Transportation Research Part E-Logistics and Transportation Review最新文献_第8页

ISAC-empowered deep reinforcement learning scheme for eVTOL approach trajectory optimization with radar point cloud 基于isac的雷达点云eVTOL进近轨迹优化深度强化学习方案

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-15 DOI: 10.1016/j.tre.2026.104681

Leyan Chen , Yuhao Wang , Shulu Chen , Yong Sun , Kai Wang

Urban air mobility (UAM) has emerged as a promising solution to alleviate ground traffic congestion by enabling the use of low-altitude airspace for passenger and cargo transportation. Among UAM, electric vertical take-off and landing (eVTOL) aircraft are expected to play a central role in future urban transport systems due to their flexibility, zero-emission operation, and compatibility with existing infrastructure. However, ensuring the safe and efficient operation of eVTOLs during the approach and landing phase remains a major challenge, especially in urban environments where intelligent unmanned aerial vehicles (UAVs) simultaneously perform logistics and monitoring tasks at low altitudes. Reliable sensing and communication are therefore essential to guarantee operational safety, connectivity, and energy efficiency. To address these challenges, this paper proposes an integrated sensing and communication (ISAC) framework for eVTOL approach trajectory optimization. The framework integrates radar point cloud (RPC) sensing with the three-dimensional radio knowledge map (RKM) to enhance environmental awareness and communication reliability in dense urban airspace. Based on this framework, a fusion deep point reinforcement learning (FDPRL) algorithm is developed to optimize eVTOL trajectories under age-of-information and energy constraints jointly. It includes the RPC feature extraction module for UAV sensing, the RKM feature extraction module for communication enhancement, and a decision-making module for trajectory control. Simulation results demonstrate that the proposed FDPRL achieves challenge performance, outperforming all the benchmarks, enabling the eVTOL to adaptively adjust its approach trajectory to avoid UAVs while maintaining efficient communication, thus achieving superior total communication capacity and maximum residual energy.

城市空中交通（UAM）已经成为缓解地面交通拥堵的一种有前途的解决方案，它允许使用低空空域进行客运和货运。在UAM中，电动垂直起降（eVTOL）飞机由于其灵活性、零排放运行以及与现有基础设施的兼容性，预计将在未来的城市交通系统中发挥核心作用。然而，确保eVTOLs在进近和着陆阶段的安全和高效运行仍然是一个主要挑战，特别是在城市环境中，智能无人机（uav）同时在低空执行后勤和监控任务。因此，可靠的传感和通信对于保证运行安全性、连接性和能源效率至关重要。为了解决这些问题，本文提出了一种用于eVTOL进近轨迹优化的集成传感和通信（ISAC）框架。该框架将雷达点云（RPC）传感与三维无线电知识地图（RKM）相结合，以增强密集城市空域的环境意识和通信可靠性。基于该框架，提出了一种融合深度点强化学习（FDPRL）算法，用于信息时代和能量约束下的eVTOL轨迹优化。它包括用于无人机感知的RPC特征提取模块、用于通信增强的RKM特征提取模块和用于轨迹控制的决策模块。仿真结果表明，所提出的FDPRL实现了挑战性能，优于所有基准测试，使eVTOL能够自适应调整其接近轨迹以避开无人机，同时保持高效通信，从而获得优越的总通信容量和最大剩余能量。

{"title":"ISAC-empowered deep reinforcement learning scheme for eVTOL approach trajectory optimization with radar point cloud","authors":"Leyan Chen , Yuhao Wang , Shulu Chen , Yong Sun , Kai Wang","doi":"10.1016/j.tre.2026.104681","DOIUrl":"10.1016/j.tre.2026.104681","url":null,"abstract":"<div><div>Urban air mobility (UAM) has emerged as a promising solution to alleviate ground traffic congestion by enabling the use of low-altitude airspace for passenger and cargo transportation. Among UAM, electric vertical take-off and landing (eVTOL) aircraft are expected to play a central role in future urban transport systems due to their flexibility, zero-emission operation, and compatibility with existing infrastructure. However, ensuring the safe and efficient operation of eVTOLs during the approach and landing phase remains a major challenge, especially in urban environments where intelligent unmanned aerial vehicles (UAVs) simultaneously perform logistics and monitoring tasks at low altitudes. Reliable sensing and communication are therefore essential to guarantee operational safety, connectivity, and energy efficiency. To address these challenges, this paper proposes an integrated sensing and communication (ISAC) framework for eVTOL approach trajectory optimization. The framework integrates radar point cloud (RPC) sensing with the three-dimensional radio knowledge map (RKM) to enhance environmental awareness and communication reliability in dense urban airspace. Based on this framework, a fusion deep point reinforcement learning (FDPRL) algorithm is developed to optimize eVTOL trajectories under age-of-information and energy constraints jointly. It includes the RPC feature extraction module for UAV sensing, the RKM feature extraction module for communication enhancement, and a decision-making module for trajectory control. Simulation results demonstrate that the proposed FDPRL achieves challenge performance, outperforming all the benchmarks, enabling the eVTOL to adaptively adjust its approach trajectory to avoid UAVs while maintaining efficient communication, thus achieving superior total communication capacity and maximum residual energy.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104681"},"PeriodicalIF":8.8,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

What if rebalancing fleets could adapt? A two-stage stochastic model for dynamic bike redistribution 如果重新平衡的船队能够适应呢？自行车动态再分配的两阶段随机模型

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-15 DOI: 10.1016/j.tre.2025.104640

Mohammadreza Eslamipirharati , Maryam Motamedi , John Doucette , Nooshin Salari

Bike-sharing systems are an important mode of transportation, enabling individuals to rent bikes for short trips and return them to any station throughout the city. However, the dynamic nature of user arrivals at each station leads to imbalances between bike supply and demand, resulting in unsatisfied users. An essential challenge lies in efficiently deploying and scheduling rebalancing vehicles for bike redistribution, as these decisions have a considerable effect on the efficiency of the system. To tackle this challenge, we propose a dynamic rebalancing model that integrates tactical and operational decisions within a single optimization framework. Unlike approaches that treat these decisions separately, our model captures the interaction between the two: in the first stage, it determines how many vehicles should be deployed over the planning horizon (tactical decision), and in the second stage, it assigns stations to dynamic rebalancing groups and allocates vehicles to these groups in response to demand realizations (operational decisions). To address the computational challenge, we propose two approaches: an Improved Integer L-shaped decomposition algorithm and a heuristic that combines machine learning with an early stopping criterion to estimate the second-stage cost function. Moreover, we generate forecasts of rental and return demand and incorporate them into the optimization model to enhance decision-making under demand uncertainty. Our numerical results show that the proposed heuristic is highly effective in minimizing the unsatisfied demand while reducing the computational costs efficiently.

自行车共享系统是一种重要的交通方式，使个人可以租用自行车进行短途旅行，并将其归还到城市的任何一个站点。然而，用户到达每个站点的动态特性导致自行车供需不平衡，导致用户不满意。一个重要的挑战在于有效地部署和调度再平衡车辆进行自行车再分配，因为这些决策对系统的效率有相当大的影响。为了应对这一挑战，我们提出了一个动态再平衡模型，该模型将战术和操作决策集成在一个单一的优化框架中。与单独处理这些决策的方法不同，我们的模型捕获了两者之间的相互作用：在第一阶段，它确定在规划范围内应该部署多少车辆（战术决策），在第二阶段，它将站点分配给动态再平衡组，并根据需求实现将车辆分配给这些组（操作决策）。为了解决计算上的挑战，我们提出了两种方法：一种改进的整数l形分解算法和一种启发式算法，该算法将机器学习与早期停止准则相结合，以估计第二阶段的成本函数。在此基础上，对租金和收益需求进行预测，并将其纳入优化模型，以增强需求不确定性下的决策能力。数值结果表明，所提出的启发式算法在最小化未满足需求的同时有效地降低了计算成本。

{"title":"What if rebalancing fleets could adapt? A two-stage stochastic model for dynamic bike redistribution","authors":"Mohammadreza Eslamipirharati , Maryam Motamedi , John Doucette , Nooshin Salari","doi":"10.1016/j.tre.2025.104640","DOIUrl":"10.1016/j.tre.2025.104640","url":null,"abstract":"<div><div>Bike-sharing systems are an important mode of transportation, enabling individuals to rent bikes for short trips and return them to any station throughout the city. However, the dynamic nature of user arrivals at each station leads to imbalances between bike supply and demand, resulting in unsatisfied users. An essential challenge lies in efficiently deploying and scheduling rebalancing vehicles for bike redistribution, as these decisions have a considerable effect on the efficiency of the system. To tackle this challenge, we propose a dynamic rebalancing model that integrates tactical and operational decisions within a single optimization framework. Unlike approaches that treat these decisions separately, our model captures the interaction between the two: in the first stage, it determines how many vehicles should be deployed over the planning horizon (tactical decision), and in the second stage, it assigns stations to dynamic rebalancing groups and allocates vehicles to these groups in response to demand realizations (operational decisions). To address the computational challenge, we propose two approaches: an Improved Integer L-shaped decomposition algorithm and a heuristic that combines machine learning with an early stopping criterion to estimate the second-stage cost function. Moreover, we generate forecasts of rental and return demand and incorporate them into the optimization model to enhance decision-making under demand uncertainty. Our numerical results show that the proposed heuristic is highly effective in minimizing the unsatisfied demand while reducing the computational costs efficiently.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104640"},"PeriodicalIF":8.8,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Impact Effects of Transport Structure Changes on Urban Traffic Congestion: A Case Study of Core Cities in China 交通结构变化对城市交通拥堵的影响——以中国核心城市为例

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-14 DOI: 10.1016/j.tre.2026.104676

Heng Chen, Naqi Yuan Liu, Yumei Yang, Yanying Wang, Qian Li, Yingji Shan

Accelerated urbanization and population expansion in developing countries have led to excessive reliance on road transport within logistics systems, exacerbating urban traffic congestion and constraining high-quality socioeconomic development. Consequently, optimizing urban logistics-transport structures becomes crucial for alleviating congestion and achieving sustainable economic growth. This study employs Chinese core cities as samples, adopting a comparative perspective on passenger/freight transport structure changes. Kernel density estimation and Markov chain models analyze spatiotemporal evolution patterns between logistics-transport restructuring and congestion, exploring their mechanistic relationships. Benchmark panel regressions and spatial econometric models provide empirical verification. Key findings include: (1) China’s urban logistics-transport structures demonstrate overall temporal optimization, specifically upward adjustments in passenger transport structures (rising ratios of railway-to-highway passenger volumes) and upward adjustments in freight transport structures (rising ratios of railway-to-highway freight volumes). Traffic congestion shows marginal mitigation. Changes in logistics-transport structures and spatial transitions of congestion remain relatively stable. China’s urban transportation exhibits a stepped imbalance pattern: “eastern superiority and central-western weakness in rising transport structures” alongside “eastern-central superiority and western inferiority in congestion alleviation”. (2) Nationally, absent spatial considerations, increased railway-to-highway passenger and freight volume ratios suppress congestion. When incorporating spatial factors, these ratio improvements alleviate local congestion while mitigating neighboring areas’ congestion. (3) Regionally without spatial effects, enhanced railway-passenger structures reduce congestion in eastern-western cities. Central cities’ inadequate transport network upgrades prevent full utilization of railway advantages, intensifying congestion. Improved railway-freight structures alleviate congestion in eastern-central cities, whereas western mountainous terrain constraints limit rail freight capacity, failing to meet growing demand and worsening congestion. With spatial effects, railway-freight improvements effectively reduce eastern-central congestion, while railway-passenger enhancements may exacerbate central cities’ congestion.

发展中国家城市化进程加快和人口扩张导致物流系统过度依赖道路运输，加剧了城市交通拥堵，制约了高质量的社会经济发展。因此，优化城市物流运输结构对于缓解拥堵和实现经济可持续增长至关重要。本研究以中国核心城市为样本，采用比较视角研究客货运输结构变化。核密度估计和马尔可夫链模型分析了物流运输重构与拥堵的时空演化规律，探讨了二者的机理关系。基准面板回归和空间计量模型提供了实证验证。主要发现包括：(1)中国城市物流运输结构呈现整体时间优化，特别是客运结构向上调整（铁路-公路客运量比上升）和货运结构向上调整（铁路-公路货运量比上升）。交通拥堵略有缓解。物流运输结构的变化和拥堵的空间转移保持相对稳定。中国城市交通呈现出“东优西弱”和“东中优西劣”的阶梯式不平衡格局。(2)在不考虑空间因素的情况下，铁路与公路客运量和货运量之比的提高抑制了交通拥堵。当纳入空间因素时，这些比率的改善缓解了局部拥堵，同时缓解了邻近地区的拥堵。③从区域上看，东西部城市的铁路客运结构增强对拥堵有一定的缓解作用，但不存在空间效应。中心城市交通网络升级不足，阻碍了铁路优势的充分利用，加剧了拥堵。铁路货运结构的改善缓解了东部中部城市的拥堵，而西部山区地形的限制限制了铁路货运能力，无法满足日益增长的需求，导致拥堵加剧。在空间效应上，铁路货运的改善可以有效缓解东中部城市的拥堵，而铁路客运的改善可能会加剧中部城市的拥堵。

{"title":"Impact Effects of Transport Structure Changes on Urban Traffic Congestion: A Case Study of Core Cities in China","authors":"Heng Chen, Naqi Yuan Liu, Yumei Yang, Yanying Wang, Qian Li, Yingji Shan","doi":"10.1016/j.tre.2026.104676","DOIUrl":"10.1016/j.tre.2026.104676","url":null,"abstract":"<div><div>Accelerated urbanization and population expansion in developing countries have led to excessive reliance on road transport within logistics systems, exacerbating urban traffic congestion and constraining high-quality socioeconomic development. Consequently, optimizing urban logistics-transport structures becomes crucial for alleviating congestion and achieving sustainable economic growth. This study employs Chinese core cities as samples, adopting a comparative perspective on passenger/freight transport structure changes. Kernel density estimation and Markov chain models analyze spatiotemporal evolution patterns between logistics-transport restructuring and congestion, exploring their mechanistic relationships. Benchmark panel regressions and spatial econometric models provide empirical verification. Key findings include: (1) China’s urban logistics-transport structures demonstrate overall temporal optimization, specifically upward adjustments in passenger transport structures (rising ratios of railway-to-highway passenger volumes) and upward adjustments in freight transport structures (rising ratios of railway-to-highway freight volumes). Traffic congestion shows marginal mitigation. Changes in logistics-transport structures and spatial transitions of congestion remain relatively stable. China’s urban transportation exhibits a stepped imbalance pattern: “eastern superiority and central-western weakness in rising transport structures” alongside “eastern-central superiority and western inferiority in congestion alleviation”. (2) Nationally, absent spatial considerations, increased railway-to-highway passenger and freight volume ratios suppress congestion. When incorporating spatial factors, these ratio improvements alleviate local congestion while mitigating neighboring areas’ congestion. (3) Regionally without spatial effects, enhanced railway-passenger structures reduce congestion in eastern-western cities. Central cities’ inadequate transport network upgrades prevent full utilization of railway advantages, intensifying congestion. Improved railway-freight structures alleviate congestion in eastern-central cities, whereas western mountainous terrain constraints limit rail freight capacity, failing to meet growing demand and worsening congestion. With spatial effects, railway-freight improvements effectively reduce eastern-central congestion, while railway-passenger enhancements may exacerbate central cities’ congestion.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104676"},"PeriodicalIF":8.8,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145961767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhancing ship collision risk assessment by integrating virtual ship-based shared nearest neighbor clustering and game-theoretic modeling 基于共享近邻聚类和博弈论建模的虚拟船舶碰撞风险评估方法

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-13 DOI: 10.1016/j.tre.2026.104675

Congcong Zhao , Zhuoyi Li , Tsz Leung Yip , Bing Wu

With the rapid growth in global shipping activities, the assessment of vessel collision risk has become a critical concern for maritime safety management. This study develops a comprehensive framework for identifying high-risk areas in congested waterways by integrating the Shared Nearest Neighbor Density-Based Spatial Clustering of Applications with Noise (SNN-DBSCAN) algorithm and Stackelberg game-theoretic model. The proposed framework first applies SNN-DBSCAN to enable robust waterway regionalization and vessel clustering under highly heterogeneous traffic densities, enhancing the accuracy and efficiency of collision risk assessment. To prevent critical crossing interactions from being fragmented by purely spatial clustering, we introduce a virtual-ship representation based on a risk-invariance principle, ensuring that high-risk encounters are preserved in the interaction set. Furthermore, a leader-follower game is employed to characterize strategic vessel responses by jointly considering safety, efficiency, and decision uncertainty to predict the next actions of the target vessel. The proposed framework is validated using empirical data from the busy waters of Hong Kong under daytime, nighttime, heavy precipitation, and strong winds. The results reveal pronounced scenario-dependent changes in vessel collision risk. Reduced nighttime visibility shifts hotspots and elevates risk, daytime port activities create new high-risk zones, and severe weather drives vessels to typhoon shelters where higher density and poorer maneuverability increase danger. The proposed approach captures these shifts and yields an interpretable, actionable tool for collision risk assessment and maritime traffic management, supporting future maritime safety management.

随着全球航运活动的快速增长，船舶碰撞风险评估已成为海上安全管理的一个重要问题。本研究通过整合基于共享最近邻密度的噪声应用空间聚类（SNN-DBSCAN）算法和Stackelberg博弈论模型，开发了一个综合框架，用于识别拥挤水道中的高风险区域。该框架首先应用SNN-DBSCAN实现了高度异构交通密度下稳健的航道区划和船舶聚类，提高了碰撞风险评估的准确性和效率。为了防止关键交叉交互被纯粹的空间聚类分割，我们引入了基于风险不变性原则的虚拟船表示，确保在交互集中保留高风险相遇。此外，通过联合考虑安全性、效率和决策不确定性来预测目标船舶的下一步行动，采用领导者-追随者博弈来表征船舶的战略反应。利用香港繁忙水域在白天、夜间、强降水和强风下的经验数据验证了所提出的框架。研究结果显示，船舶碰撞风险随场景变化而显著变化。夜间能见度降低会转移热点并增加风险，白天港口活动会产生新的高风险区域，恶劣天气会迫使船只前往密度更高、机动性更差的台风避难所，从而增加危险。拟议的方法抓住了这些变化，并为碰撞风险评估和海上交通管理提供了一种可解释、可操作的工具，为未来的海上安全管理提供支持。

{"title":"Enhancing ship collision risk assessment by integrating virtual ship-based shared nearest neighbor clustering and game-theoretic modeling","authors":"Congcong Zhao , Zhuoyi Li , Tsz Leung Yip , Bing Wu","doi":"10.1016/j.tre.2026.104675","DOIUrl":"10.1016/j.tre.2026.104675","url":null,"abstract":"<div><div>With the rapid growth in global shipping activities, the assessment of vessel collision risk has become a critical concern for maritime safety management. This study develops a comprehensive framework for identifying high-risk areas in congested waterways by integrating the Shared Nearest Neighbor Density-Based Spatial Clustering of Applications with Noise (SNN-DBSCAN) algorithm and Stackelberg game-theoretic model. The proposed framework first applies SNN-DBSCAN to enable robust waterway regionalization and vessel clustering under highly heterogeneous traffic densities, enhancing the accuracy and efficiency of collision risk assessment. To prevent critical crossing interactions from being fragmented by purely spatial clustering, we introduce a virtual-ship representation based on a risk-invariance principle, ensuring that high-risk encounters are preserved in the interaction set. Furthermore, a leader-follower game is employed to characterize strategic vessel responses by jointly considering safety, efficiency, and decision uncertainty to predict the next actions of the target vessel. The proposed framework is validated using empirical data from the busy waters of Hong Kong under daytime, nighttime, heavy precipitation, and strong winds. The results reveal pronounced scenario-dependent changes in vessel collision risk. Reduced nighttime visibility shifts hotspots and elevates risk, daytime port activities create new high-risk zones, and severe weather drives vessels to typhoon shelters where higher density and poorer maneuverability increase danger. The proposed approach captures these shifts and yields an interpretable, actionable tool for collision risk assessment and maritime traffic management, supporting future maritime safety management.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104675"},"PeriodicalIF":8.8,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145961769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

With whom to ally? Alliance strategy for EV battery supplier considering echelon utilization and disassembly recycling 与谁结盟？考虑梯次利用和拆解回收的电动汽车电池供应商联盟策略

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-12 DOI: 10.1016/j.tre.2025.104656

Zhangzhen Fang , Yuhan Guo , Gaoxiang Lou , Zhixuan Lai , Haicheng Ma , Li Zhou

The rapid expansion of the electric vehicle (EV) industry has heightened the need for sustainable and efficient closed-loop supply chains (CLSC) that can simultaneously improve economic returns and mitigate environmental impacts. To address this challenge, this study develops a game-theoretic model from the perspective of the power battery supplier and examines four inter-firm alliance modes: Non-alliance (N), supplier-manufacturer alliance (SM), supplier-recycler alliance (SR), and comprehensive alliance (SMR). The results reveal that (1) in the forward supply chain, suppliers under the SM and SMR modes consistently achieve higher battery capacity and EV sales. In the reverse supply chain, suppliers in alliance modes (SM, SR, SMR) are able to pay lower recycling prices while securing higher recycling quantities. (2) When recycling competition is weak, alliance with the manufacturer improves economic performance, whereas that with the recycler enhances environmental outcomes; however, the two benefits cannot be achieved simultaneously. By contrast, under intense recycling competition, forming a comprehensive alliance allows suppliers to improve both environmental and economic performance. (3) When extending the analysis to include suppliers’ investment in echelon utilization technology innovation, increased recycling competition intensity leads to a decline in the supplier’s echelon utilization performance, thereby amplifying the advantage of the comprehensive alliance.

电动汽车（EV）行业的快速扩张，提高了对可持续、高效的闭环供应链（CLSC）的需求，这种供应链可以同时提高经济回报和减轻环境影响。为了解决这一挑战，本文从动力电池供应商的角度建立了博弈论模型，并考察了四种企业间联盟模式：非联盟(N)、供应商-制造商联盟（SM）、供应商-回收商联盟（SR）和综合联盟（SMR）。结果表明：(1)在正向供应链中，SM模式和SMR模式下的供应商始终实现更高的电池容量和电动汽车销量。在逆向供应链中，联盟模式（SM、SR、SMR）的供应商能够支付较低的回收价格，同时获得较高的回收数量。(2)当回收竞争较弱时，与制造商联盟提高经济绩效，与回收商联盟提高环境绩效；然而，这两个好处不能同时实现。相比之下，在激烈的回收竞争中，形成一个全面的联盟可以使供应商提高环境和经济绩效。(3)将分析扩展到供应商在梯队利用技术创新方面的投入，回收竞争强度的增加导致供应商梯队利用绩效的下降，从而放大了综合联盟的优势。

{"title":"With whom to ally? Alliance strategy for EV battery supplier considering echelon utilization and disassembly recycling","authors":"Zhangzhen Fang , Yuhan Guo , Gaoxiang Lou , Zhixuan Lai , Haicheng Ma , Li Zhou","doi":"10.1016/j.tre.2025.104656","DOIUrl":"10.1016/j.tre.2025.104656","url":null,"abstract":"<div><div>The rapid expansion of the electric vehicle (EV) industry has heightened the need for sustainable and efficient closed-loop supply chains (CLSC) that can simultaneously improve economic returns and mitigate environmental impacts. To address this challenge, this study develops a game-theoretic model from the perspective of the power battery supplier and examines four inter-firm alliance modes: Non-alliance (N), supplier-manufacturer alliance (SM), supplier-recycler alliance (SR), and comprehensive alliance (SMR). The results reveal that (1) in the forward supply chain, suppliers under the SM and SMR modes consistently achieve higher battery capacity and EV sales. In the reverse supply chain, suppliers in alliance modes (SM, SR, SMR) are able to pay lower recycling prices while securing higher recycling quantities. (2) When recycling competition is weak, alliance with the manufacturer improves economic performance, whereas that with the recycler enhances environmental outcomes; however, the two benefits cannot be achieved simultaneously. By contrast, under intense recycling competition, forming a comprehensive alliance allows suppliers to improve both environmental and economic performance. (3) When extending the analysis to include suppliers’ investment in echelon utilization technology innovation, increased recycling competition intensity leads to a decline in the supplier’s echelon utilization performance, thereby amplifying the advantage of the comprehensive alliance.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104656"},"PeriodicalIF":8.8,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145956534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Algorithmic pricing in supply chains: implications for product quality, pricing, and profits 供应链中的算法定价：对产品质量、定价和利润的影响

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-12 DOI: 10.1016/j.tre.2026.104679

Kui Song , Jing Chen , Hui Yang , Bintong Chen , Honghu Huang

The rapid adoption of algorithmic pricing by retailers, enabled by big data analytics, is reshaping decisions in supply chains and affecting consumer surplus. We develop a game-theoretic model to examine how a retailer’s operation under an algorithmic-pricing regime, compared with a uniform-pricing regime, influences the manufacturer’s product quality and wholesale pricing decisions, as well as profits and consumer surplus. We uncover three key findings. First, algorithmic pricing affects product quality through two opposing effects: the demand segmentation effect, which encourages quality improvement by better matching products to heterogeneous consumers, and the profit compression effect, which discourages quality investment when the consumer distribution is highly skewed. Second, algorithmic pricing generates asymmetric profit impacts for supply chain members. While the retailer benefits more directly from pricing precision, both firms can benefit, particularly under a balanced mix of consumer types, through increased market coverage and reduced channel conflict. Third, when algorithmic reliability is high and consumer heterogeneity is moderate, algorithmic pricing can improve consumer surplus by aligning prices with willingness-to-pay and incentivizing higher quality. As reliability improves and the consumer distribution becomes more balanced, the system can achieve a tripartite win–win that benefits the manufacturer, the retailer, and consumers. These findings highlight the dual, condition-dependent role of algorithmic pricing as both a coordination tool and a quality-enhancement mechanism in supply chains. They also offer managerial implications for the strategic deployment of algorithmic pricing tools and inform policy debates on regulating algorithm-driven markets.

在大数据分析的推动下，零售商迅速采用算法定价，正在重塑供应链决策，并影响消费者剩余。我们开发了一个博弈论模型来研究零售商在算法定价制度下的运作，与统一定价制度相比，如何影响制造商的产品质量和批发定价决策，以及利润和消费者剩余。我们发现了三个关键发现。首先，算法定价通过两种相反的效应影响产品质量：一是需求细分效应，通过更好地将产品与异质消费者匹配来鼓励质量提高；二是利润压缩效应，当消费者分布高度倾斜时，利润压缩效应阻碍质量投资。其次，算法定价对供应链成员产生不对称的利润影响。虽然零售商更直接地受益于定价的准确性，但两家公司都可以受益，特别是在消费者类型的平衡组合下，通过增加市场覆盖和减少渠道冲突。第三，当算法可靠性高且消费者异质性适中时，算法定价可以通过调整价格与支付意愿和激励更高质量来提高消费者剩余。随着可靠性的提高和消费者分布的更加平衡，该系统可以实现制造商、零售商和消费者三方共赢。这些发现突出了算法定价作为供应链中协调工具和质量提升机制的双重、条件依赖的作用。它们还为算法定价工具的战略部署提供了管理意义，并为规范算法驱动市场的政策辩论提供了信息。

{"title":"Algorithmic pricing in supply chains: implications for product quality, pricing, and profits","authors":"Kui Song , Jing Chen , Hui Yang , Bintong Chen , Honghu Huang","doi":"10.1016/j.tre.2026.104679","DOIUrl":"10.1016/j.tre.2026.104679","url":null,"abstract":"<div><div>The rapid adoption of algorithmic pricing by retailers, enabled by big data analytics, is reshaping decisions in supply chains and affecting consumer surplus. We develop a game-theoretic model to examine how a retailer’s operation under an algorithmic-pricing regime, compared with a uniform-pricing regime, influences the manufacturer’s product quality and wholesale pricing decisions, as well as profits and consumer surplus. We uncover three key findings. First, algorithmic pricing affects product quality through two opposing effects: the demand segmentation effect, which encourages quality improvement by better matching products to heterogeneous consumers, and the profit compression effect, which discourages quality investment when the consumer distribution is highly skewed. Second, algorithmic pricing generates asymmetric profit impacts for supply chain members. While the retailer benefits more directly from pricing precision, both firms can benefit, particularly under a balanced mix of consumer types, through increased market coverage and reduced channel conflict. Third, when algorithmic reliability is high and consumer heterogeneity is moderate, algorithmic pricing can improve consumer surplus by aligning prices with willingness-to-pay and incentivizing higher quality. As reliability improves and the consumer distribution becomes more balanced, the system can achieve a tripartite win–win that benefits the manufacturer, the retailer, and consumers. These findings highlight the dual, condition-dependent role of algorithmic pricing as both a coordination tool and a quality-enhancement mechanism in supply chains. They also offer managerial implications for the strategic deployment of algorithmic pricing tools and inform policy debates on regulating algorithm-driven markets.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104679"},"PeriodicalIF":8.8,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145956533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A risk-averse two-stage stochastic programming model for vessel schedule recovery in liner shipping service 班轮运输船舶进度恢复的风险规避两阶段随机规划模型

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-10 DOI: 10.1016/j.tre.2025.104655

Shuaiqi Zhao , Hualong Yang , Yadong Wang , Zaili Yang

Significant delays caused by disruption events, coupled with regular uncertainties, pose challenges to risk avoidance and vessel schedule recovery problem (RA-VSRP) in liner container shipping services. To address this, we propose a new optimization framework that incorporates a hybrid risk aversion measure with three recovery strategies, including sailing speed adjustment, port skipping, and transshipment. The framework systematically combines ex-ante decision-making and in-progress decision-making. The former helps shorten vessel schedule recovery time and costs by quickly responding to disruption events, while the latter improves the flexibility of selecting vessel schedule recovery strategies. By adopting a scenario-based approach to jointly capture regular uncertainties and disruption events, RA-VSRP is formulated as a chance-constrained two-stage stochastic programming model, where conditional value-at-risk (CVaR) is used as the risk measure. An exact Benders decomposition-based branch-and-cut algorithm is employed to efficiently solve the computationally challenging model. We develop two algorithmic variants based on alternative representations of CVaR. Extensive numerical experiments demonstrate the applicability of the model and the computational efficiency of the algorithm. The results show that the proposed framework can provide reliable vessel schedule recovery solutions through sailing speed adjustments, port skipping, and transshipment. The findings provide managerial insights for shipping companies regarding schedule recovery, risk aversion, and cost control.

在集装箱班轮运输服务中，由于中断事件造成的重大延误，加上常规的不确定性，对风险规避和船舶进度恢复问题（RA-VSRP）提出了挑战。为了解决这个问题，我们提出了一个新的优化框架，该框架结合了混合风险规避措施和三种恢复策略，包括航行速度调整、港口跳过和转运。该框架系统地将事前决策与事前决策相结合。前者通过快速响应中断事件，有助于缩短船舶调度恢复时间和成本，而后者提高了选择船舶调度恢复策略的灵活性。通过采用基于场景的方法来联合捕获常规不确定性和中断事件，RA-VSRP被制定为机会约束的两阶段随机规划模型，其中条件风险值（CVaR）被用作风险度量。采用基于精确Benders分解的分支切断算法有效地解决了计算困难的模型。我们基于CVaR的替代表示开发了两个算法变体。大量的数值实验证明了该模型的适用性和算法的计算效率。结果表明，该框架可以通过航速调整、跳港和转运提供可靠的船舶调度恢复方案。研究结果为航运公司提供了关于进度恢复、风险规避和成本控制的管理见解。

{"title":"A risk-averse two-stage stochastic programming model for vessel schedule recovery in liner shipping service","authors":"Shuaiqi Zhao , Hualong Yang , Yadong Wang , Zaili Yang","doi":"10.1016/j.tre.2025.104655","DOIUrl":"10.1016/j.tre.2025.104655","url":null,"abstract":"<div><div>Significant delays caused by disruption events, coupled with regular uncertainties, pose challenges to risk avoidance and vessel schedule recovery problem (RA-VSRP) in liner container shipping services. To address this, we propose a new optimization framework that incorporates a hybrid risk aversion measure with three recovery strategies, including sailing speed adjustment, port skipping, and transshipment. The framework systematically combines ex-ante decision-making and in-progress decision-making. The former helps shorten vessel schedule recovery time and costs by quickly responding to disruption events, while the latter improves the flexibility of selecting vessel schedule recovery strategies. By adopting a scenario-based approach to jointly capture regular uncertainties and disruption events, RA-VSRP is formulated as a chance-constrained two-stage stochastic programming model, where conditional value-at-risk (CVaR) is used as the risk measure. An exact Benders decomposition-based branch-and-cut algorithm is employed to efficiently solve the computationally challenging model. We develop two algorithmic variants based on alternative representations of CVaR. Extensive numerical experiments demonstrate the applicability of the model and the computational efficiency of the algorithm. The results show that the proposed framework can provide reliable vessel schedule recovery solutions through sailing speed adjustments, port skipping, and transshipment. The findings provide managerial insights for shipping companies regarding schedule recovery, risk aversion, and cost control.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104655"},"PeriodicalIF":8.8,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145956562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Who should invest in EV charging infrastructure? Policy design under ZEV mandates 谁应该投资电动汽车充电基础设施？ZEV授权下的政策设计

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-10 DOI: 10.1016/j.tre.2025.104642

Ting Chen , Kannan Govindan , Wenting Yang , Qiuwei Wang , Lu Liu

Governments worldwide have implemented multiple incentive measures to promote electric vehicle (EV) adoption, including consumer subsidies, infrastructure investment subsidies, and Zero Emission Vehicle (ZEV) mandates. Motivated by these policy instruments, various stakeholders (including automakers and EV component suppliers) are actively investing in charging infrastructure. Which subsidy policy is most cost-effective, and which entity’s infrastructure investment yields optimal outcomes? This study constructs a Stackelberg game model comprising government, component supplier, and competing fuel/electric vehicle manufacturers to explore optimal policy structures under three investment modes (i.e., supplier-led infrastructure investment, EV manufacturer-led investment, and traditional automaker-led investment). We systematically evaluate the performance in minimizing government expenditure while achieving adoption targets for each investment mode. More importantly, we extend our analysis to include ZEV mandates and their impact on subsidy effectiveness and government expenditure. Our results provide several interesting and counterintuitive insights. First, we show that supplier-led infrastructure investment (i.e., Mode S) consistently achieves the lowest policy expenditure and highest social welfare. Second, when automakers invest (Modes E and F), consumer subsidies alone are never optimal, contradicting widespread policy practice. Mode S consistently achieves the lowest expenditure and highest social welfare, yet Modes E and F generate superior economic benefits for firms—revealing a fundamental trade-off between public cost-efficiency and private profitability. Third, although ZEV mandates are intended to substitute for costly subsidies and reduce fiscal burden, our analysis reveals they may paradoxically increase government expenditure and discourage EV production when requirements become overly stringent.

世界各国政府已经实施了多种激励措施来促进电动汽车（EV）的采用，包括消费者补贴、基础设施投资补贴和零排放汽车（ZEV）授权。在这些政策工具的推动下，各种利益相关者（包括汽车制造商和电动汽车零部件供应商）都在积极投资充电基础设施。哪种补贴政策最具成本效益，哪个实体的基础设施投资产生最优结果？本文构建了由政府、零部件供应商和燃油/电动汽车竞争厂商组成的Stackelberg博弈模型，探讨了供应商主导基础设施投资、电动汽车制造商主导投资和传统汽车制造商主导投资三种投资模式下的最优政策结构。我们系统地评估了在实现每种投资模式的采用目标的同时最小化政府支出的表现。更重要的是，我们扩展了我们的分析，包括ZEV要求及其对补贴有效性和政府支出的影响。我们的研究结果提供了一些有趣的、违反直觉的见解。首先，我们证明了供应商主导的基础设施投资（即模式S）始终实现最低的政策支出和最高的社会福利。其次，当汽车制造商投资（模式E和模式F）时，仅靠消费者补贴从来都不是最优的，这与广泛的政策实践相矛盾。模式S始终实现最低的支出和最高的社会福利，而模式E和模式F为企业带来了更大的经济效益——揭示了公共成本效率和私人盈利能力之间的基本权衡。第三，尽管ZEV法规旨在取代昂贵的补贴并减轻财政负担，但我们的分析表明，当要求过于严格时，它们可能会矛盾地增加政府支出并阻碍电动汽车生产。

{"title":"Who should invest in EV charging infrastructure? Policy design under ZEV mandates","authors":"Ting Chen , Kannan Govindan , Wenting Yang , Qiuwei Wang , Lu Liu","doi":"10.1016/j.tre.2025.104642","DOIUrl":"10.1016/j.tre.2025.104642","url":null,"abstract":"<div><div>Governments worldwide have implemented multiple incentive measures to promote electric vehicle (EV) adoption, including consumer subsidies, infrastructure investment subsidies, and Zero Emission Vehicle (ZEV) mandates. Motivated by these policy instruments, various stakeholders (including automakers and EV component suppliers) are actively investing in charging infrastructure. Which subsidy policy is most cost-effective, and which entity’s infrastructure investment yields optimal outcomes? This study constructs a Stackelberg game model comprising government, component supplier, and competing fuel/electric vehicle manufacturers to explore optimal policy structures under three investment modes (i.e., supplier-led infrastructure investment, EV manufacturer-led investment, and traditional automaker-led investment). We systematically evaluate the performance in minimizing government expenditure while achieving adoption targets for each investment mode. More importantly, we extend our analysis to include ZEV mandates and their impact on subsidy effectiveness and government expenditure. Our results provide several interesting and counterintuitive insights. First, we show that supplier-led infrastructure investment (i.e., Mode S) consistently achieves the lowest policy expenditure and highest social welfare. Second, when automakers invest (Modes E and F), consumer subsidies alone are never optimal, contradicting widespread policy practice. Mode S consistently achieves the lowest expenditure and highest social welfare, yet Modes E and F generate superior economic benefits for firms—revealing a fundamental trade-off between public cost-efficiency and private profitability. Third, although ZEV mandates are intended to substitute for costly subsidies and reduce fiscal burden, our analysis reveals they may paradoxically increase government expenditure and discourage EV production when requirements become overly stringent.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104642"},"PeriodicalIF":8.8,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep reinforcement learning for the vehicle routing problem with route balancing 基于深度强化学习的车辆路径平衡问题

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-10 DOI: 10.1016/j.tre.2025.104632

Jianhua Xiao , Detian Kong , Zhiguang Cao , Jingyi Zhao

The Vehicle Routing Problem with Route Balancing (VRPRB) is a multi-objective combinatorial optimization (MOCO) problem that aims to balance workload distribution while minimizing overall travel costs. Unlike traditional Vehicle Routing Problems (VRP), VRPRB introduces fleet size constraints to improve resource utilization and reduce operational costs. Existing deep reinforcement learning (DRL) approaches for VRP rarely address multi-objective optimization and often assume an unlimited fleet size, limiting their practical applicability. To address this issue, we propose an Equity Attention Model (E-AM), a problem-tailored DRL framework designed to generate Pareto-optimal solutions for VRPRB. Our E-AM formulates the problem as a sequential decision-making process, where each decision involves pairing a vehicle with a customer. E-AM is built on an attention-based architecture, incorporating a node encoder, a vehicle context encoder, and a decoder, with a hyper-network technique to efficiently handle multi-objective optimization. Experimental results demonstrate that our approach finds better solutions than current state-of-the-art methods on VRPRB benchmark instances while maintaining higher computational efficiency. By fully leveraging the strengths of deep reinforcement learning, our approach provides a scalable and adaptive alternative to traditional heuristic and exact algorithms, achieving high-quality solutions for complex real-world routing problems. To enhance scalability and training efficiency, we introduce a two-stage reinforcement learning strategy that enables E-AM to solve VRPRB instances with up to 1000 customers. To promote transparency and reproducibility, we have open-sourced our implementation¹.

具有路径平衡的车辆路径问题（VRPRB）是一个多目标组合优化（MOCO）问题，其目标是平衡工作负载分配，同时使总体出行成本最小化。与传统的车辆路径问题（VRP）不同，VRPRB引入了车队规模限制，以提高资源利用率并降低运营成本。现有的深度强化学习（DRL） VRP方法很少解决多目标优化问题，并且通常假设无限的车队规模，限制了它们的实际适用性。为了解决这个问题，我们提出了一个公平注意模型（E-AM），这是一个针对问题定制的DRL框架，旨在为VRPRB生成帕累托最优解决方案。我们的E-AM将问题表述为一个连续的决策过程，其中每个决策都涉及将车辆与客户配对。E-AM建立在基于注意力的架构上，结合节点编码器、车辆上下文编码器和解码器，采用超网络技术有效处理多目标优化。实验结果表明，我们的方法在VRPRB基准实例上找到了比目前最先进的方法更好的解决方案，同时保持了更高的计算效率。通过充分利用深度强化学习的优势，我们的方法为传统的启发式和精确算法提供了可扩展和自适应的替代方案，为复杂的现实世界路由问题提供了高质量的解决方案。为了提高可扩展性和训练效率，我们引入了一种两阶段强化学习策略，使E-AM能够解决多达1000个客户的VRPRB实例。为了提高透明度和可重复性，我们已经开源了我们的实现。

{"title":"Deep reinforcement learning for the vehicle routing problem with route balancing","authors":"Jianhua Xiao , Detian Kong , Zhiguang Cao , Jingyi Zhao","doi":"10.1016/j.tre.2025.104632","DOIUrl":"10.1016/j.tre.2025.104632","url":null,"abstract":"<div><div>The Vehicle Routing Problem with Route Balancing (VRPRB) is a multi-objective combinatorial optimization (MOCO) problem that aims to balance workload distribution while minimizing overall travel costs. Unlike traditional Vehicle Routing Problems (VRP), VRPRB introduces fleet size constraints to improve resource utilization and reduce operational costs. Existing deep reinforcement learning (DRL) approaches for VRP rarely address multi-objective optimization and often assume an unlimited fleet size, limiting their practical applicability. To address this issue, we propose an Equity Attention Model (E-AM), a problem-tailored DRL framework designed to generate Pareto-optimal solutions for VRPRB. Our E-AM formulates the problem as a sequential decision-making process, where each decision involves pairing a vehicle with a customer. E-AM is built on an attention-based architecture, incorporating a node encoder, a vehicle context encoder, and a decoder, with a hyper-network technique to efficiently handle multi-objective optimization. Experimental results demonstrate that our approach finds better solutions than current state-of-the-art methods on VRPRB benchmark instances while maintaining higher computational efficiency. By fully leveraging the strengths of deep reinforcement learning, our approach provides a scalable and adaptive alternative to traditional heuristic and exact algorithms, achieving high-quality solutions for complex real-world routing problems. To enhance scalability and training efficiency, we introduce a two-stage reinforcement learning strategy that enables E-AM to solve VRPRB instances with up to 1000 customers. To promote transparency and reproducibility, we have open-sourced our implementation<span><span><sup>1</sup></span></span>.</div></div>","PeriodicalId":49418,"journal":{"name":"Transportation Research Part E-Logistics and Transportation Review","volume":"208 ","pages":"Article 104632"},"PeriodicalIF":8.8,"publicationDate":"2026-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Corrigendum to “Cost allocation in a robust two-stage resource allocation game: fairness and robustness”. [Trans. Res. Part E: Logist. Trans. Rev. 207 (2026) 104633] “稳健的两阶段资源分配博弈中的成本分配：公平与稳健”的勘误表。(反式。答：E部分：医生。反式。Rev. 207 (2026) 104633]

IF 8.8 1区工程技术 Q1 ECONOMICS

Transportation Research Part E-Logistics and Transportation Review

Pub Date : 2026-01-10 DOI: 10.1016/j.tre.2026.104680

Menghang Wang, Lan Lu, Lindong Liu, Jie Wu

引用次数: 0