Top Read

Published in last 1 year |

In last 2 years |

In last 3 years |

All

Please wait a minute...


For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails

Select

A Review of AI-driven Trajectory Prediction Methods for Autonomous Vehicles TIAN Daxin, XIAO Xiao, ZHOU Jianshan Journal of Transportation Systems Engineering and Information Technology    2025, 25 (5): 1-24.   DOI: 10.16097/j.cnki.1009-6744.2025.05.001 Abstract （1878）      PDF （1838KB）（859）       In autonomous driving systems, trajectory prediction plays an important role in connecting vehicle's perception and decision-making, and enhancing driving safety and overall system robustness. In recent years, with the continuous advancement of artificial intelligence (AI), the AI-driven trajectory prediction methods have seen significant progress in terms of accuracy, adaptability, and the capability to model complex traffic environment. This paper provides a systematic review of mainstream trajectory prediction methods in autonomous driving, with a focus on predictive model frameworks. It first revisits traditional physics-based approaches, and then highlights current research trends, including modeling paradigms based on classical machine learning, deep neural networks, and reinforcement learning. Additionally, recent developments in explainable AI techniques aimed at improving model transparency and safety are discussed. Based on comparative analysis, the paper evaluates the strengths and limitations of various models in interaction modeling, multimodal uncertainty, and generalization capability. Furthermore, it organizes trajectory prediction evaluation metrics and publicly available trajectory prediction datasets according to their characteristics and application scenarios, and summarizes representative real-world deployments from both domestic and international sources. At last, considering the existing research bottlenecks and future development trends, the paper outlines potential directions for future studies, such as enhancing model interpretability, effectively integrating multimodal information, and designing unified frameworks for joint prediction and planning. The purpose of this review is to provide insights and references that can be used in the future research and applications.

Select

Drone Delivery: A Systematic Review on Technology, Efficiency, and Applications WU Jingqiong, DIAN Ran, ZI Taisheng, LI Yunqi Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 34-49.   DOI: 10.16097/j.cnki.1009-6744.2025.06.004 Abstract （1863）      PDF （2284KB）（775）    PDF(English version) （466KB）（9）    With the rapid development of e-commerce and a surge in demand for instant delivery, drone delivery, as an innovative solution in the logistics sector, is driving profound transformations in the logistics system. This paper synthesizes findings from 74 relevant articles published between 2015 and 2024, comprehensively examining drone delivery research advancements across key dimensions including critical technologies, economic benefits, environmental sustainability, application potential, and system synergy. The results indicate that drone delivery primarily relies on path planning algorithms, energy management, and multi-drone collaboration as its core technologies. Related optimization research has evolved from single-objective to multi-objective coordination, with algorithms transitioning from classical heuristics to intelligent approaches, effectively reducing solution time and optimizing costs. However, nonlinear effects of payload and wind resistance, along with adaptability to harsh weather conditions, remain bottlenecks. In terms of economic benefits, drone-vehicle collaborative systems can significantly reduce customer waiting time, delivery costs, and labor demands through optimized path planning and resource scheduling. Integration with public transit systems (bus/subway) effectively expands service coverage while reducing energy consumption. Multi-objective optimization models dynamically balance energy consumption, cost, and timeliness to further enhance synergistic benefits. Nevertheless, economic viability remains constrained by payload and range limitations, showing greater advantages in short distance, lightweight deliveries, particularly for emergency cargo deliveries. Environmental benefit analyses demonstrate that the operational phase of drone delivery exhibits significantly lower carbon emissions than traditional transportation methods, though a comprehensive lifecycle assessment encompassing manufacturing, operation, and recycling phases is required. Regarding applications, drone delivery technology demonstrates unique value in medical supply distribution, emergency logistics, and urban "last- mile" delivery, with particular advantages in remote areas and urgent emergency scenarios. However, challenges persist regarding safety risks, technological innovation gaps, limited social acceptance, and imperfect policy regulations. Future research should prioritize battery technology breakthroughs, intelligent path planning optimization, privacy/security safeguards, and cross regional policy coordination to accelerate drone delivery commercialization.

Select

Intermodal Transportation Network Design Optimization Considering Demand Uncertainty Under "Dual Carbon" Background HUANG Rui, ZHAO Xu, WANG Jingyun Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 1-12.   DOI: 10.16097/j.cnki.1009-6744.2025.06.001 Abstract （1083）      PDF （2693KB）（567）    PDF(English version) （1131KB）（7）    A central challenge in modern intermodal transportation planning is the simultaneous consideration of "Dual Carbon" goals and growing fluctuations in freight demand. To address this challenge, this study presents an optimization model for intermodal transport network design. First, a bi-level bi-objective optimization model is developed, with the strategy planner serving as the upper-level leader and shippers as the lower-level followers. The upper level jointly determines the capacity expansion investment, low-carbon investment, and subsidy policies, with the objective of maximizing total revenue while minimizing total carbon emissions. The lower layer solves the network cargo flow allocation under user equilibrium based on generalized transportation costs. Then, the theory of real options is introduced, and geometric Brownian motion is used to describe the stochastic process of transportation demand fluctuations. This enables the quantification of the option value of delayed optimization to determine the optimal timing for strategy implementation. Based on the model characteristics, a nested Frank Wolfe multi-objective evolutionary algorithm based on decomposition (MOEA/D) is designed to solve the deterministic model, combined with a least squares Monte Carlo simulation algorithm to get the optimal implementation timing. Empirical analysis along the Western Land-Sea New Corridor shows that the proposed method simultaneously balances the economic, low-carbon, and operational efficiency optimization goals, which results in a 16.58% decrease in unit transportation costs, a 27.11% decrease in total carbon emissions, and a robust 5.41% increase in total revenue. Under demand uncertainty, delaying the implementation of optimization strategies can generate additional option value. In the case study, delaying to the third period can increase expected revenue by 4.70% and reduce total carbon emissions by 5.03%.

Select

Review of Connected and Autonomous Vehicle Dedicated Lane Setup CHENG Guozhu, WANG Wenzhi, YANG Zihan, WANG Guopeng, CHEN Yongsheng, GU Shuang Journal of Transportation Systems Engineering and Information Technology    2025, 25 (5): 25-39.   DOI: 10.16097/j.cnki.1009-6744.2025.05.002 Abstract （996）      PDF （2429KB）（491）    PDF(English version) （737KB）（27）    With the advancement of information and communication technologies, autonomous vehicles (AV) and connected and autonomous vehicles have emerged as promising solutions to address traffic congestion, to enhance traffic safety, and to improve overall traffic efficiency. This paper provides a comprehensive review of the methods for setting up dedicated lanes for connected and autonomous vehicle (CAV). It begins by tracing the evolution of CAV dedicated lanes and elaborating on the background and significance of their implementation. Based on relevant literatures, it then delves into the methodologies for calculating road capacity, providing a foundation for predicting the impact of CAV dedicated lanes on traffic operations, evaluating strategies, and making necessary adjustments. Furthermore, the paper conducts an in-depth analysis on the strategies for setting up CAV dedicated lanes, including the conditions based on CAV penetration rates and traffic demand. It also explores the determination of the number and location of lanes, access methods, and lane separation approaches under various influencing factors. Finally, the paper proposes that future research should focus on understanding the changes in influencing factors post-implementation of CAV dedicated lanes and their alignment with real-world traffic conditions. It also emphasizes the need for establishing specific standards for setting up CAV dedicated lanes to ensure their function effectively across different traffic scenarios.

Select

Cooperative Optimization of Lane Allocation and Vehicle Trajectory at Intersections Under Connected-and-Automated-Vehicle Environment SONG Lang, HU Xiaowei, YU Shanchuan, AN Shi Journal of Transportation Systems Engineering and Information Technology    2025, 25 (5): 59-71.   DOI: 10.16097/j.cnki.1009-6744.2025.05.005 Abstract （888）      PDF （2673KB）（332）    PDF(English version) （1984KB）（29）    In the collaborative optimization of intersection signal timing and Connected and Automated Vehicle (CAV) trajectory planning, the CAV exit, left turn, through, and right turn lanes can be assigned dynamically in the operation period. Based on the characteristics of CAV technology, this paper proposes a set of dynamic control rules for lane assignment under CAV, named as "flexible lane strategy". Compared to the existing fixed lane strategy, the proposed strategy can adjust the number of exit lanes and entrance lanes (including left turn, through, right turn) for different directions of traffic flow during operation. Lane assignment, signal timing and CAV trajectory planning are incorporated into a unified optimization framework to build a mixed integer linear programming optimization model. Meanwhile, feasible phase and sequence schemes can be automatically generated according to lane assignment in each direction, and the effectiveness of the model is verified through a case study. The results show that the optimization model can generate the optimal lane assignment scheme according to the traffic demand of each flow direction, especially when the lane assignment of the fixed lane strategy does not match the traffic composition of each flow direction, the flexible lane strategy helps to improve the intersection traffic efficiency. In low flow scenario, the flexible Lane strategy reduces average vehicle delay by 4.08%. In high-traffic scenarios, the fixed lane strategy at the intersection will be in a supersaturated state, while the flexible lane strategy can still meet the demand.

Select

Optimization Algorithm for Ride-hailing Dispatch Under Concentrated Supply and Sparse Demand WANG Jiangfeng, SONG Zhifan, LI Yunfei, QI Chongkai, YAN Xuedong, LI Qingshan Journal of Transportation Systems Engineering and Information Technology    2025, 25 (5): 280-290.   DOI: 10.16097/j.cnki.1009-6744.2025.05.025 Abstract （868）      PDF （2451KB）（197）       This study addresses the structural imbalance of “concentrated supply and sparse demand” in ride-hailing services across emerging urban areas by introducing incentive strategies to optimize dispatch decisions and alleviate spatial mismatches between supply and demand. A reinforcement learning-based dispatch optimization algorithm is proposed, integrating incentive mechanisms into four key modules: environment construction, feasible decision selection, incentive embedding, and policy optimization. Specifically, a Markov Decision Process-based environment is developed to model tasks such as order matching, idle vehicle repositioning, and charging management. A joint incentive strategy combining direct financial incentives (via reduced platform commission) and indirect regulatory adjustments (via prioritized dispatch guarantees) is designed, and an Actor-Critic algorithm is used to maximize platform revenue. Empirical analysis using data from Xiong’an New Area in Hebei Province demonstrates that the proposed method significantly improves the service rate of sparse-demand orders with sparse demand. Under fixed fleet size, the service rates of sparse-demand orders respectively increased by 24.70%, 2.53%, and 26.09% through commission reduction, prioritized dispatch, and joint incentives, while maintaining or increasing overall service rate and platform revenue. As the ride-hailing vehicle fleet expands, the sparse-demand and overall order fulfillment rates reach 61.20% and 81.55% respectively, when the fleet size reaches 120 vehicles. Sensitivity analysis further indicates that a 24% commission rate and a 4-period dispatch guarantee yield a favorable balance between platform revenue and global service performance.

Select

《交通运输系统工程与信息》编辑部 Journal of Transportation Systems Engineering and Information Technology    2026, 26 (1): 1-.   Abstract （854）      PDF （774KB）（224）

Select

Optimization of Operation Plans for Full-length and Short-turn Routings of Urban Rail Transit Based on Flexible Train Composition LIU Bin, ZHAO Jinhui, TIAN Zhiqiang, MA Chaofan, LIANG Hui, LI Hebi Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 143-152.   DOI: 10.16097/j.cnki.1009-6744.2025.06.013 Abstract （829）      PDF （2072KB）（599）    PDF(English version) （623KB）（177）    To address capacity redundancy or shortages caused by uneven passenger flow distribution in urban rail transit, this study optimizes train operation plan under flexible train composition during peak hours to achieve precise demand-capacity alignment, enhance cost control, and improve resource utilization efficiency. Focusing on asymmetric passenger flow patterns, this study develops a bi-objective nonlinear integer programming model with the goals of minimizing operator costs and maximizing average load factor and the constraints such as standing-passenger density limits. A hierarchical three-stage algorithm and Technique for Order Preference by Similarity to Ideal Solution are designed for solution prioritization. The model and algorithm are validated through case studies of an operational metro line. Results demonstrate that flexible train composition outperforms fixed-composition strategies (single routing and combined large/small routing). During peak hours, the operating costs decreased by 25.67% compared with the single routing method, and 7.34% compared to the combined large/small routing method. The average load factor improvements are respectively 29.55% and 26.02% compared to the single routing and combined large/small routing. Sensitivity analysis indicates that allowing a moderate increase in standing-passenger density for small routing can further reduce operational costs, though requires a balance between passenger comfort and safety. When the standing-passenger density increases from 7 to 9 persons ⋅ m -2, the flexible train composition reduces costs by 15.38%, decreases the load factor ratio 14.83%, and lower the required number of train sets by 15.79%. The flexible train composition effectively adapts to spatiotemporal passenger flow fluctuations, and could achieve synergistic optimization of dynamic capacity allocation and cost control.

Select

Resource Allocation and Multi-machinery Cooperative Scheduling Optimization in Automated Dry Bulk Terminals JI Mingjun, LI Jiawei, HU Hanlin, GAO Zhendi Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 317-326.   DOI: 10.16097/j.cnki.1009-6744.2025.06.029 Abstract （794）      PDF （2707KB）（295）       This paper addresses the difficulties of multi-resource and multi-equipment coordination scheduling in automated dry bulk terminals by proposing a berth-ship loader-stockyard collaborative scheduling model with dual objectives of minimizing ship port time and terminal operation costs. Due to the large-scale and nonlinear characteristics of the problem which bring significant computational challenges, this paper proposes a two-stage algorithm based on grey wolf optimization for model solution. The first stage employs an improved algorithm of grey wolf optimization to solve the berth-ship loader allocation scheme, while the second stage uses an algorithm of distribution machine-stockyard allocation to screen feasible solutions that satisfy operational line and stockyard constraints. Finally, the feasibility and algorithm superiority of model are validated using the data from Anhui Changjiu Inland River Terminal. The numerical results demonstrate that the established optimization model is highly suitable for the operation scenarios in automated dry bulk terminal involving multi-resource and multi-equipment collaborative operations, which conforms to actual operational constraints, and can fully utilize terminal resources and mechanical equipment. The proposed algorithm effectively solves this optimization problem, achieving the solution improvements of 8.1%, 8.7%, 6.5%, 2.4%, and 4.5% in total cost through comparing with the algorithms of particle swarm optimization, genetic, grey wolf optimization, whale optimization, and Harris hawks optimization, respectively.

Select

Safety Optimization Control for Connected Vehicle Platoon Under Acceptable Spacing Policy YANG Haifei, TANG Yong, GUO Yanyong, LI Hongwei, ZHAO Enze Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 50-61.   DOI: 10.16097/j.cnki.1009-6744.2025.06.005 Abstract （776）      PDF （2581KB）（238）       The existing Three-Phase Adaptive Cruise Control (TPACC) based on the acceptable spacing policy realizes speed synchronization with linear controller, improving stability over the constant headway policy but is more likely to induce traffic conflicts when facing complex traffic disturbances. To improve this, this paper uses Model Predictive Control (MPC) to optimize the safety of Coperative TPACC (CTPACC) with the support of vehicle network technology. A kinematic model of CTPACC is developed considering the system delay. The relationship between stability and safety under acceptable spacing policy is clarified using numerical experiments. Then, a safety optimization policy of MPC based on speed synchronization is proposed, and terminal constraints and Bayesian optimization are introduced to maintain the stability. A delay compensation mechanism is designed to improve the control performance. At last, the effectiveness of the proposed policy is verified by typical working conditions. The results show that, excessively high stability margin of the linear controller will lead to rear-end accidents or emergency safety mode intervention, and safety and stability show a non-monotonic relationship. To address this, the proposed MPC without delay compensation achieves an overall improvement in safety while maintaining stability. The delay compensation mechanism further enables comprehensive optimization of both margins of stability and safety. In the theoretical working condition, the reductions in the safety risk indicators of time-integrated time-to-collision, time-exposed time-to-collision, the platoon oscillation indicators of average maximum overshoot, total absolute jerk reach 24.4%~61.7%, 29.7%~57.4%, 52.7%~90.8%, and 13.9%~81.3%. Moreover, sensitivity experiments with disturbance intensity and actual working condition tests consistently demonstrate the same trend. In addition, the proposed policy suppresses the emergency safety mode intervention and improves the smoothness of acceleration control. Phase transition analysis indicates that the delay-compensated MPC, while mitigating congestion nuclei, reduces two safety risk indicators by 50.4% and 53.3% compared to the typical linear controller.

Select

Integrated Optimization of Electric Bus Timetable and Vehicle Scheduling with Piecewise Charging Strategy GAO Wanchen, LU Shichang, ZHAO Yatong, LIU Kai Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 239-248.   DOI: 10.16097/j.cnki.1009-6744.2025.06.022 Abstract （768）      PDF （2366KB）（191）       To reduce the costs of bus enterprises and enhance passenger satisfaction, this paper proposes a dual-objective model to optimize the timetable and vehicle scheduling considering multiple depots and multiple vehicle types. The model combines the piecewise charging strategy, the time-of-use pricing, and time-dependent parameters such as dwelling time, travel time, passenger boarding rate, and passenger alighting rate. The objectives of the model include minimizing the passenger travel time and the cost of bus enterprises. The former includes passenger waiting time and the riding time, while the latter includes fixed costs, deadheading costs and charging costs. The improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) was designed to solve the model. The bus line No. 26 in Zhuhai City was selected for case analysis. The findings indicate that during the morning rush hour, the noon off-peak hour, and the evening rush hour, the solutions derived from both the current plan employed by bus companies and the sequential method are inferior to the Pareto optimal solutions achieved through an integrated optimization approach. Compared with the current plan of bus enterprises, the integrated optimization approach reduces the cost of bus enterprises by an average of 5.1% and the travel time of passengers by 4.9%. Compared with the sequential method, the average travel time of passengers decreased by 1.6%, among which the waiting time of passengers decreased by 1.8% and the riding time of passengers decreased by 1.6%. This research provides references for the scientific dispatching and management of urban bus transit.

Select

Campus Bus Route Design Considering Passenger Transfers ZHAO Ying, LIANG Jinpeng, BAO Yue Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 220-228.   DOI: 10.16097/j.cnki.1009-6744.2025.06.020 Abstract （766）      PDF （2365KB）（260）       In response to the growing internal mobility challenges caused by the expansion of university campuses, this paper proposes an optimization model for campus bus route design and departure frequency setting that incorporates the transfer behavior of passengers. By constructing a bi-layer network structure consisting of travel nodes and transfer nodes, the formulation of flow balance constraints that accurately capture transfer dynamics is designed. The model is formulated as a mixed-integer programming problem, aiming to minimize the travel time of total passengers and the operation costs of buses. It also optimize the decisions of route selection, frequency setting, and passenger path allocation jointly. In the route generation stage, a constrained depth-first search algorithm is proposed to construct a set of candidate routes based on the spatial distribution of travel demand. A case study based on the Xiong'an campus of Beijing Jiaotong University demonstrates the effectiveness of the approach: three optimized routes achieve full coverage of 11 stations, with 70.8% of travel demands satisfied via direct service and the remainder completed with only one transfer. Additionally, 75% of trips incur less than 5 minutes of additional travel time, and over 52% of route segments operate with a load factor above 70%. The results provide both theoretical support and practical guidance for the scientific planning and efficient operation of the campus bus systems.

Select

Cascading Failure Analysis of Urban Rail Transit Network with Bus Substitution Effect LI Guiyang, XIE Binglei, LI Xiaodan Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 185-196.   DOI: 10.16097/j.cnki.1009-6744.2025.06.017 Abstract （764）      PDF （2487KB）（297）       This paper proposes an extended coupled map lattice (SCML) model that incorporates the bus substitution effect to evaluate the mitigating impact of bus networks on cascading failures triggered by disruptions in urban rail transit stations. The bus substitution effect is characterized in terms of transportation accessibility and quantified as a weighted combination of pedestrian transfer convenience and station access opportunities through the CRITIC-TOPSIS method. A mapping function is formulated to depict the regulatory role of the bus substitution effect on station states, and it is integrated into the CML framework to simulate the dynamic evolution of cascading failures. Several vulnerability indicators are defined, including the stepwise proportion of failed stations. The Shenzhen metro-bus system is used for the empirical validation. The results show that the bus substitution effect exhibits spatial heterogeneity, with higher values in the urban core and lower values in peripheral areas. Under low to moderate disturbance intensity, it substantially slows down and reduces the spread of failures, while its effect becomes limited under high-intensity disturbances. Compared with the CML model, the SCML model more accurately captures the vulnerability characteristics of urban rail transit network, raising the average perturbation intensity threshold for network collapse from 1.7 to 2.2. The findings provide quantitative evidence for assessing and enhancing the resilience of multimodal transportation systems.

Select

Impact of Weather on Invulnerability of Low-Altitude Route Network CHENG Ming, HUANG Hongming Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 101-108.   DOI: 10.16097/j.cnki.1009-6744.2025.06.009 Abstract （761）      PDF （2120KB）（267）       To accurately quantify the impact of weather factors and node failures on the destructivity of low-altitude logistics route networks, this paper proposes an evaluation framework that integrates meteorological data spatial interpolation and complex network cascade failures. Using the published elevation data (DEM) as a covariate, the study uses the local thin disk smooth spline method to perform high-precision spatial interpolation of wind speed and rainfall data on low-altitude routes, which overcomes the sparsity of station data and accurately depicts the distribution of meteorological risks on the route. Based on six indicators, including the centrality of the degree of entropy weight method and the centrality of the intermediary, the M si is introduced to identify the key nodes of the network. The comprehensive destructibility index of convergence network efficiency, maximum connected subgraph size, and network density is defined, H , and a cascading failure model considering node failure and load redistribution is developed. The low-altitude logistics route network built by 88 operating stations in Shenzhen is used as the object for simulation. The simulation results show that the comprehensive destructive resistance index, H can fully reflect the change of network destructivity, and the entropy-weighted node importance index M si has a significant effect on identifying key nodes, and the removal of the first 10 nodes leads to a decrease of 80% in H and the network collapse. The quantitative analysis of meteorological impact shows that the constructed network effectively avoids the high wind speed area in Shenzhen, and the rainfall in 60% of the area is lower than the UAV operation threshold, and when the top 30 nodes affected by weather are removed, H only decreases by 28.3%, which proves that the impact of weather on the overall destructive resistance of the network is limited. The resulting destructive optimization scheme are adding 4 alternate landing fields near key nodes. Simulation verification shows that the destructive resistance of the network is significantly improved after optimization, and when the first 10 nodes are removed according to the degree value, H increases from 0.08 to 0.25 before optimization. This study provides safety evaluation tools and optimization strategies for the safe operation and destructive resistance improvement of low-altitude logistics networks in bad weather.

Select

Optimization of Multimodal Transport Considering Coordination of Production and Departure Schedules LI Yajun, XUE Longjiang, ZHENG Jianfeng Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 23-33.   DOI: 10.16097/j.cnki.1009-6744.2025.06.003 Abstract （749）      PDF （2584KB）（242）       In order to solve the problem of the disconnection between production and transportation, this paper constructs a bi objective optimization model which integrates the allocation of production line, production sequence and multimodal transportation route selection. With the purpose of minimizing the total cost and time, two railway departure modes, scheduled departure and fixed-frequency departure, are introduced into the model to truly describe the characteristics of multimodal transportation. To solve the model, an enhanced algorithm of multi-objective particle swarm optimization is proposed, which improves the global search capabilities and convergence speed by dynamically adjusting the parameter of algorithm and combining with the simulated annealing mechanisms. Finally, the numerical experiments are carried out with an example from Busan to Hamburg/Rotterdam. The results show that the optimal scheduling scheme can reduce the total cost by 11.4% and the total time by 580 hours. Compared with the other multi-objective optimization algorithms, the proposed algorithm has significant advantages in the diversity and accuracy of solutions. Scenario analyses reveal that optimizing either production or transportation scheduling separately fails to achieve overall optimality, and increase costs by over 18% potentially. Sensitivity analyses indicate that the moderate adjustments of departure schedule and unit storage costs can effectively influence the selection of transportation modes under the balance cost and time.

Select

Multi-objective Optimization of Truck-Speedboat Coordination for Emergency Material Distribution in Mountainous Floods CHENG Jiahao, HAO Zhidan, LI Guoqi, LIU Sijing Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 294-304.   DOI: 10.16097/j.cnki.1009-6744.2025.06.027 Abstract （748）      PDF （2460KB）（179）       Floods in mountainous areas often disrupt roads, create complex networks of land and water routes and challenge emergency material distribution. This paper develops a bi-objective mixed-integer programming model considering warehouse storage limits and post-disaster network functional differentiation. A land-water intermodal distribution network is established with warehouses, land, and water routes. The proposed model uses the lateral transshipment and the truck-speedboat collaborative strategy to minimize total transportation time and maximize the average demand satisfaction rate. An improved Non-dominated Sorting Genetic Algorithm II (INSGA-II) is designed to solve the model, which includes a two-phase heuristic, hybrid genetic operators, and variable neighborhood search. A case study of a large-scale construction project shows that a hybrid collaborative strategy, allowing speedboats to depart from warehouses, improves the maximum average satisfaction rate by 4.85% and reduces the shortest transportation time by 2.17% compared to a dependent collaborative mode.

Select

Platooning Strategies Study for Connected Autonomous Vehicles on Superhighways HE Yongming, LU Yangpeng, LIU Huiyang, LI Xinran Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 62-73.   DOI: 10.16097/j.cnki.1009-6744.2025.06.006 Abstract （747）      PDF （3005KB）（200）       Robust platoon control strategies are crucial for enabling safe and efficient operation of Connected and Automated Vehicles (CAV) in superhighway scenarios. This paper proposes a CAV platooning strategy to enhance the stability and safety of platoons on superhighways. First, a comprehensive full velocity difference model considering communication delays is developed as the fundamental car-following foundation. Then, under a leader-follower information topology, a nonlinear platoon control law is designed, where the optimal velocity function of the underlying car-following model is inverted to dynamically compute a speed dependent nonlinear desired spacing. This approach ensures consistency between the steady-state platoon target and individual vehicle driving equilibrium. Furthermore, both local and string stability of the proposed model are rigorously analyzed using transfer function methods, and controller parameters are systematically optimized via the H∞ norm. Simulation results demonstrate that, in dynamic speed scenarios ranging from 120 km·h-1 to 160 km·h -1, the proposed Delayed Leader-Follower Control (DLFC) strategy achieves significantly better string stability and tracking performance compared with the benchmark Cooperative Adaptive Cruise Control (CACC) strategy, with smoother acceleration responses that effectively prevent disturbance amplification along the platoon. In the extreme emergency braking scenarios with random disturbances, the comprehensive Monte Carlo simulations verify that the DLFC strategy provides higher safety margins, effectively avoids collision risks, and reduces the probability of high-risk states to below 20%, thereby substantially improving platoon safety under extreme superhighway conditions.

Select

Trajectory Control Method for UAV Spiral Search Oriented to High-rise Building Emergency Rescue CHEN Deqi, ZHANG Zishe, ZHANG Wenhui, YAN Xuedong, JIANG Xiancai Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 87-100.   DOI: 10.16097/j.cnki.1009-6744.2025.06.008 Abstract （741）      PDF （2924KB）（193）       During the golden rescue period following a disaster, unmanned aerial vehicles (UAVs) can be delivered first to reach the damaged buildings for spiral full-coverage scanning and search for survivors. However, due to the complex and dynamic environment at disaster sites, UAVs often encounter challenges such as low trajectory tracking accuracy and high collision risks during close-range three-dimensional (3D) scanning. To address these issues, this paper proposes a Prioritized Experience Replay Soft Actor-Critic (PER-SAC) control model and establishes a high-fidelity simulation platform based on a 6-degree-of-freedom (6 DOF) nonlinear dynamic model. By prioritizing the learning of key experiences with high Temporal-Difference error (TD-error), the model enhances learning efficiency and policy robustness in complex tasks. Comparative simulation experiments demonstrate that the proposed PER-SAC strategy outperforms both Soft Actor-Critic (SAC) and Proximal Policy Optimization (PPO) algorithms in terms of convergence speed and final performance. In static trajectory tracking tasks, PER-SAC achieves a success rate of 99.0%, with an average trajectory error reduced by 66.3% compared to SAC. For dynamic obstacle avoidance tasks, the success rate is 97.0%, exhibiting smoother and more efficient evasion maneuvers, thereby fully validating its robustness. The incorporation of prioritized experience replay significantly improves UAVs' autonomous flight performance in unknown dynamic environments. The proposed PER-SAC strategy represents an advanced control method that effectively balances control precision, flight quality, and safety. It can be directly applied to autonomous spiral scanning of high-rise damaged buildings post-disaster, enabling stable flight attitudes to capture high-definition imagery. This capability assists rescue teams in rapidly locating trapped individuals, thereby enhancing emergency search and rescue efficiency

Select

Exploration of Transfer Time at Metro Transfer Stations with Automatic Fare Collection Data LIU Chenhui, TAO Mengxin Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 175-184.   DOI: 10.16097/j.cnki.1009-6744.2025.06.016 Abstract （725）      PDF （2623KB）（162）       Transfer efficiency is a key factor affecting the service quality of urban rail transit systems, and analysis on transfer time is essential to develop effective improvement strategies. This study investigates the transfer time based on the Automated Fare Collection (AFC) data of Changsha Metro system in one week. A breadth-first search (BFS) algorithm with transfer-count constraints was developed to identify shortest transfer paths. Based on the identified trips, a linear regression model was constructed, and parameters were estimated by using the ordinary least squares (OLS) method to quantify transfer times. Transfer characteristics were analyzed from three perspectives: spatial distribution, temporal variation, and relative duration within the entire trip. The results show that the average transfer time in Changsha Metro is 6.2 minutes, accounting for 17.6% of the total travel time. Transfer times on weekdays are generally longer than that on weekends, with noticeable efficiency degradation during peak hours. In terms of spatial distribution, significant differences exist among transfer stations, and K-means clustering was further applied to classify stations by transfer performance. Seven transfer stations with passageway-type layouts were identified as low-efficiency nodes. Based on the findings, this paper proposes optimization strategies including structural improvements, real time passenger guidance, and inter-line interoperability to improve the transfer efficiency in metro system.

Select

Method for Short-term Passenger Flow Prediction in Urban Rail Transit Networks Considering Data and Model Uncertainty MU Liang, KANG Yurui, YAN Zixu, ZHU Guangyu Journal of Transportation Systems Engineering and Information Technology    2025, 25 (6): 118-128.   DOI: 10.16097/j.cnki.1009-6744.2025.06.011 Abstract （723）      PDF （3604KB）（152）       This paper proposes a short-term probabilistic model for forecasting passenger flow in urban rail transit networks, termed PD-STGCN, which considers the uncertainties of both data and model. It aims to obtain the probabilistic information of network-wide passenger volume at future time steps. The model constructs a collaborative framework consisting of the Spatio Temporal Uncertainty Prediction Module (STUPM) and Probabilistic Quantification Module (PQM). In the STUPM, addressing uncertainties inherent in both the passenger flow data and forecasting model, a novel loss function is developed by integrating Gaussian Negative Log-Likelihood (GNLL) and Monte Carlo Dropout (MC Dropout) techniques to quantify these dual uncertainties. Within the PQM, the discrete sample sets are obtained through the random normal sampling based on prediction results, and continuous probabilistic forecasting outputs are generated using the Gaussian Kernel Density Estimation (KDE). Using the data of passenger flow from a urban rail transit system in a large city as a case study, the model is validated under both weekday and non-weekday scenarios. The results demonstrate that, compared to the baseline forecasting models, PD-STGCN improves the Prediction Interval Coverage Probability (PICP) and Continuous Ranked Probability Score (CRPS) by 8.01% and 20.77%, respectively, which provides a better coverage of actual passenger flow values and has higher forecasting accuracy. Ablation experiments confirm that uncertainty is the most significant factor affecting model performance. The model considering dual uncertainties leads to the improvements of at least 1.91% in PICP and 4.02% in CRPS over that considering only a single type of uncertainty.