摘要
针对地铁接驳高铁客流疏散场景中乘客拥挤、候车时间过长及交通资源浪费等问题,提出基于多智能体强化学习(Multi-Agent Reinforcement Learning,MARL)的地铁接驳高铁客流疏散优化方法.该方法通过动态调整地铁时刻表,提高乘客疏散效率,减少拥挤情况,并提高交通资源利用率.首先,根据地铁的时空信息及乘客换乘的时空参数,将地铁接驳高铁客流疏散优化问题建模为马尔可夫博弈过程,并设计通用状态特征、行为空间和奖励函数.然后,采用Actor-Critic(AC)框架建立多智能体的决策模型,并在集中式训练和分布式执行的框架下设计一种异步动作协同机制,以提高方法的训练效率.最后,以天津西站换乘地铁为案例进行优化研究.研究结果表明:优化地铁接驳高铁客流疏散能显著降低乘客候车时间,并提高地铁的运行效率;乘客平均候车时间减少了26.80%,地铁的平均运行效率提高了14.11%.
To address challenges such as passenger crowding,excessive waiting times,and inefficient use of transportation resources in metro-to-high-speed rail transfer scenarios,this study proposes an optimization method for metro-to-high-speed rail passenger flow dispersion based on Multi-Agent Re-inforcement Learning(MARL).The method dynamically adjusts metro timetables to enhance passen-ger dispersion efficiency,alleviate crowding,and improve the utilization of transportation resources.First,the metro-to-high-speed rail passenger flow dispersion optimization problem is formulated as a Markov game by integrating the spatiotemporal information of metro operations and the spatiotempo-ral characteristics of passenger transfers,with general state features,action space,and a reward func-tion specifically designed.Second,a multi-agent decision-making model is then developed using the Actor-Critic(AC)framework,and an asynchronous action coordination mechanism is introduced within a centralized training and distributed execution architecture to enhance training efficiency.Fi-nally,an optimization study is conducted using the Tianjin West railway station as a case study.Re-sults indicate that the proposed method significantly reduces passenger waiting times and improves metro operational efficiency.The average passenger waiting time decreases by 26.80%,while the av-erage metro operational efficiency increases by 14.11%.
作者
孙峣
柯水平
贾宁
辛秀颖
SUN Yao;KE Shuiping;JIA Ning;XIN Xiuying(School of Mechanical Engineering,Dalian University of Technology,Dalian Liaoning 116024,China;Tianjin Municipal Engineering Design&Research Institute,Tianjin 300051,China;College of Management and Econom-ics,Tianjin University,Tianjin 300072,China;School of Economics and Management,Tianjin University of Sci-ence and Technology,Tianjin 300222,China)
出处
《北京交通大学学报》
北大核心
2025年第4期19-28,共10页
JOURNAL OF BEIJING JIAOTONG UNIVERSITY
基金
国家自然科学基金(52372313)
天津市人社局“131”创新团队项目(2019-44)。
关键词
多智能体强化学习
地铁接驳
客流疏散
异步动作协同机制
multi-agent reinforcement learning
metro connection
passenger flow dispersion
asyn-chronous action coordination mechanism
