Sequential-Agent Game Lane-Changing Model for CAVs Approaching Off-Ramps

IF 9.1 1区工程技术 Q1 ENGINEERING, CIVIL

IEEE Transactions on Intelligent Transportation Systems Pub Date : 2026-10-01 Epub Date: 2026-08-10 DOI:10.1109/TITS.2026.3705746

Qinglu Ma;Zixuan Li;Shengping Niu;Xinyu Wang;Saleem Ullah

{"title":"Sequential-Agent Game Lane-Changing Model for CAVs Approaching Off-Ramps","authors":"Qinglu Ma;Zixuan Li;Shengping Niu;Xinyu Wang;Saleem Ullah","doi":"10.1109/TITS.2026.3705746","DOIUrl":null,"url":null,"abstract":"To improve the safety and efficiency of Connected and Autonomous Vehicles (CAVs) approaching off-ramps, this study proposes a Sequential-Agent Game Lane-Changing Model (S-GLM), which couples sequential game theory with agent-oriented LSTM-DQN collaborative training under explicit safety constraints. Through a sequential game process, S-GLM extends lane-changing decision-making beyond single-step reasoning toward continuous interaction optimization. Field data were collected from Masangxi Bridge to Dadukou Interchange on Chongqing’s Inner Ring Expressway using autonomous aerial vehicles (AAVs) and fixed cameras. The proposed S-GLM was validated through joint simulation using SUMO, Python, and TraCI. The results show that S-GLM increases the Modified Time to Collision (MTTC) by 9.15% and 14.89% compared with the Pure Reinforcement Learning baseline (PRL-b) and the Mixed-Strategy Game Lane-Changing Model (M-GLM), respectively, while improving the average vehicle speed by 2.09% and 1.72%. Overall, these improvements highlight the significance of S-GLM for research on CAV off-ramp diverging.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"27 10","pages":"12946-12958"},"PeriodicalIF":9.1000,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Intelligent Transportation Systems","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/11647261/","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2026/8/10 0:00:00","PubModel":"Epub","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}

引用次数: 0

Abstract

To improve the safety and efficiency of Connected and Autonomous Vehicles (CAVs) approaching off-ramps, this study proposes a Sequential-Agent Game Lane-Changing Model (S-GLM), which couples sequential game theory with agent-oriented LSTM-DQN collaborative training under explicit safety constraints. Through a sequential game process, S-GLM extends lane-changing decision-making beyond single-step reasoning toward continuous interaction optimization. Field data were collected from Masangxi Bridge to Dadukou Interchange on Chongqing’s Inner Ring Expressway using autonomous aerial vehicles (AAVs) and fixed cameras. The proposed S-GLM was validated through joint simulation using SUMO, Python, and TraCI. The results show that S-GLM increases the Modified Time to Collision (MTTC) by 9.15% and 14.89% compared with the Pure Reinforcement Learning baseline (PRL-b) and the Mixed-Strategy Game Lane-Changing Model (M-GLM), respectively, while improving the average vehicle speed by 2.09% and 1.72%. Overall, these improvements highlight the significance of S-GLM for research on CAV off-ramp diverging.

查看原文本刊更多论文

自动驾驶汽车驶近匝道的序贯智能体博弈变道模型

为了提高网联自动驾驶汽车（cav）驶近匝道的安全性和效率，本研究提出了一种序列-智能体博弈变道模型（S-GLM），该模型将序列博弈论与明确安全约束下面向智能体的LSTM-DQN协同训练相结合。通过一个顺序博弈过程，S-GLM将变道决策从单步推理扩展到连续交互优化。现场数据是利用自主飞行器（aav）和固定摄像头从重庆内环高速公路的马山溪大桥到大渡口立交收集的。通过使用SUMO、Python和TraCI进行联合仿真，验证了所提出的S-GLM。结果表明，与纯强化学习基线（PRL-b）和混合策略博弈变道模型（M-GLM）相比，S-GLM的修正碰撞时间（MTTC）分别提高了9.15%和14.89%，平均车速提高了2.09%和1.72%。综上所述，这些改进凸显了S-GLM在CAV出入口分流研究中的重要意义。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

IEEE Transactions on Intelligent Transportation Systems 工程技术-工程：电子与电气

CiteScore

14.80

自引率

12.90%

发文量

1872

审稿时长

7.5 months

期刊介绍： The theoretical, experimental and operational aspects of electrical and electronics engineering and information technologies as applied to Intelligent Transportation Systems (ITS). Intelligent Transportation Systems are defined as those systems utilizing synergistic technologies and systems engineering concepts to develop and improve transportation systems of all kinds. The scope of this interdisciplinary activity includes the promotion, consolidation and coordination of ITS technical activities among IEEE entities, and providing a focus for cooperative activities, both internally and externally.