Journal of Transportation Engineering, Part A: Systems

Flooding can severely disrupt transportation systems. In many cases, flooding only results in road closures rather than neighborhood evacuation. This paper introduces a framework that provides routing assistance to vehicles exposed to flooding by identifying them based on origins, destinations, anticipated paths, and departure times. Warning messages are disseminated to vehicles not directly impacted by the flood. The framework leverages vehicle connectivity that allows the enhanced exchange of information between equipped vehicles and a traffic management center. The proposed framework is evaluated on two transportation networks based on sections of Virginia Beach, Virginia. The evaluations of the scalability to different network sizes and the sensitivity to various flood characteristics, policy-related variables, and other dependencies are performed using simulated vehicle data and hypothetical flood scenarios. The computation times depend on the network size and flood depth but have an average of 1.47 s for the most widely tested network and deepest tested flood. The framework has the potential to alleviate the impacts and inconveniences associated with nuisance flooding.

This paper analyzes the distribution characteristics and influencing factors of traffic accidents in the multiethnic-gathering plateau mountain area. Using data from an investigation of 56,651 accident records in Guiyang, China, in 2015, we analyze the spatial and temporal distribution characteristics of accidents, and build a Gray Markov model to predict regional traffic accidents. Accident types of rear-end collisions and yielding not according to regulations are the most common ones in Guiyang, and the majority of these types of accidents occur along main roads and near commercial areas. Meanwhile, logistic regression analyses are applied to determine the correlations between road type, traffic flow type, and traffic accidents. It shows that road type and traffic flow type have no significant interactive impacts on accidents, while road type has a significant impact on the number of accidents. The accident distribution is more intensive in the northern part of the city where many ethnic groups gather together than others. Therefore, it is recommended that the traffic management department consider ethnic differences when improving traffic safety in this area.

Probe vehicle data is changing the landscape of transportation engineering. The availability of vehicle trajectory data, or GPS waypoint data, has expanded the utility of probe data. However, the low penetration rate of vehicles prevents signal-performance assessment during short-term or low-volume periods, such as special events, seasonal traffic patterns, and overnight timing plans. Current research has used high-ping frequency data or the temporal distributions of waypoints of less than 2 s. This paper evaluates different approaches for using low-ping frequency data to measure delays at signalized intersections. The results of statistical testing show that 30- and 60-s ping data provide delay values that are not significantly different from 1-s ping data. These sampling frequencies increase the number of observable trajectories by 700%. This data allows for scalable approaches to immediately measure delays at signalized intersections nationwide in the US, thereby reducing costly infrastructure needed for signalized performance measures.

Unlike personal cars for daily commuting, freight vehicles demonstrate vastly different traveling behaviors with longer spatial-temporal activity that is composed of multiday trip chains. Quantitatively identifying and describing the trip chains of freight vehicles could help in understanding typical freight behaviors and, thereby, provide a new perspective for analyzing freight systems. Therefore, based on the large-scale and prolonged vehicle trajectory datasets from global positioning system (GPS) equipment, a multiscale depot-identified method based on the density-based spatial clustering of applications with noise (DBSCAN) algorithm is proposed. The base depots and trip ends, which are critical components for multiday freight trip chains, are acquired to construct the complete multiday trip chains. Additionally, a new structure with multifeatures for synthetically depicting multiday trip chains is proposed. Finally, by discriminating the trip chain characteristics, the multiday trip-chaining patterns of freight vehicles are extracted, and their distributions across different vehicle types are analyzed. The results show that some travel patterns are limited to specific vehicle types. For example, the travel pattern in Cluster 3 only occurs for medium-sized ordinary trucks (METs) and tractor vehicles (TRVs). Additionally, the same travel pattern may occur for different vehicle types. The travel patterns of METs and TRVs are the same, but their proportions are different. The discovered patterns could be used in freight demand modeling, freight system simulations, or other customized management for operators.

This study solves the electric transit network design problem (ETNDP) by simultaneously optimizing the layout of bus routes, the service frequency, and the location of charging depots. To ensure the rational design and operational feasibility of an electric transit network, an optimization model of the ETNDP with the constraints of route, depot, operation, and charging is developed in consideration of achieving overall operating cost effectiveness, while guaranteeing adequate operating buses to meet all passenger demands and satisfy the recharging demands of all operating buses without delays or congestion. An improved artificial fish swarm algorithm (AFSA) with the crossover and mutation operators is developed to solve the proposed model. For example, the transit network in an urban region of a city in China is studied in this research. It is confirmed that the optimization model solved by the improved AFSA is able to appropriately provide the optimal solution to the design of a relatively large-scaled electric transit network for its efficient operation.

The categorization analysis of car-following behaviors is beneficial to enrich the current car-following models and the applications of connected and automated vehicles (CAVs) in a mixed traffic environment. Previous studies categorized the car-following behaviors during the traffic oscillations using artificially designed behavior patterns, but they are not quietly flexible and are limited to distinguish the complicated car-following behaviors. To address such a problem, the study proposes a wavelet-based time series clustering approach to automatically categorize the car-following behaviors. First, the response time series of the car-following behaviors are extracted using general Newell’s car-following model. Second, the discrete wavelet transformation algorithm is employed to extract the following behavior features from the original time series. Finally, the hierarchical clustering algorithm is used to categorize the car-following behaviors according to the calculated similarity between the transformed time series. Numerical tests on Next Generation Simulation (NGSIM) show that the proposed algorithm can effectively and automatically categorize the typical car-following behavior patterns summarized in the previous studies. The proposed algorithm is also flexibly implemented to discover the potential car-following behavior patterns. Findings suggest that a wavelet-based time series clustering by combing the Haar wavelet transformation algorithm and hierarchical clustering algorithm is a superior approach to automatically categorize car-following behaviors with a time series trajectory.

Real-time performance measures are important for agencies to maintain their roadways during the winter season. However, infrastructure can be expensive to deploy and maintain and may be sparse in rural areas, while speed data alone may not provide enough fidelity in borderline conditions. This study looks at high-frequency vehicle data from the in-vehicle bus to detect changes in the vehicle and driver behavior during changing winter road conditions. The data is reported to the cloud via cellular communication and is viewable in real-time using a map-based web dashboard. Three winter weather events are assessed using in-vehicle data collected from the February–March 2018 period. Using a two-sample Kolmogorov-Smirnov test, we found high significance in the reductions of applied brake pressures and rates of braking in winter versus fair weather conditions before vehicle intervention is necessary. The paper concludes that pairwise comparison of driver change in brake pressure may be a valuable data source indicative of deteriorating road conditions before more severe indicators such as traction control, antilock brake, and/or hazard indicators are activated.

This study develops optimal variable, lane group–based, speed limits for traffic control at freeway lane drop areas (e.g., work zones). The proposed approach adopts a simulation-optimization framework that utilizes a calibrated and validated macroscopic traffic flow model METANET, along with the microscopic traffic simulation model VISSIM, to develop the optimal speed limits. A multiobjective optimization framework is implemented whereby the model primarily seeks to improve traffic safety by reducing the average number of stops, while taking other objectives, such as the average travel time and throughput, into consideration. For optimization, the heuristic, biologically-inspired optimization technique known as particle swarm optimization (PSO), is utilized, and the $\epsilon$-constraint method is adopted to allow for considering multiple objectives in the optimization process. The proposed traffic control strategy is then evaluated for a hypothetical freeway lane drop area under a real-world congested traffic scenario. The research findings show that the proposed lane group–based control strategy outperforms other link-based, variable speed limits reported in the literature.

In order to promote the accuracy of short-term traffic flow forecasting, an adaptive real-time model consisting of two important stages is proposed. The first stage encloses a novel online sequence extreme learning machine with forgetting factor (FFOS-ELM) that effectively averts the influence of early data on model accuracy induced by the time variability of short-term traffic flow and adaptively corrects the model parameters. In the second stage, based on the optimal estimation on the particle filter system, optimized real-time forecasting of future traffic volume is accomplished by filtering out the noise in the original traffic volume. Finally, the validity and feasibility of the proposed model are verified by a case study. Microwave data from the main road of a city in China was selected to extract the traffic volume as the model data set, and the accuracy of the proposed model is compared with five traditional offline algorithm models and two online algorithm models. Forecasting results indicate that the two-stage adaptive model produces more accurate and stable predictions and shows potential in forecasting the short-term traffic flow under uncontainable conditions.

Cycling is a healthier and greener travel mode that city planners and policymakers have encouraged for short-distance trips. Because cycling provides an efficient way to improve public health and reduce energy consumption, analyzing the contributing factors to bicycle usage on roadway segments is essential to quantify the impact of certain attributes on bicycle volume and to further provide a better cycling environment for cyclists to encourage nonmotorized travel. To gain a better understanding of the attributes that have a significant impact on cycling, this study collects crowdsourced bicycle data from Strava and combines the data with other supporting data, such as road characteristics, demographic information, temporal factors, geometry features, and bike facilities. An ordered probit model is then developed to analyze Strava users’ bicycle usage on each road segment in the city of Charlotte, North Carolina. The results reveal that road segment length, number of through lanes, median household income, total households in a census block, cycling on a suggested bike route, greenway, US route, and one-way road all have a positive impact on Strava user counts on a road segment from 6 a.m. to 6 p.m. Conversely, the variables for cycling on weekdays, total families in a census block, slope, signed bike routes, and suggested bike routes with low comfort have a negative impact on the Strava user counts on a road segment. Based on the modeling results, recommendations are also made to assist in improving the cycling environment and increasing future bicycle volume.

On single-track rail corridors, meets between trains are a significant source of train delay. From the stopping train’s perspective, a meet can be divided into three distinct phases: braking into a siding, waiting for higher-priority trains to pass, and accelerating to operating speed on the main track. Meet delay can be further divided into fixed and variable components depending on the number of trains partaking in the meet. Advanced train control systems incorporating moving blocks and innovative dispatching strategies such as train fleeting promise to reduce minimum meet times. Using a spreadsheet-based calculation, it was found that train fleeting distributes fixed delays among more train conflicts, resulting in more efficient conflict resolution. Complementarily, moving blocks minimize variable delays. Combining moving blocks and fleeting can be highly effective, producing the lowest meet delay across a variety of track speeds and dispatching strategies. The results from this study can help railway practitioners evaluate the benefits of train fleeting, moving blocks, shorter train lengths, and extended fleet-length sidings when developing operating and capital plans.

Optimized and cost-effective railway track repair and maintenance planning requires periodic evaluation of railway track performance for which a track condition index is necessary. In particular, rail internal defects have not been sufficiently considered in the currently used rail quality indices, although they have been frequently reported as the main causes of railway derailments. This study is devoted to develop a new rail condition index which takes into account rail surface/visual defects as well as all reported internal defects for the first time. The index is based on data obtained from visual and automated measurements using an ultrasonic technique. The results obtained indicate that the consideration of rail internal defects in the evaluation of rail conditions has a considerable effect on the accuracy of rail quality predictions. It was shown that the new assessment model can considerably improve track safety by providing more precise repair strategies.

At a signalized intersection with a permitted left-turning phase, left-turning and opposing through vehicles always compete for priority according to actual traffic conditions instead of following strict priority rules. It is a critical cause of safety challenges. To alleviate the problem, this article develops a working model to determine the form of left-lane line extensions. In the proposed model, left-lane line extensions were approximated as a combination of (1) straight lines, and (2) circular curves and straight lines. In addition, they were constrained by the entering point, exiting point, and controlling point. As a result, a left-turning vehicle traveling along left-lane line extensions would be forced to yield the right-of-way to opposing through streams. Additionally, the location of the left-lane stop line was also optimized to meet the demand of left-lane line extensions. On this basis, a recommended application procedure was further introduced to make it easy for traffic engineers to use. Finally, three case studies with different intersection geometries were conducted in Changchun, China, to demonstrate the effectiveness of the proposed model. The proposed models are valuable because they could provide a guideline on designing left-lane line extensions and also help to make intersections safer.

Beijing, the capital city of China, currently experiences serious traffic congestion and extreme air pollution. To mitigate these issues, Beijing Municipality is currently considering implementing a congestion charging system that requires drivers to pay for entering congested areas. A stated preference survey of 5,468 Beijing residents was conducted mid-May 2016 to investigate public attitudes toward the proposal. This research analyzed the survey data using a multinomial logit (MNL) model and a cluster analysis to identify relationships between residents’ characteristics (including the basic socioeconomic and demographic attributes; and their workday travel characteristics) and their attitudes toward Beijing congestion charging. Women are more likely to remain neutral toward congestion pricing. Among employment categories, students are most likely to support, the retired/unemployed are most likely to oppose, and the employed tend to fall between students and the retired/unemployed when it comes to acceptability of congesting pricing. Travelers who use public transit, bicycle, or walk to work are more likely to support congestion charging than car commuters. Income and car ownership have nonlinear relationships with acceptability. High- and low-income commuters are more likely to support congestion charging than middle-income commuters. Residents with no access to cars, with access to one car, or to more than two cars are more likely to support congestion charging than people with access to two cars. There is also a nonlinear relationship between the monetary amount people are willing to pay for the charge and their acceptance of congestion pricing, which merits additional research. The perceived effectiveness of congestion charging strongly influences their acceptance, and so does a traveler’s familiarity with the general congestion charging concept and the specific plan to be implemented in Beijing.

The traffic volume of each movement at signalized intersections can provide valuable information on real-time traffic conditions that enable traffic control systems to dynamically respond to the fluctuated traffic demands. Real-time movement-based traffic volume prediction is challenging due to various nonlinear spatial relationships at different locations/approaches and the complicated underlying temporal dependencies. In this study, a novel deep intersection spatial-temporal network (DISTN) is developed for real-time movement-based traffic volume prediction at signalized intersections, which considers both spatial and temporal features by the convolutional neural network (CNN) and long short-term memory (LSTM), respectively. In addition, the within-day, daily, and weekly periodic trends of traffic volume are also considered in the proposed model. This is the first time that a deep-learning method has been applied for movement-based traffic volume prediction at signalized intersections. In the numerical experiment, the proposed model is evaluated using real-world data and simulation data to demonstrate its effectiveness. The impacts of various structures of traffic networks on the proposed model are also discussed. Results show that the proposed model outperforms some of the state-of-the-art volume prediction methods currently in the literature.

Rail track inspection is essential to urban rail infrastructure management in terms of safety. The track geometry car is one of the important inspection vehicles for track condition measurement. Routing a track geometry car on a large-scale subway network has two objectives: inspection cost and inspection time interval consistency of each line during one planning horizon. An optimization model is developed to route a track geometry car on a large-scale subway network. Because of the model’s nondeterministic polynomial time (NP)-hardness, a recursive hybrid heuristic algorithm using a novel double-layer chromosome genetic algorithm as the global search strategy and a simulated annealing algorithm as the local search strategy are proposed to attain suboptimal inspection schedules. The proposed algorithm is applied to the Beijing subway network for two months, and the 877.260-km track mileage was inspected through one track geometry car. The schedule currently adopted in practice instructs Beijing subway’s track geometry car to complete all prescribed inspections at inconsistent time intervals in 23 days with a dead mileage of 601.180 km. The schedule attained via the proposed algorithm routes the track geometry car to inspect all tracks at consistent time intervals and saves 301.806-km of dead mileage (50.2%) and five days.

A turnout, a device to guide tracks, is critical to the safety of high-speed railways. Detecting gaps in switch machines can monitor a turnout’s working performance. Existing gap-detection systems, however, can barely perform at high accuracy and with a low false alarm rate for a long time due to the complex operating conditions of switch machines. This study proposes an approach combining YOLO-based object detection architecture with image processing algorithms, of which YOLO is a deep learning network for object detection. First, YOLO detects target areas in gap images, and then image-processing algorithms identify gaps and calculate gap sizes. This approach targets various types of switch machines and particularly complicated situations. Experiments on gap images of S700K switch machines demonstrate that the accuracy of detecting gaps is 100%, and the accuracy of calculating gap sizes is higher than 99%. Additionally, the proposed approach can exhibit the same high performance on complex images, like overexposed and tilted ones.

Railway track geometry is generally understood to be influenced by the deformation of the track infrastructure. This study developed a spatial–temporal identification model for the deformation of the underlying high-speed railway infrastructure, including simply supported beams and track slabs based upon track geometry data collected between 2016 and 2019. To achieve this, we first preprocessed the data, including data collection and cleaning. Next, we developed a track irregularity degradation indicator (TIDI) for different track infrastructures using wavelet coefficients. Then, we combined the TIDIs of 40 inspection runs over 3 years to obtain the TIDI distribution matrixes in the spatial and temporal domains for different track infrastructures. In the spatial domain, we extracted the most probable abnormal position of the track slab using kernel density estimation. In the temporal domain, we developed a logarithmic–linear regression model and noted that the growth trend of the TIDI of the abnormal bridge gradually slows with time.

In this paper, a disturbance-based methodology was proposed to represent the safety level of a car-following scenario. According to the probabilistic causal model, the following vehicles always take evasive actions to avoid a collision in the crash mechanism. In this study, the probabilistic model is modified to evaluate the maximum disturbance a car-following scenario can accommodate corresponding to the maximum evasive action taken by the following vehicle. This paper aims to investigate the safety level of a car-following scenario by estimating its capability index on accommodating disturbance. The surrogate measure, Disturbance Accommodate Index (DAI), is thus proposed to represent the stability of a car-following scenario by measuring its maximum capability on accommodating disturbance. Further, a case study is conducted to evaluate the performance of DAI by using the traffic and crash data provided by the Department of Transport and Main Roads in Queensland. The results show that the DAI outperforms the other crash surrogate measures (e.g., Aggregated Crash Index, Time to Collision, and Proportion of Stopping Distance) in representing rear-end crash risk, followed by the discussion.

Accurate short-term traffic flow prediction is essential for real-time traffic control. A linear hybrid method and a nonlinear hybrid method for short-term traffic flow prediction are proposed with vehicle type as one concern. Traffic flow data are divided into the similar, volatile, and irregular parts. The selected methods are the autoregressive integrated moving average and generalized autoregressive conditional heteroscedasticity (ARIMA-GARCH) model, the Markov model with state membership degree, and the wavelet neural network. The ARIMA-GARCH model is used to predict the similar and volatile parts, and the other methods are adopted to predict the irregular part. This paper aims at providing better prediction methods for short-term traffic flow, and comparing the advantages and disadvantages of the linear and nonlinear hybrid methods. Additionally, the impacts of vehicle type on the predicted values are analyzed. The proposed methods are tested using field data from Dalian, China, and Hefei, China. The results suggest that the developed nonlinear hybrid method should be used with vehicle type and sampling interval as concerns.

Truck-based transportation is the predominant mode used to transport goods and raw materials within the United States. While trucks play a major role in local commerce, a significant portion of truck activity is also long haul in nature. Long-haul truck drivers are continuously faced with the problem of not being able to secure a safe parking spot since many rest areas become fully occupied, and information about parking and availability is limited. Truck drivers faced with full parking lots/facilities either continue driving until a safe parking spot is located or park illegally. Both scenarios pose a hazard to the truck driver, as well as the surrounding road users. Disseminating forecasts of parking availability to truck drivers may help mitigate this hazard, since many truck drivers plan their parking in advance of arrival. Building on 1 year of nearly continuous truck parking data collection, this paper proposes and demonstrates a method for developing a dynamic forecasting model that can predict truck parking occupancy for any specified time within the present day, using only truck parking occupancy data from a trucking logistics facility in the northern San Joaquin Valley during 2016. Different versions of the dynamic model were studied and verified against successive weekdays with performance measured using the root-mean-square error (RMSE). Results indicated that for a particular day, the maximum error can range between 13 and 40 trucks, about 5% of the absolute maximum capacity of the facility.