Journal of Advanced Transportation最新文献

This study aims to develop a machine learning-based framework for predicting the severity of highway traffic accidents by leveraging high-resolution data from Taiwan’s Electronic Toll Collection (ETC) system. Unlike traditional accident-reporting systems, the ETC infrastructure provides a uniquely comprehensive and precise dataset that captures spatiotemporal traffic patterns and environmental conditions across the national highway network. This rich dataset enabled the integration of data mining and data visualization techniques to uncover nontypical contributing factors to accident severity. Feature engineering was conducted using random forest and LASSO regression, while extreme gradient boosting and the Apriori algorithm were employed to identify key associations between accident severity and contextual variables. Based on human factor and traffic psychology theory, influential factors include poor lighting at night, adverse weather conditions, late-night hours (20:00–06:00), specific geographic regions (e.g., Yilan County), speed limits of 100 km/h, and vehicle types such as taxis and large trucks. The findings not only enhance the understanding of environmental influences on accident outcomes but also offer actionable insights for improving highway safety. Moreover, Taiwan’s ETC system serves as a model for countries seeking to integrate tolling infrastructure with traffic safety analytics.

Driving behavior modeling is a crucial yet challenging task in the development of traffic simulation systems. Advances in machine learning and data-driven vehicle trajectory extraction technologies have significantly advanced research in this area. However, the performance of such models can be affected by numerous factors often overlooked by the existing methods, including the complexity of real-world road environments and driver characteristics. In this paper, we introduce a novel modeling approach, termed the customized generative adversarial imitation learning (Cus-GAIL) method, designed to capture these complex factors. Our approach incorporates a conditional imitation learning technique that utilizes traffic’s prior knowledge to train a reinforcement learning (RL) model. In addition, we have innovatively developed a collision avoidance mechanism that markedly improves the reliability of microscopic traffic simulation. To address variations in driving styles, we have also created a driver classifier. Moreover, we propose a method for synthesizing small-sample vehicle trajectory data to enhance the RL model’s ability to perceive rare data scenarios. By integrating these components, our model effectively encapsulates a wide range of external and internal factors. To validate the efficacy of the Cus-GAIL method, we employ an unmanned aerial vehicle (UAV) to monitor the two road segments and gather video data of actual vehicle trajectories. The experimental results demonstrate that the Cus-GAIL method outperforms established baselines on both microscopic and macroscopic metrics.

The disruptions in supply chains have put small- and medium-sized enterprises (SMEs) in dire need of resilient supply chains through which they can improve their performance. Based on the resource dependence theory, this study proposes a mediation model to improve the environmental performance (EP) of SMEs. The purpose of this study is to investigate the effect of supply chain resilience (SCR) on EP mediated by ambidextrous green innovation (AMGI). We proved a structural equation model based on questionnaire data from 261 companies in Iraq to test our hypotheses. The results show that SCR has a positive effect on AMGI for proactive and exploitative green innovation dimensions and positive impact on SMEs’ EP. AMGI plays a mediating role and positively affects EP in dimensions. Building SCR requires management support through proactive and reactive measures to address disruption risks. AMGI necessitates integration with supply chain members, including external suppliers and customers, and involves them in developing a corporate strategy that supports environmental issues. By emphasizing EP improvement, this study will guide practitioners in developing innovative techniques that contribute to improving the EP of SMEs and urging decision-makers to support SMEs that include EP within their strategy.

Infrastructure-to-infrastructure (I2I) communication enables the exchange of traffic data between intersections, which brings a new challenge to urban traffic control. This paper proposes a novel deep reinforcement learning (DRL) framework for urban traffic signal control within the vehicle-to-everything (V2X) ecosystem. The framework incorporates a joint-state representation integrating traffic data from both I2I and vehicle-to-infrastructure (V2I) communications, while considering communication range effects. The reward function is designed to optimize both local intersection conditions and global network performance, facilitating adaptive signal coordination. A simulation case study in Huzhou, China, evaluates the proposed model against conventional pretimed, actuated, nonjoint-state DRL, and joint-state reinforcement learning (RL) models. Results demonstrate superior performance of the joint-state DRL model in episode rewards, average speed, and average time loss, particularly during peak traffic periods. To address data limitations, Monte Carlo cross-validation (MCCV) is employed, further validating the model’s robustness. Results show consistent performance advantages in average speed and time loss across peak and off-peak periods, with slight variations compared to joint-state RL models in certain intervals. The impacts of the communication range are also discussed with the proposed model. Pearson correlation analysis reveals a strong positive correlation between the communication range and convergence time across all traffic periods. Meanwhile, correlations between the communication range and reward, average speed, and average time loss vary by traffic period. Findings highlight the transformative potential of integrating DRL with V2X communication technologies for enhancing traffic signal control in complex urban environments. The proposed model offers a flexible, adaptive approach to traffic management, optimizing flow while maintaining safety standards, with implications for future smart city developments.

This paper constructs a dynamic crossing model for motorized and nonmotorized vehicles at right-turn intersections based on the Lotka–Volterra framework. The study explores equilibrium points and stability conditions within the symbiotic evolution model. By incorporating the Allee effect, the model simulates various stability scenarios and proposes strategies for optimizing the symbiotic evolution of motorized and nonmotorized vehicles. The findings reveal that the evolutionary trends of the interaction system align closely with the predicted trajectories of the proposed model. The stability of the symbiotic evolution model is influenced by the threshold of the Allee effect, initial crossing scales, and the cooperation and suppression coefficients between motorized and nonmotorized vehicles. These insights not only provide a novel perspective for understanding the interaction dynamics of motorized and nonmotorized vehicles but also offer theoretical guidance for improving safety and efficiency at right-turn intersections.

It is important to establish appropriate performance indicators so that decision-makers can better determine the best alternatives for sustainable urban freight distribution systems. This literature about urban logistics and routing problems is structured and analyzed through a systematic literature review of a total of 201 papers from 2002 to 2023. Three main axes were considered: problem modeling and solution approaches, multimodal transportation, and indicators to assess the performance and sustainability of the distribution networks. There is a growing trend of research on this topic. Indeed, the paper highlighted the academic interest in the analysis of case studies to test the scenarios and network configurations and proposed solution approaches, as well as the adoption of greener transportation modes. To the best of our knowledge, no previous studies have analyzed the literature from the thematic lines proposed in this review, especially those that refer to performance indicators to assess both the freight distribution networks and the transportation modes considered. Advancing stochastic modeling, expanding case studies to underrepresented regions, integrating AI-driven multimodal logistics, and developing social impact indicators are key research directions to enhance the sustainability and efficiency of urban logistics. This review provides a structured foundation for future research by identifying gaps in the literature and offering a thematic roadmap to advance the study and implement sustainable urban logistics solutions. In addition, its findings can assist decision-makers and logistics planners in evaluating current practices, identifying opportunities for improvement, and supporting the development of more sustainable and efficient distribution strategies.

Mode choice behavior directly affects the layout of the urban transportation system and serves as the foundation for the development of policies about the planning and administration of urban transportation. This study focused on the city’s various transportation options, aiming to pinpoint and investigate several factors influencing transportation mode selection. The data included the traffic survey; traveler interviews were gathered using a questionnaire survey, structured interviews, and secondary documents. The collected data from the questionnaire survey were then analyzed using a multinomial logit (MNL) model to assess the relationships between different parameters and mode choice. The investigation considered several concerns, including travel distance, travel time, travel cost, safety, environmental impact, health benefits, and comfort. Both qualitative and quantitative methods of sustainability have been integrated into this study. The MNL model’s pseudo-R-squared value illustrates the apparent correlation between the independent and dependent variables. The multilayer perceptron (MLP) model was used as a comparison model. The results show that MLP has higher predictive performance than the MNL model in assessing transport mode choice in the city. The study reveals that travel distance, time, availability, health benefits, comfort, safety, cost, and environmental impact significantly influence mode choice for work trips. Public services are safer, less environmentally impactful, and more accessible, while walking is safest, offers health benefits, and is more environmentally friendly but is preferred by the youngest. Private vehicle users offer more safety but are less cost-effective. Minibus users provide better cost-benefit and safety but take longer travel times. Overall, the study was used to understand passenger preferences and critical factors in transport options, thereby aiding policymakers in making informed decisions and suggestions for improving the transport system in similar cities.

In order to better understand the factors that affect Hangzhou residents’ acceptance of the license plate–based restriction (LPR) policy, new factors such as fairness, family life cycle factors, and preacceptance of alternative measures were added to explore new interactions between different factors. A questionnaire survey was completed among 958 residents of Hangzhou City, and a partial least squares structural equation model (PLS-SEM) was established to analyze the factors that affect the acceptance of the LPR policy. An analysis of socioeconomic attributes is conducted to explore the impact of education, age, and family life cycle factors on the acceptance of the LPR policy. The results indicate that perceived cost-effectiveness, social norms, policy cognition, fairness, important goals, and preacceptance of alternative measures have significant direct effects on the postacceptance of the LPR policy, while fairness and important goals have indirect effects through social norms. Regarding postacceptance, perceived effectiveness can only indirectly affect postacceptance of the LPR policy through policy cognition and perceived cost-effectiveness. Responsibility attribution can only indirectly affect postacceptance through important goals. As the education level and age increase, residents’ acceptance of the LPR policy will decrease; young families without children and families with minor children have lower acceptance of the LPR policy than families with all adult members and elder families without children.

Vehicle forward collision warning based on machine vision can help to reduce the incidence of traffic accidents. Many researchers have studied this topic in recent years. However, most of the existing studies only focus on one stage of the process such as vehicle detection and distance measurement. It will face many issues in practical application. To solve these problems, we propose a framework for forward collision warning. This study applies the YOLO algorithm to detect the vehicle and uses the Kalman filter to track the vehicle. The monocular vision distance measuring method is used to estimate the distance and travel speed. Finally, we adopt the time to collision (TTC) to decide whether to trigger the warning process. In the speed measurement stage, we design an appropriate time interval to calculate the relative speed of the front vehicle. In the collision warning segment, a TTC threshold is set by considering not only vehicle safety guarantees but also avoiding hard barking that would make drivers uncomfortable. Furthermore, we set a warning area to filter the false warning when the car overtakes and meets. Experiments with real traffic scenes demonstrate that the performance of the proposed model is good to make accurate collision prediction and warning.

The customized bus in operation faces numerous random factors that affect the service level and attractiveness to passengers. Therefore, this paper investigates the optimization problem of customized bus routes considering random vehicle travel times and the capability to respond to dynamic requests in real time. We developed a stochastic programming model that minimizes total cost and passenger travel time. The innovation lies in the model’s ability to respond to requests made by passengers during service and to model the randomness of vehicle travel times using a known distribution. Furthermore, we propose a heuristic algorithm combining the nondominated sorting genetic algorithm II (NSGA-II) and a variable neighborhood search operator. This algorithm starts by generating an optimized initial path based on initial reservation demands and then employs a dynamic adjustment mechanism to respond to real-time requests. The effectiveness and superiority of our algorithm are validated through an illustrative example. Finally, numerical experiments using taxi trajectory data demonstrate that considering both randomness and real-time aspects can significantly reduce the total cost and penalties for early and late arrivals and improve the bus service level.