Communications in Transportation Research最新文献

Machine learning-based real-time crash risk forecasting for pedestrians 基于机器学习的行人实时碰撞风险预测

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-12-01 DOI: 10.1016/j.commtr.2025.100224

Fizza Hussain , Yuefeng Li , Shimul Md Mazharul Haque

Recent developments in artificial intelligence (AI) have made significant improvements in understanding and enhancing pedestrian safety—a vulnerable road user group that receives less attention than motorized road users do. Specifically, AI-based video analytics have provided insight into facilitating real-time safety at signalized intersections. However, past studies have not fully realized the essence of real-time analysis, which underpins forecasting pedestrian collision likelihood by analyzing how past extreme events influence future risk over sequential intervals. To this end, we combine extreme value theory and machine learning models for real-time pedestrian collision risk forecasting. Traffic conflicts and their associated variables were identified from 288 h of video footage obtained from three signalized intersections in Queensland, Australia, via computer vision techniques, including YOLO and DeepSORT, to obtain the post encroachment time for vehicle‒pedestrian interactions. A Bayesian non-stationary peak over threshold (POT) is developed to obtain real-time pedestrian crash risk at the signal cycle level. The performance of the POT model is compared with observed crashes, and the results demonstrate the reasonable accuracy of the model. The estimated pedestrian crash risk at each signal cycle forms contiguous univariate time series data (which serve as ground truth), which are used as input to develop time series machine learning models (recurrent neural networks (RNNs) and long short-term memory (LSTM)). Both of these models forecast pedestrian crash risk, with the RNN model outperforming the competing model and demonstrating that pedestrian crash risk can be reliably estimated 30−33 min in advance.

人工智能（AI）的最新发展在理解和加强行人安全方面取得了重大进展，行人是一个弱势的道路使用者群体，与机动道路使用者相比，他们受到的关注较少。具体来说，基于人工智能的视频分析为促进信号交叉口的实时安全提供了见解。然而，过去的研究并没有充分认识到实时分析的本质，实时分析是通过分析过去的极端事件对未来风险的影响来预测行人碰撞可能性的基础。为此，我们将极值理论与机器学习模型相结合，进行实时行人碰撞风险预测。利用计算机视觉技术（包括YOLO和DeepSORT），从澳大利亚昆士兰州三个信号交叉口的288 h视频片段中识别交通冲突及其相关变量，获得车辆与行人相互作用的侵占后时间。提出了一种贝叶斯非平稳超阈值峰值（POT）方法，在信号周期水平上实时获取行人碰撞风险。将POT模型的性能与实际碰撞进行了比较，结果证明了该模型的合理精度。每个信号周期估计的行人碰撞风险形成连续的单变量时间序列数据（作为基础事实），这些数据用作开发时间序列机器学习模型（循环神经网络（rnn）和长短期记忆（LSTM））的输入。这两种模型都预测行人碰撞风险，其中RNN模型优于竞争模型，并证明行人碰撞风险可以提前30 - 33分钟可靠地估计。

{"title":"Machine learning-based real-time crash risk forecasting for pedestrians","authors":"Fizza Hussain , Yuefeng Li , Shimul Md Mazharul Haque","doi":"10.1016/j.commtr.2025.100224","DOIUrl":"10.1016/j.commtr.2025.100224","url":null,"abstract":"<div><div>Recent developments in artificial intelligence (AI) have made significant improvements in understanding and enhancing pedestrian safety—a vulnerable road user group that receives less attention than motorized road users do. Specifically, AI-based video analytics have provided insight into facilitating real-time safety at signalized intersections. However, past studies have not fully realized the essence of real-time analysis, which underpins forecasting pedestrian collision likelihood by analyzing how past extreme events influence future risk over sequential intervals. To this end, we combine extreme value theory and machine learning models for real-time pedestrian collision risk forecasting. Traffic conflicts and their associated variables were identified from 288 h of video footage obtained from three signalized intersections in Queensland, Australia, via computer vision techniques, including YOLO and DeepSORT, to obtain the post encroachment time for vehicle‒pedestrian interactions. A Bayesian non-stationary peak over threshold (POT) is developed to obtain real-time pedestrian crash risk at the signal cycle level. The performance of the POT model is compared with observed crashes, and the results demonstrate the reasonable accuracy of the model. The estimated pedestrian crash risk at each signal cycle forms contiguous univariate time series data (which serve as ground truth), which are used as input to develop time series machine learning models (recurrent neural networks (RNNs) and long short-term memory (LSTM)). Both of these models forecast pedestrian crash risk, with the RNN model outperforming the competing model and demonstrating that pedestrian crash risk can be reliably estimated 30−33 min in advance.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100224"},"PeriodicalIF":14.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145617648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MoTIF: An end-to-end multimodal road traffic scene understanding foundation model 主题：一个端到端的多模式道路交通场景理解基础模型

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-12-01 DOI: 10.1016/j.commtr.2025.100227

Zihe Wang , Haiyang Yu , Changxin Chen , Zhiyong Cui , Yufeng Bi , Yilong Ren , Zijian Wang , Delan Kong , Jing Tian , Shoutong Yuan , Zhiqiang Li

Video-based road intelligent detection constitutes a critical component in modern intelligent transportation systems, serving as a crucial role for comprehensive transportation planning and emergency traffic management. Current traffic scene perception methodologies relying on conventional deep learning architectures present inherent limitations, including heavy dependence on extensive manual annotations of specific traffic scenarios and predefined rule configurations. These approaches demonstrate constrained semantic representation capacity and limited generalizability across heterogeneous traffic scenarios. To address these challenges, this study proposes a novel end-to-end multimodal foundation model architecture that jointly generates dynamic traffic event detection outcomes and semantic-rich contextual descriptions. Through integration of low-rank adaptation (LoRA) and prompt fine-tuning as parameter-efficient fine-tuning strategies, we develop the multimodal road traffic scene understanding foundation model (MoTIF), which establishes cross-modal alignment between visual patterns and textual semantics. This framework demonstrates enhanced capability in extracting salient traffic targets and generating hierarchical scene representations, significantly improving automated detection efficiency in road video analytics. Notably, MoTIF exhibits contextual reasoning capabilities for implicit traffic event interpretation. Extensive evaluations on two real-world datasets encompassing urban road intersection scenarios in Tianjin and highway monitoring systems in Shandong Province reveal that MoTIF achieves superior performance metrics: 65.81 average score on multimodal scene understanding assessment and 83.33% event detection accuracy, outperforming mainstream benchmarks in both precision and computational efficiency. This research advances multimodal learning paradigms for intelligent transportation systems while providing practical insights for adaptive traffic management applications.

基于视频的道路智能检测是现代智能交通系统的重要组成部分，在综合交通规划和应急交通管理中发挥着重要作用。当前依赖于传统深度学习架构的交通场景感知方法存在固有的局限性，包括严重依赖于特定交通场景的大量手动注释和预定义的规则配置。这些方法在异构流量场景中表现出有限的语义表示能力和有限的泛化能力。为了应对这些挑战，本研究提出了一种新颖的端到端多模态基础模型架构，该架构可共同生成动态交通事件检测结果和语义丰富的上下文描述。通过将低阶自适应（low-rank adaptation, LoRA）和快速微调（prompt fine-tuning）作为参数高效的微调策略，我们开发了多模式道路交通场景理解基础模型（MoTIF），该模型建立了视觉模式和文本语义之间的跨模式对齐。该框架在提取显著交通目标和生成分层场景表示方面表现出增强的能力，显著提高了道路视频分析中的自动检测效率。值得注意的是，MoTIF展示了隐性交通事件解释的上下文推理能力。通过对天津市城市道路交叉口场景和山东省公路监控系统两个真实数据集的广泛评估，MoTIF实现了卓越的性能指标：多模式场景理解评估平均分65.81分，事件检测准确率83.33%，在精度和计算效率方面均优于主流基准。本研究推进了智能交通系统的多模式学习范式，同时为自适应交通管理应用提供了实践见解。

{"title":"MoTIF: An end-to-end multimodal road traffic scene understanding foundation model","authors":"Zihe Wang , Haiyang Yu , Changxin Chen , Zhiyong Cui , Yufeng Bi , Yilong Ren , Zijian Wang , Delan Kong , Jing Tian , Shoutong Yuan , Zhiqiang Li","doi":"10.1016/j.commtr.2025.100227","DOIUrl":"10.1016/j.commtr.2025.100227","url":null,"abstract":"<div><div>Video-based road intelligent detection constitutes a critical component in modern intelligent transportation systems, serving as a crucial role for comprehensive transportation planning and emergency traffic management. Current traffic scene perception methodologies relying on conventional deep learning architectures present inherent limitations, including heavy dependence on extensive manual annotations of specific traffic scenarios and predefined rule configurations. These approaches demonstrate constrained semantic representation capacity and limited generalizability across heterogeneous traffic scenarios. To address these challenges, this study proposes a novel end-to-end multimodal foundation model architecture that jointly generates dynamic traffic event detection outcomes and semantic-rich contextual descriptions. Through integration of low-rank adaptation (LoRA) and prompt fine-tuning as parameter-efficient fine-tuning strategies, we develop the multimodal road traffic scene understanding foundation model (MoTIF), which establishes cross-modal alignment between visual patterns and textual semantics. This framework demonstrates enhanced capability in extracting salient traffic targets and generating hierarchical scene representations, significantly improving automated detection efficiency in road video analytics. Notably, MoTIF exhibits contextual reasoning capabilities for implicit traffic event interpretation. Extensive evaluations on two real-world datasets encompassing urban road intersection scenarios in Tianjin and highway monitoring systems in Shandong Province reveal that MoTIF achieves superior performance metrics: 65.81 average score on multimodal scene understanding assessment and 83.33% event detection accuracy, outperforming mainstream benchmarks in both precision and computational efficiency. This research advances multimodal learning paradigms for intelligent transportation systems while providing practical insights for adaptive traffic management applications.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100227"},"PeriodicalIF":14.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145736818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Quantitative assessment of mid-air collision probability in urban air mobility: A safety barrier-based framework for integrated operations 城市空中机动中空中碰撞概率的定量评估：基于安全屏障的综合作战框架

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-12-01 DOI: 10.1016/j.commtr.2025.100230

Jinpeng Zhang , Yan Xu , Kaiquan Cai , Victor Gordo , Gokhan Inalhan

Risk assessment is a key issue when handling the increasing use of unmanned aircraft systems (UASs), especially in integrated operational urban airspace. This study proposes a method to systematically quantify the mid-air collision (MAC) risk for different operation types in urban air mobility (UAM). Specifically, a novel assessment framework is formulated that includes three timelines to track risk evolution for integrated operations (i.e., cooperative, non-cooperative, and manned aircraft) and four safety barriers (i.e., procedural, strategic, tactical, and collision avoidance mitigation) to prevent a conflict ending in MAC. Under the framework, analytical models are established considering trajectory uncertainty, latency time, and detection and avoidance (DAA) system risk ratio to calculate the failure probability of each barrier, and thus an overall MAC probability. Result-oriented, process-oriented and comprehensive Monte Carlo simulations are constructed to validate the proposed models and the MAC assessment timeline, followed by demonstrating four operational scenarios in real-world environment to illustrate the assessment process of the method. Results show that the simulation probability curves closely match the theoretical predictions. The cooperative UAS contributes the highest MAC risk in our designed integrated environment, with strategic mitigation failures accounting for the largest proportion, and thus effective strategic trajectory planning is crucial for maintaining the safety of integrated operations.

风险评估是处理无人机系统（UASs）日益增加的使用时的一个关键问题，特别是在综合作战城市空域。本文提出了一种系统量化城市空中交通不同运行类型的空中碰撞风险的方法。具体而言，本文制定了一个新的评估框架，其中包括跟踪综合作战（即合作、非合作和有人驾驶飞机）的风险演变的三个时间表，以及防止冲突以MAC结束的四个安全屏障（即程序、战略、战术和避撞缓解）。在该框架下，建立了考虑轨迹不确定性、延迟时间、和检测与规避（DAA）系统风险比来计算每个屏障的失效概率，从而得到总体MAC概率。构建了以结果为导向、以过程为导向的综合蒙特卡罗仿真，验证了所提出的模型和MAC评估时间表，并在实际环境中演示了四种操作场景，以说明该方法的评估过程。结果表明，仿真概率曲线与理论预测吻合较好。在我们设计的综合环境中，合作型无人机系统造成的MAC风险最高，其中战略缓解故障所占比例最大，因此有效的战略轨迹规划对于维持综合作战安全至关重要。

{"title":"Quantitative assessment of mid-air collision probability in urban air mobility: A safety barrier-based framework for integrated operations","authors":"Jinpeng Zhang , Yan Xu , Kaiquan Cai , Victor Gordo , Gokhan Inalhan","doi":"10.1016/j.commtr.2025.100230","DOIUrl":"10.1016/j.commtr.2025.100230","url":null,"abstract":"<div><div>Risk assessment is a key issue when handling the increasing use of unmanned aircraft systems (UASs), especially in integrated operational urban airspace. This study proposes a method to systematically quantify the mid-air collision (MAC) risk for different operation types in urban air mobility (UAM). Specifically, a novel assessment framework is formulated that includes three timelines to track risk evolution for integrated operations (i.e., cooperative, non-cooperative, and manned aircraft) and four safety barriers (i.e., procedural, strategic, tactical, and collision avoidance mitigation) to prevent a conflict ending in MAC. Under the framework, analytical models are established considering trajectory uncertainty, latency time, and detection and avoidance (DAA) system risk ratio to calculate the failure probability of each barrier, and thus an overall MAC probability. Result-oriented, process-oriented and comprehensive Monte Carlo simulations are constructed to validate the proposed models and the MAC assessment timeline, followed by demonstrating four operational scenarios in real-world environment to illustrate the assessment process of the method. Results show that the simulation probability curves closely match the theoretical predictions. The cooperative UAS contributes the highest MAC risk in our designed integrated environment, with strategic mitigation failures accounting for the largest proportion, and thus effective strategic trajectory planning is crucial for maintaining the safety of integrated operations.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100230"},"PeriodicalIF":14.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145683860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Interpretable machine learning for traffic congestion prediction: Unveiling the impact of different COVID-19 periods 交通拥堵预测的可解释机器学习：揭示不同COVID-19时期的影响

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-12-01 DOI: 10.1016/j.commtr.2025.100226

Dan Zhu , ChiSin Ng , Litian Xie , Yang Liu

Traffic congestion prediction plays a crucial role in mitigating congestion. However, the COVID-19 pandemic and associated government control measures have significantly altered urban travel behavior, increasing the complexity of traffic congestion prediction. This study aims to predict traffic congestion in Alameda County in the San Francisco Bay Area, USA, during the prelockdown, lockdown, and postlockdown periods. We incorporate three external categories of data, i.e., weather conditions, seasonality factors, and COVID-19-related variables, and use recursive feature elimination with cross-validation to identify important features across different periods and avoid potential overfitting. On this basis, multiple advanced machine learning (ML) models, including support vector regression (SVR), multiple linear regression (MLR), recurrent neural network (RNN), and long short-term memory (LSTM) networks, are trained and optimized through extensive experimentation and parameter tuning. Since LSTM has more hyperparameters and is more sensitive to tuning than the other ML methods used, we employ an adaptive parameter selection approach to optimize its hyperparameters, enhancing model accuracy and efficiency, rather than manually tuning parameters for SVR and RNN. These models are evaluated via the normalized root mean square error. The results indicate that the bidirectional LSTM (Bi-LSTM) consistently outperforms the other models across all COVID-19 periods. This superior performance can be attributed to the Bi-LSTM's bidirectional architecture, which effectively captures temporal dependencies by analyzing data both forward and backward in time. To address the limited interpretability of ML methods and provide valuable insights, we apply the integrated gradient (IG) technique to interpret the best-performing and differentiable Bi-LSTM predictions. Our analysis revealed that new COVID-19 cases had a negative influence on traffic congestion during the lockdown and postlockdown periods. The observed reduction in traffic can be explained by heightened public risk awareness, voluntary reductions in travel, and compliance with government-imposed mobility restrictions. We also apply SHapley Additive exPlanations to SVR, given that IG is not applicable to this model. The results indicate that in the postpandemic period, people have become more cautious—high new hospitalization discourages travel, reducing traffic congestion, whereas high fuel prices do not deter a shift toward private vehicle use, leading to increased congestion.

交通拥堵预测在缓解交通拥堵中起着至关重要的作用。然而，新冠肺炎大流行和相关政府控制措施显著改变了城市出行行为，增加了交通拥堵预测的复杂性。本研究旨在预测美国旧金山湾区阿拉米达县在封城前、封城后和封城后的交通拥堵情况。我们纳入了天气条件、季节性因素和covid -19相关变量这三种外部数据类别，并使用递归特征消除和交叉验证来识别不同时期的重要特征，避免潜在的过拟合。在此基础上，通过广泛的实验和参数调整，训练和优化多个高级机器学习（ML）模型，包括支持向量回归（SVR）、多元线性回归（MLR）、循环神经网络（RNN）和长短期记忆（LSTM）网络。由于LSTM具有更多的超参数，并且比使用的其他ML方法对调谐更敏感，我们采用自适应参数选择方法来优化其超参数，提高模型的精度和效率，而不是手动调整SVR和RNN的参数。这些模型通过标准化均方根误差进行评估。结果表明，双向LSTM （Bi-LSTM）在所有COVID-19期间的表现始终优于其他模型。这种优越的性能可以归因于Bi-LSTM的双向架构，该架构通过在时间上向前和向后分析数据来有效地捕获时间依赖性。为了解决机器学习方法的有限可解释性并提供有价值的见解，我们应用集成梯度（IG）技术来解释性能最佳且可微的Bi-LSTM预测。我们的分析显示，新冠肺炎病例对封锁期间和封锁后的交通拥堵产生了负面影响。观察到的交通量减少可以通过提高公众风险意识、自愿减少旅行以及遵守政府施加的流动限制来解释。考虑到IG不适用于该模型，我们还将SHapley Additive explanation应用于SVR。结果表明，在大流行后时期，人们变得更加谨慎——高新增住院率阻碍了出行，减少了交通拥堵，而高油价并没有阻止人们转向使用私家车，导致交通拥堵加剧。

{"title":"Interpretable machine learning for traffic congestion prediction: Unveiling the impact of different COVID-19 periods","authors":"Dan Zhu , ChiSin Ng , Litian Xie , Yang Liu","doi":"10.1016/j.commtr.2025.100226","DOIUrl":"10.1016/j.commtr.2025.100226","url":null,"abstract":"<div><div>Traffic congestion prediction plays a crucial role in mitigating congestion. However, the COVID-19 pandemic and associated government control measures have significantly altered urban travel behavior, increasing the complexity of traffic congestion prediction. This study aims to predict traffic congestion in Alameda County in the San Francisco Bay Area, USA, during the prelockdown, lockdown, and postlockdown periods. We incorporate three external categories of data, i.e., weather conditions, seasonality factors, and COVID-19-related variables, and use recursive feature elimination with cross-validation to identify important features across different periods and avoid potential overfitting. On this basis, multiple advanced machine learning (ML) models, including support vector regression (SVR), multiple linear regression (MLR), recurrent neural network (RNN), and long short-term memory (LSTM) networks, are trained and optimized through extensive experimentation and parameter tuning. Since LSTM has more hyperparameters and is more sensitive to tuning than the other ML methods used, we employ an adaptive parameter selection approach to optimize its hyperparameters, enhancing model accuracy and efficiency, rather than manually tuning parameters for SVR and RNN. These models are evaluated via the normalized root mean square error. The results indicate that the bidirectional LSTM (Bi-LSTM) consistently outperforms the other models across all COVID-19 periods. This superior performance can be attributed to the Bi-LSTM's bidirectional architecture, which effectively captures temporal dependencies by analyzing data both forward and backward in time. To address the limited interpretability of ML methods and provide valuable insights, we apply the integrated gradient (IG) technique to interpret the best-performing and differentiable Bi-LSTM predictions. Our analysis revealed that new COVID-19 cases had a negative influence on traffic congestion during the lockdown and postlockdown periods. The observed reduction in traffic can be explained by heightened public risk awareness, voluntary reductions in travel, and compliance with government-imposed mobility restrictions. We also apply SHapley Additive exPlanations to SVR, given that IG is not applicable to this model. The results indicate that in the postpandemic period, people have become more cautious—high new hospitalization discourages travel, reducing traffic congestion, whereas high fuel prices do not deter a shift toward private vehicle use, leading to increased congestion.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100226"},"PeriodicalIF":14.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145617647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multimodal traffic assignment from privacy-protected OD data 多模式交通分配从隐私保护OD数据

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-12-01 DOI: 10.1016/j.commtr.2025.100223

Guoyang Qin , Shidi Deng , Qi Luo , Jian Sun

The (static) traffic assignment (TA) problem, which computes network equilibrium flows from origin–destination (OD) demand under flow conservation, is central to transportation modeling. As multimodal transportation systems (MTSs) grow, sharing detailed OD data – such as trip counts, timestamps, and routes – raises serious privacy concerns. Differential privacy (DP) has emerged as the leading standard for releasing such data, offering adjustable protection beyond traditional anonymization. However, current methods mostly apply extrinsic DP by adding noise to aggregate OD matrices before release, without fully addressing its effects on traffic modeling. This reveals TA’s unpreparedness for privacy-protected data and calls for redesigned methods that operate reliably under such constraints. To fill this gap, we propose the privacy-preserving traffic assignment (PPTA) framework, which embeds DP intrinsically within the TA process. Instead of externally perturbing aggregate demand, PPTA injects structured noise at the individual trip level. This preserves privacy while ensuring equilibrium feasibility through chance-constrained optimization, unifying privacy protection and traffic assignment. The framework supports various discrete choice models and noise types, using a moment-based approximation to boost computational efficiency. Our results show PPTA attains a privacy-utility balance beyond extrinsic methods, enabling robust, privacy-aware multimodal routing, network design, and pricing.

静态交通分配（TA）问题是交通建模的核心问题，它计算流量守恒条件下从起点到终点的网络平衡流量。随着多式联运系统（mts）的发展，共享详细的OD数据（如旅行次数、时间戳和路线）引发了严重的隐私问题。差分隐私（DP）已经成为发布此类数据的主要标准，提供超越传统匿名化的可调整保护。然而，目前的方法大多是通过在释放之前在聚合OD矩阵中添加噪声来应用外在DP，而没有充分解决其对流量建模的影响。这揭示了TA对隐私保护数据的准备不足，并要求重新设计在这种约束下可靠运行的方法。为了填补这一空白，我们提出了隐私保护流量分配（PPTA）框架，该框架内在地将DP嵌入到流量分配过程中。而不是外部干扰总需求，PPTA注入结构化噪声在个别行程水平。通过机会约束优化，统一隐私保护和流量分配，在保证均衡可行性的同时，保护隐私。该框架支持各种离散选择模型和噪声类型，使用基于矩的近似来提高计算效率。我们的研究结果表明，PPTA在外部方法之外实现了隐私-效用平衡，实现了健壮的、隐私敏感的多模式路由、网络设计和定价。

{"title":"Multimodal traffic assignment from privacy-protected OD data","authors":"Guoyang Qin , Shidi Deng , Qi Luo , Jian Sun","doi":"10.1016/j.commtr.2025.100223","DOIUrl":"10.1016/j.commtr.2025.100223","url":null,"abstract":"<div><div>The (static) traffic assignment (TA) problem, which computes network equilibrium flows from origin–destination (OD) demand under flow conservation, is central to transportation modeling. As multimodal transportation systems (MTSs) grow, sharing detailed OD data – such as trip counts, timestamps, and routes – raises serious privacy concerns. Differential privacy (DP) has emerged as the leading standard for releasing such data, offering adjustable protection beyond traditional anonymization. However, current methods mostly apply extrinsic DP by adding noise to aggregate OD matrices before release, without fully addressing its effects on traffic modeling. This reveals TA’s unpreparedness for privacy-protected data and calls for redesigned methods that operate reliably under such constraints. To fill this gap, we propose the privacy-preserving traffic assignment (PPTA) framework, which embeds DP intrinsically within the TA process. Instead of externally perturbing aggregate demand, PPTA injects structured noise at the individual trip level. This preserves privacy while ensuring equilibrium feasibility through chance-constrained optimization, unifying privacy protection and traffic assignment. The framework supports various discrete choice models and noise types, using a moment-based approximation to boost computational efficiency. Our results show PPTA attains a privacy-utility balance beyond extrinsic methods, enabling robust, privacy-aware multimodal routing, network design, and pricing.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100223"},"PeriodicalIF":14.5,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145683864","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Scalable and reliable multi-agent reinforcement learning for traffic assignment 交通分配的可扩展和可靠的多智能体强化学习

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-11-21 DOI: 10.1016/j.commtr.2025.100225

Leizhen Wang , Peibo Duan , Cheng Lyu , Zewen Wang , Zhiqiang He , Nan Zheng , Zhenliang Ma

The evolution of metropolitan cities and increasing travel demand impose stringent requirements on traffic assignment methods. Multi-agent reinforcement learning (MARL) approaches outperform traditional methods in modeling adaptive routing behavior without requiring explicit system dynamics, making them attractive for real-world deployment. However, existing MARL frameworks face scalability and reliability challenges when managing large-scale networks with substantial and variable demand. This study proposes MARL-OD-DA, a novel framework that redefines agents as origin–destination (OD) pair routers and employs a continuous simplex-constrained action space. This reformulation reduces the agent population from O(N) (number of travelers) to O(|D|) (number of OD pairs), achieving at least two orders of magnitude fewer agents in practice while preserving convexity and enabling efficient adaptation to demand variation, thus significantly improving scalability. In contrast to prior MARL studies constrained to small-sized networks (up to 70 nodes, 2100 travelers) and fixed demand, MARL-OD-DA is validated on medium-sized networks (up to 416 nodes, 1406 OD pairs, and 360,600 travelers) under varying demand scenarios, demonstrating substantial improvements in scalability and applicability. To further enhance reliability, the framework integrates a Dirichlet-based policy, action pruning, and a relative gap-based reward. Theoretical analysis demonstrates that the Dirichlet-based policy reduces gradient bias, stabilizes variance, and enables sparse routing decisions, in contrast to the commonly used softmax-based policy. Experiments on three benchmark networks show that MARL-OD-DA significantly improves assignment quality and convergence speed. On the SiouxFalls network, the trained agents converge within 10 iterations during deployment, reducing the relative gap by 94.99% compared to conventional baselines.

大城市的发展和日益增长的出行需求对交通分配方法提出了严格的要求。多智能体强化学习（MARL）方法在建模自适应路由行为方面优于传统方法，而不需要显式的系统动力学，这使得它们对现实世界的部署具有吸引力。然而，当管理具有大量和可变需求的大规模网络时，现有的MARL框架面临着可扩展性和可靠性方面的挑战。本研究提出了一种新的框架MARL-OD-DA，该框架将智能体重新定义为原点-目的地（OD）对路由器，并采用连续的简单约束动作空间。这种重新表述将agent种群从O(N)（出行者数量）减少到O(|D|) （OD对数量），在保持凸性的同时实现了至少两个数量级的agent减少，并能够有效地适应需求变化，从而显著提高了可扩展性。与之前仅限于小型网络（最多70个节点，2100个出行者）和固定需求的MARL研究相比，MARL-OD- da在不同需求场景下的中型网络（最多416个节点，1406个OD对，360,600个出行者）上进行了验证，证明了可扩展性和适用性的显著提高。为了进一步提高可靠性，该框架集成了基于dirichlet的策略、行动修剪和相对基于缺口的奖励。理论分析表明，与常用的基于softmaxs的策略相比，基于dirichlet的策略减少了梯度偏差，稳定了方差，并实现了稀疏路由决策。在三个基准网络上的实验表明，MARL-OD-DA算法显著提高了分配质量和收敛速度。在SiouxFalls网络上，经过训练的代理在部署期间的10次迭代内收敛，与传统基线相比，将相对差距减少了94.99%。

{"title":"Scalable and reliable multi-agent reinforcement learning for traffic assignment","authors":"Leizhen Wang , Peibo Duan , Cheng Lyu , Zewen Wang , Zhiqiang He , Nan Zheng , Zhenliang Ma","doi":"10.1016/j.commtr.2025.100225","DOIUrl":"10.1016/j.commtr.2025.100225","url":null,"abstract":"<div><div>The evolution of metropolitan cities and increasing travel demand impose stringent requirements on traffic assignment methods. Multi-agent reinforcement learning (MARL) approaches outperform traditional methods in modeling adaptive routing behavior without requiring explicit system dynamics, making them attractive for real-world deployment. However, existing MARL frameworks face scalability and reliability challenges when managing large-scale networks with substantial and variable demand. This study proposes MARL-OD-DA, a novel framework that redefines agents as origin–destination (OD) pair routers and employs a continuous simplex-constrained action space. This reformulation reduces the agent population from <em>O</em>(<em>N</em>) (number of travelers) to <em>O</em>(|<em>D</em>|) (number of OD pairs), achieving at least two orders of magnitude fewer agents in practice while preserving convexity and enabling efficient adaptation to demand variation, thus significantly improving scalability. In contrast to prior MARL studies constrained to small-sized networks (up to 70 nodes, 2100 travelers) and fixed demand, MARL-OD-DA is validated on medium-sized networks (up to 416 nodes, 1406 OD pairs, and 360,600 travelers) under varying demand scenarios, demonstrating substantial improvements in scalability and applicability. To further enhance reliability, the framework integrates a Dirichlet-based policy, action pruning, and a relative gap-based reward. Theoretical analysis demonstrates that the Dirichlet-based policy reduces gradient bias, stabilizes variance, and enables sparse routing decisions, in contrast to the commonly used softmax-based policy. Experiments on three benchmark networks show that MARL-OD-DA significantly improves assignment quality and convergence speed. On the SiouxFalls network, the trained agents converge within 10 iterations during deployment, reducing the relative gap by 94.99% compared to conventional baselines.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100225"},"PeriodicalIF":14.5,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571109","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PriorFusion: Unified integration of priors for robust road perception in autonomous driving PriorFusion：用于自动驾驶稳健道路感知的先验统一集成

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-11-18 DOI: 10.1016/j.commtr.2025.100229

Xuewei Tang , Mengmeng Yang , Tuopu Wen , Peijin Jia , Le Cui , Mingshan Luo , Kehua Sheng , Bo Zhang , Kun Jiang , Diange Yang

With the growing interest in autonomous driving, there is an increasing demand for accurate and reliable road perception technologies. In complex environments without high-definition map support, autonomous vehicles must independently interpret their surroundings to ensure safe and robust decision-making. However, these scenarios pose significant challenges due to the large number, complex geometries, and frequent occlusions of road elements. A key limitation of existing approaches lies in their insufficient exploitation of the structured priors inherently present in road elements, resulting in irregular, inaccurate predictions. To address this, we propose PriorFusion, a unified framework that effectively integrates semantic, geometric, and generative priors to enhance road element perception. We introduce an instance-aware attention mechanism guided by shape-prior features, then construct a data-driven shape template space that encodes low-dimensional representations of road elements, enabling clustering to generate anchor points as reference priors. We design a diffusion-based framework that leverages these prior anchors to generate accurate and complete predictions. Experiments on large-scale autonomous driving datasets demonstrate that our method significantly improves perception accuracy, particularly under challenging conditions. Visualization results further confirm that our approach produces more accurate, regular, and coherent predictions of road elements.

随着人们对自动驾驶的兴趣日益浓厚，对准确可靠的道路感知技术的需求也越来越大。在没有高清地图支持的复杂环境中，自动驾驶汽车必须独立解读周围环境，以确保安全可靠的决策。然而，由于数量庞大、几何形状复杂、道路元素频繁遮挡，这些场景带来了重大挑战。现有方法的一个关键限制在于它们没有充分利用道路要素固有的结构化先验，从而导致不规则和不准确的预测。为了解决这个问题，我们提出了PriorFusion，这是一个统一的框架，有效地集成了语义、几何和生成先验，以增强道路要素感知。我们引入了一种由形状先验特征引导的实例感知注意力机制，然后构建了一个数据驱动的形状模板空间，该空间对道路元素的低维表示进行编码，使聚类能够生成锚点作为参考先验。我们设计了一个基于扩散的框架，利用这些先前的锚来生成准确和完整的预测。在大规模自动驾驶数据集上的实验表明，我们的方法显著提高了感知精度，特别是在具有挑战性的条件下。可视化结果进一步证实，我们的方法产生了更准确、更规律、更连贯的道路元素预测。

{"title":"PriorFusion: Unified integration of priors for robust road perception in autonomous driving","authors":"Xuewei Tang , Mengmeng Yang , Tuopu Wen , Peijin Jia , Le Cui , Mingshan Luo , Kehua Sheng , Bo Zhang , Kun Jiang , Diange Yang","doi":"10.1016/j.commtr.2025.100229","DOIUrl":"10.1016/j.commtr.2025.100229","url":null,"abstract":"<div><div>With the growing interest in autonomous driving, there is an increasing demand for accurate and reliable road perception technologies. In complex environments without high-definition map support, autonomous vehicles must independently interpret their surroundings to ensure safe and robust decision-making. However, these scenarios pose significant challenges due to the large number, complex geometries, and frequent occlusions of road elements. A key limitation of existing approaches lies in their insufficient exploitation of the structured priors inherently present in road elements, resulting in irregular, inaccurate predictions. To address this, we propose PriorFusion, a unified framework that effectively integrates semantic, geometric, and generative priors to enhance road element perception. We introduce an instance-aware attention mechanism guided by shape-prior features, then construct a data-driven shape template space that encodes low-dimensional representations of road elements, enabling clustering to generate anchor points as reference priors. We design a diffusion-based framework that leverages these prior anchors to generate accurate and complete predictions. Experiments on large-scale autonomous driving datasets demonstrate that our method significantly improves perception accuracy, particularly under challenging conditions. Visualization results further confirm that our approach produces more accurate, regular, and coherent predictions of road elements.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100229"},"PeriodicalIF":14.5,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571166","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

From global open multi-source data to network-wide traffic flow: A large-scale case study across multiple cities 从全球开放多源数据到网络范围的交通流量：跨多个城市的大规模案例研究

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-11-18 DOI: 10.1016/j.commtr.2025.100222

Zijian Hu , Zhenjie Zheng , Monica Menendez , Wei Ma

Network-wide traffic flow, which captures dynamic traffic volume on each link of a general network, is fundamental to smart mobility applications. However, the observed traffic flow from sensors is usually limited across the entire network due to the associated high installation and maintenance costs. To address this issue, existing research uses various supplementary data sources to compensate for insufficient sensor coverage and estimate the unobserved traffic flow. Although these studies have shown promising results, the inconsistent availability and quality of supplementary data across cities make their methods typically face a trade-off challenge between accuracy and generality. In this research, we first advocate using the global open multi-source (GOMS) data within an advanced deep learning framework to break the trade-off. The GOMS data mainly refers to publicly available multi-type datasets, including road topology, building footprints, and population density, which can be consistently collected across cities. More importantly, these GOMS data are closely related to the traffic flow dynamics, thereby creating opportunities for accurate network-wide flow estimation. Furthermore, we use map images to represent GOMS data, instead of traditional tabular formats, to capture richer and more comprehensive geographical and demographic information. To address multi-source data fusion, we develop an attention-based graph neural network that effectively extracts and synthesizes information from GOMS maps while simultaneously capturing spatiotemporal traffic dynamics from observed traffic data. A large-scale case study across 15 cities in Europe and North America was conducted. The results demonstrate stable and satisfactory estimation accuracy across these cities, which suggests that the trade-off challenge can be successfully addressed using our approach.

网络范围内的流量是智能移动应用的基础，它捕获了一般网络中每个链路上的动态流量。然而，由于相关的高安装和维护成本，从传感器观察到的流量通常在整个网络中受到限制。为了解决这个问题，现有的研究使用各种补充数据源来补偿传感器覆盖范围的不足，并估计未观察到的交通流量。尽管这些研究显示出有希望的结果，但各城市补充数据的可用性和质量不一致，使得他们的方法通常面临准确性和普遍性之间的权衡挑战。在本研究中，我们首先倡导在先进的深度学习框架内使用全球开放多源（GOMS）数据来打破权衡。GOMS数据主要是指可公开获取的多类型数据集，包括道路拓扑、建筑足迹、人口密度等，这些数据集可以跨城市统一收集。更重要的是，这些GOMS数据与交通流量动态密切相关，从而为准确估计全网流量创造了机会。此外，我们使用地图图像来表示GOMS数据，而不是传统的表格格式，以获取更丰富、更全面的地理和人口信息。为了解决多源数据融合问题，我们开发了一种基于注意力的图神经网络，该网络可以有效地从GOMS地图中提取和综合信息，同时从观测到的交通数据中捕获时空交通动态。在欧洲和北美的15个城市进行了大规模的案例研究。结果表明，这些城市的估计精度稳定且令人满意，这表明使用我们的方法可以成功地解决权衡挑战。

{"title":"From global open multi-source data to network-wide traffic flow: A large-scale case study across multiple cities","authors":"Zijian Hu , Zhenjie Zheng , Monica Menendez , Wei Ma","doi":"10.1016/j.commtr.2025.100222","DOIUrl":"10.1016/j.commtr.2025.100222","url":null,"abstract":"<div><div>Network-wide traffic flow, which captures dynamic traffic volume on each link of a general network, is fundamental to smart mobility applications. However, the observed traffic flow from sensors is usually limited across the entire network due to the associated high installation and maintenance costs. To address this issue, existing research uses various supplementary data sources to compensate for insufficient sensor coverage and estimate the unobserved traffic flow. Although these studies have shown promising results, the inconsistent availability and quality of supplementary data across cities make their methods typically face a trade-off challenge between accuracy and generality. In this research, we first advocate using the global open multi-source (GOMS) data within an advanced deep learning framework to break the trade-off. The GOMS data mainly refers to publicly available multi-type datasets, including road topology, building footprints, and population density, which can be consistently collected across cities. More importantly, these GOMS data are closely related to the traffic flow dynamics, thereby creating opportunities for accurate network-wide flow estimation. Furthermore, we use map images to represent GOMS data, instead of traditional tabular formats, to capture richer and more comprehensive geographical and demographic information. To address multi-source data fusion, we develop an attention-based graph neural network that effectively extracts and synthesizes information from GOMS maps while simultaneously capturing spatiotemporal traffic dynamics from observed traffic data. A large-scale case study across 15 cities in Europe and North America was conducted. The results demonstrate stable and satisfactory estimation accuracy across these cities, which suggests that the trade-off challenge can be successfully addressed using our approach.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100222"},"PeriodicalIF":14.5,"publicationDate":"2025-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Public air routes for low-altitude economies: Priority infrastructure beneficially associated with ground roads 低空经济的公共航线：与地面道路相关的优先基础设施

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-11-18 DOI: 10.1016/j.commtr.2025.100228

Chenchen Xu , Zongru Li , Hongbo He , Anqi Tang , Xiaohan Liao

The emerging low-altitude economy (LAE) hinges on effectively managing high-density aerial traffic flows. Establishing a system of low-altitude public air routes offers a feasible solution to handle this congestion. In this context, these routes are conceptualized as foundational “sky–road” infrastructure, and it is proposed that they be treated as fixed assets endowed with tradable or transferable property rights. Previous analysis suggests that building a public low-altitude air-route network can generate significant socioeconomic benefits, potentially beyond those of conventional ground transportation. Although “building sky roads” poses technological and regulatory challenges, extending the principles of terrestrial road networks into low-altitude airspace allows existing planning standards and governance mechanisms to be largely adapted to this new domain. This approach can transform many of these challenges into opportunities, laying the groundwork for a robust LAE in the future.

新兴的低空经济（LAE）取决于对高密度空中交通流的有效管理。建立低空公共航线系统为解决这一拥堵问题提供了一个可行的解决方案。在这种情况下，这些路线被概念化为基础的“空中道路”基础设施，并建议将它们视为具有可交易或可转让产权的固定资产。先前的分析表明，建立一个公共低空航线网络可以产生显著的社会经济效益，可能超过传统的地面运输。尽管“建造空中道路”带来了技术和监管方面的挑战，但将地面道路网络的原则扩展到低空空域，可以使现有的规划标准和治理机制在很大程度上适应这一新领域。这种方法可以将许多挑战转化为机遇，为未来强大的LAE奠定基础。

引用次数: 0

Efficient and stable ride-pooling through a multi-level coalition formation game 通过多级联盟形成博弈实现高效稳定的拼车

IF 14.5 Q1 TRANSPORTATION

Communications in Transportation Research

Pub Date : 2025-11-15 DOI: 10.1016/j.commtr.2025.100220

Yaotian Tan , Shuyue Qian , Aoyong Li , Haiyang Yu , Jie Gao

Ride-pooling has the potential to offer a sustainable solution for urban mobility by reducing vehicle use and emissions through shared trips. However, its adoption remains limited due to poor matching performance. Many requests fail to form feasible pools, and even successful matches often involve long detours or minimal cost savings. These inefficiencies largely arise from fragmented market structures: most operators act independently, restricting matching to their own request pools and limiting the formation of beneficial coalitions. Aggregation platforms improve efficiency by integrating regional operators through unified dispatch systems, but raise concerns over long-term stability. Differences in operator cost structures and market shares may incentivize deviation, at the same time, passengers may reject assigned payments if more attractive alternatives exist. To address these challenges, we propose a multi-level coalition formation game that jointly models operator and passenger collaboration. At the upper level, operators play a non-cooperative game to decide coalition partners. At the lower level, passengers are grouped into shared trips through a cooperative game that ensures individually rational payments. The two layers are coupled via constraint propagation, forming a unified decision-making process. We evaluate our framework using real-world data from three Chinese regions—Chengdu, Haikou, and the Ningxia Hui Autonomous Region—chosen to reflect diverse urban and regional contexts. Compared to independent operations, our approach increases vehicle occupancy by 14%–28%, reduces total costs by 10%–15%, and shortens average travel distances by 4%–5%. The system maintains stable coalition structures with operator deviation rates below 6.81% and near-zero passenger deviation rates.

拼车有可能通过共享出行减少车辆使用和排放，为城市交通提供可持续的解决方案。然而，由于匹配性能差，其采用仍然受到限制。许多请求不能形成可行的池，即使成功的匹配也常常需要绕很长一段路或节省很少的成本。这些低效率很大程度上源于分散的市场结构：大多数运营商独立行动，限制了他们自己的请求池的匹配，限制了有利联盟的形成。集散平台通过统一调度系统整合地区运营商，提高了效率，但也引发了对长期稳定性的担忧。运营商成本结构和市场份额的差异可能会激励偏差，同时，如果存在更有吸引力的替代方案，乘客可能会拒绝分配的付款。为了应对这些挑战，我们提出了一个多级联盟形成游戏，共同模拟运营商和乘客的合作。在上层，运营商通过非合作博弈来决定联盟伙伴。在较低的层次，乘客通过合作游戏被分组到共享行程中，以确保个人合理支付。两层通过约束传播耦合，形成统一的决策过程。我们使用来自中国三个地区（成都、海口和宁夏回族自治区）的真实数据来评估我们的框架，这些地区的选择反映了不同的城市和区域背景。与独立运营相比，我们的方法使车辆占用率提高了14%-28%，总成本降低了10%-15%，平均行驶距离缩短了4%-5%。该系统保持稳定的联盟结构，操作人员的偏差率低于6.81%，乘客的偏差率接近于零。

{"title":"Efficient and stable ride-pooling through a multi-level coalition formation game","authors":"Yaotian Tan , Shuyue Qian , Aoyong Li , Haiyang Yu , Jie Gao","doi":"10.1016/j.commtr.2025.100220","DOIUrl":"10.1016/j.commtr.2025.100220","url":null,"abstract":"<div><div>Ride-pooling has the potential to offer a sustainable solution for urban mobility by reducing vehicle use and emissions through shared trips. However, its adoption remains limited due to poor matching performance. Many requests fail to form feasible pools, and even successful matches often involve long detours or minimal cost savings. These inefficiencies largely arise from fragmented market structures: most operators act independently, restricting matching to their own request pools and limiting the formation of beneficial coalitions. Aggregation platforms improve efficiency by integrating regional operators through unified dispatch systems, but raise concerns over long-term stability. Differences in operator cost structures and market shares may incentivize deviation, at the same time, passengers may reject assigned payments if more attractive alternatives exist. To address these challenges, we propose a multi-level coalition formation game that jointly models operator and passenger collaboration. At the upper level, operators play a non-cooperative game to decide coalition partners. At the lower level, passengers are grouped into shared trips through a cooperative game that ensures individually rational payments. The two layers are coupled via constraint propagation, forming a unified decision-making process. We evaluate our framework using real-world data from three Chinese regions—Chengdu, Haikou, and the Ningxia Hui Autonomous Region—chosen to reflect diverse urban and regional contexts. Compared to independent operations, our approach increases vehicle occupancy by 14%–28%, reduces total costs by 10%–15%, and shortens average travel distances by 4%–5%. The system maintains stable coalition structures with operator deviation rates below 6.81% and near-zero passenger deviation rates.</div></div>","PeriodicalId":100292,"journal":{"name":"Communications in Transportation Research","volume":"5 ","pages":"Article 100220"},"PeriodicalIF":14.5,"publicationDate":"2025-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0