefly introduced in this paper, and the commonly used expansion joints are discussed and reviewed systematically from the aspects of noise generation mechanism, noise influencing factors, noise reduction effect and application status. The suitability and feasibility of these noise reduction measures of expansion joints are discussed, which could provide certain references for the type selection of expansion joints in practical engineering. Based on the domestic and foreign research data, the noise reduction problems needed to be solved in the future are also discussed. The paperprovides new perspectives and basic data for academic researches on noise reduction expansion joints.

Heavy-duty vehicles will produce greater dynamic load, likely leading to premature pavement damage. Ensuring driving safety and road friendliness of heavy-duty vehicles is an important issue to be considered. In this paper, the three-dimensional (3D) model of pavement roughness was constructed based on fractal theory and fractal interpolation. According to the 3D contact characteristics of tire-pavement, the improved re-revised 3D flexible roller contact tire model was established. The 3D three-axle heavy-duty vehicle vibration model with 9 degrees of freedom (DOF) was re-derived and the driving safety was analyzed. The results show that the driving safety of grade A roughness pavement has always met the requirements. For B-level roughness pavement, the speed should be controlled at 100 km/h. For grade C uneven pavement, the speed should not exceed 80 km/h. The research results will provide a theoretical basis for the safe driving of heavy vehicles.

The construction of the foundation pit has a negative impact on the existing shield tunnel below, and if not properly controlled, it could lead to construction risks and affect operational safety. In this paper, the construction of a viaduct pit over an existing metro tunnel is used as an example. We analyzed the soil and tunnel deformation under various construction conditions by numerical method and quantified the effect of strengthening measure. The results show that the shield tunnel produced a large vertical deformation and a small horizontal displacement during the construction, while grouting can reduce the ground uplift by about 30% and the deformation of the existing tunnel by 50%.

The assembled underground stations of Changchun Railway Line 2 use a series of closed cavity thin-walled components to reduce the structural weight. As for the hollow component, the shear lag effect is essential for the safety of the structure. Thus, the numerical simulation is used to analyze the normal stress of the component, which contains a simple straight model and an actual station model. The results show that the shear lag effect occurs in the intersection of flange and rib of the component, and the maximum amplification factor of stress is 1.02 which should be considered for relevant design.

A new kind of closed cavity thin-walled component was developed to reduce the weight of the precast assembly underground station in Changchun Metro Line 2. The mechanical properties and force transmission paths of this component are still lacking in research, so this study sets up models for the representative straight component and the actual curved component with the finite element method. The distribution of shear stress is analyzed and the shearing sharing ratio of web and flange is obtained. Based on the results, a formula to calculate the shear capacity for this component is proposed and the improving coefficient of the flange to shear strength is about 1.2 for all types of components.

When pouring concrete in large stations, the temperature change caused by the hydration heat leads to cracks in the structure and the installation of the induced joint is an effective way to solve this problem. This study presents a 3D numerical simulation of the concrete pouring process in a subway station based on a real project and analyses the effect of temperature stress on internal force and deformation. Based on this, the effect of the induced joint on the station is quantified. The results show that the induced joint can reduce the tensile stress by sacrificing the deformation resistance of the structure. In addition, complementary measures that can control the development of cracks are proposed for similar projects.

The construction of the foundation pit that is close to a maglev line will lead to the deformation of the maglev line, thus affecting safe operation. Based on the foundation pit construction of the underground station of Shanghai airport connecting line and the fire resistance experimental building of Shanghai Tongji University, this paper studies the influence of foundation pit construction on the deformation of adjacent maglev structure and its deformation development law. Through the arrangement of a variety of deformation monitoring points and the statistical analysis of a variety of monitoring data, this paper reveals the correlation between maglev structure deformation and foundation deformation which includes the deformation of underground continuous wall and deep soil deformation and so on, the key factors and methods that can effectively control maglev structure deformation are put forward. The research results have certain reference significance for follow-up similar projects.

The establishment of curb parking spaces can effectively relieve the tight situation of urban parking supply and demand. However, road resources are always limited. As parking spaces occupy road resources and squeeze road width, the capacity of road sections will be reduced. It is of theoretical significance and practical value to study the influence of curb parking on traffic capacity for alleviating traffic congestion and improving road traffic safety. This paper will analyze the reasons for the impact of curb parking behavior on road capacity and the influence mechanism of various influencing factors of curb parking on traffic capacity. By using the method of traffic conflict technology, the influence of curb parking behavior on traffic capacity will be studied and a model of the influence of curb parking on road capacity will be constructed.

In view of the interference of social groups to the pedestrian flow in large-scale activities, the micro-simulation model including grouping behavior was formulated, and the simulation platform of pedestrians in hall corridor was built based on the model. With the platform, the pedestrian flow behavior in corridors with different conditions was simulated. In the simulation, the corridor is divided into individual travelling area and social group travelling area by guardrail to explore the influence of different guardrail settings on the evacuation efficiency. The results show that the evacuation efficiency can be influenced by the position and length of the guardrail. With the increase of the corridor width, the number of evacuees in unit time tends to increase. The research can provide theoretical support for managers to optimize the diversion facilities in corridors in large-scale activities.

For the first time in China, precast assembly technology was used to build the large underground stations and a new kind of closed cavity thin-walled component was developed to reduce the weight of the structure in Changchun Metro Line 2. This paper set up a numerical model for the typical straight closed cavity thin-walled component and conducted a parameter study to obtain the influence of the end plate on stress transfer of the component. The result shows that the best performance is achieved when the ratio of the end plate thickness to height is 0.43 and over-thick end plate does not improve the force performance significantly.

This paper presents a study on basalt fiber reinforced vinyl ester resin composites exposed to extreme thermal cycling conditions. The mechanical properties of basalt fiber surface mat, needle felt, and unidirectional fabric reinforced vinyl ester resin composites were evaluated, which were thermally cycled between the temperatures of 150ºC and −196ºC in a gaseous N₂ atmosphere for 6 and 12 cycles, respectively. No significant changes of the tensile strength and failure modes of BF/VE composites were observed under extreme thermal cycling conditions. The morphology of the composites after thermal cycling was characterized by SEM and FTIR, which indicated that the interior structure of basalt fiber reinforced vinyl ester resin composites was hardly changed after extreme thermal cycles. Therefore, it is reasonable to conclude that the composite has satisfied mechanical property to meet applicability requirement under extreme thermal cycling conditions.

To enhance the parking efficiency and user satisfication, the strategies of parking duration limit and parking pricing are explored in this paper. Based on the inherent characteristics of parking spaces and the stability of parking duration distribution, the parking space are divided into several areas, with each area set up with specific parking duration limit and parking prices. A bilevel programming model is proposed to solve the problem of how to minimize the average parking time (crusing time and walking time) and the personal cost (parking time and parking prices). A Chaos Genetic Algorithm is utilized to achieve the appropriate results. To evaluate the availability of the methodology, a case study is presented. The results show that compared with the plan of a single area, the average parking time of the three-area zoning plan decreased by 7 s and the parking revenue increased by $125.50 in a day.

In this paper, we present an optimization model to address the allocation problem for a one-way electric carsharing system. The proposed model is a linear mixed-integer program with the objective of maximizing the total profit, which involves all incomes and costs in an electric carsharing system. In particular, the profit is formulated as the mathematical expectation of a discrete random variable, which allows for a comprehensive consideration of all possible future demands. A linear solution method is applied to obtain the global optimal. At the end, the model is validated using real data from 20 electric carsharing stations. The results indicate the daily improvement of profit could be as high as 21.5% with an average of 15.72%.

Various disruptions such as natural disasters, malevolent attacks, and emergencies are still inevitably posing threats on road networks in the driverless environment. However, most studies have focused on the impacts of micro mixed traffic flow characteristics, which cannot explicitly measure road network vulnerability in driverless environment. This paper develops a bi-objective framework to construct driverless environment based on travel demand. An improved accessibility-based methodology integrating traffic flow and driverless conditions evaluates the vulnerability of road network under disruptions. The proposed methodology was applied to the highway network in Florida, USA. Results demonstrate that the vulnerability of road network would be decreased with the increase of the penetration rate of autonomous vehicles. When the penetration rate gets in the range of 40%–80%, network vulnerability would significantly reduce, and they show linear relationship. Findings of this study could provide new sights into managing and planning driverless road network in future.

The city scale expanding leads to the imbalance of job-housing and increases commuting distance. Based on the analysis of urban building census data, cellular signaling data and Gaode POI data in Kunming, it is found that the area of the second ring road is a concentrated area of population and employment. Although rail transit plays an important role in the whole urban traffic network, it accounts for a low proportion of commuting. Further, the degree of job-housing balance index and the influencing factors system of job-housing balance is established. According to the result of regression model, the land mixing, the area of public service facilities, residential facilities and school facilities, and public transport accessibility significantly influence job-housing balance. Then, the paper suggests the layout optimization of construction land, and the depth of the fusion between urban land and rail transit in order to improve the level of job-housing balance.

The practical experience of traffic development indicates that there are some problems in increasing network traffic capacity by traditional means ignoring travel behavior. In order to increase the current capacity, this paper focuses on studying the relationship between network traffic capacity and travel behavior choice and explores a way to select individuals as objectives to study the network traffic capacity. Based on the space-time consumption method, the relationships between travel speed, travel distance, travel mode, and traffic capacity are discussed, respectively; the specific influence trends of these three aspects are obtained. This paper constructs a simple traffic capacity model considering mode choice under a certain modal ratio, and the internal binary logit model is calibrated by revealed preference (RP) survey in peak hour in Tai’an. This paper can provide reference for the calculation and modeling of traffic capacity, as well as suggestions for relevant departments in traffic demand management.

There is a close interaction between the urban built environment and the spatial behavior characteristics of individuals in time. Youths are an important group in urban development, and their travel affects the overall efficiency of the city. This study uses cell phone signaling data and POI data to build linear regression models and geographically weighted regression models, respectively, to analyze the impact of built-up environment on the peak hour travel of youth groups. It is found that there are similarities in the morning and evening peak models, and there are significant differences in travel time and space; residential population and job density have significant effects on the travel of the youth population, and commuting travel is still the most important which exists a separation of jobs and residences; construction of new cities in Shanghai is steadily advancing, and accessibility and equity of urban services and facilities need to be strengthened.

Behind promotion and application of battery-electric buses in China, many cities in cold areas add fuel boilers for battery-electric buses to warm the cabin in winter, which has reduced the effect of energy-saving and emission reduction. This paper aims to understand installing oil-fired boilers and comparing the two schemes, with and without fuel boilers. Sixty-four public transportation companies in Northeast China are investigated, and the results show 92% of them have added fuel boilers in winter. The main reasons are the range of electric buses cannot support the operation for a day, and the cabin temperature is low. By comparing the actual usage with fuel boilers (Yichun, Heilongjiang) and without (Fushun, Liaoning), it is found that the average daily mileage and charging times. Furthermore, when the electricity price is low, no fuel boilers are more advantageous, indicating that more bus companies could do without fuel boilers through preferential electricity price policies.

Rail transit passenger flow and its distribution characteristics are markedly determined by multiple influencing factors such as the land use characteristics and the built environment. Most known studies in the relationship between traffic flow and land use tend to involve simple linear assumptions. This study introduced the real data set and analyzed the nonlinear relationship between rail transit passenger flow and its influencing factors. A numerical analysis was conducted in passenger flow collected in the rail transit stations of Shanghai. Firstly, network crawler technology was used to obtain the points of interest (POIs) for presenting land-use characteristics around subject sites. Subsequently, the gradient boosting decision tree (GBDT) model was established and extracts influencing factors and analyzes the distribution characteristics of rail transit passenger flow. The findings can provide some targeted suggestions for traffic planning and guide reasonable land use to relieve the traffic congestion caused by the over-concentrated traffic demand.