ITSC 2025 Paper Abstract

Paper WE-EA-T10.5

Zhao, Zixu (Beijing JiaoTong University), Lyu, Jidong (Beijing Jiaotong University), Luo, Zhengwei (Beijing Jiaotong University), Lyu, Jiahui (Beijing Jiaotong University), Chai, Ming (Beijing Jiaotong University), Liu, Hongjie (Beijing Jiaotong University), Chen, Junqiang (Traffic Control Technology Co., Ltd)

Coordinated Trajectory Planning for High-Speed Train Group Operation Based on Deep Reinforcement Learning

Scheduled for presentation during the Regular Session "S10b-Cooperative and Connected Autonomous Systems" (WE-EA-T10), Wednesday, November 19, 2025, 14:50−14:50, Cooleangata 4

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 19, 2025

Keywords Cooperative Driving Systems and Vehicle Coordination in Multi-vehicle Scenarios, Autonomous Rail Systems and Advanced Train Control Technologies, Real-time Motion Planning and Control for Autonomous Vehicles in ITS Networks

Abstract

The high-speed train group (HST-G) is one of the emerging modes to improve the transportation efficiency of high-speed railways. Coordinated trajectory planning is of a great challenge for HST-G considering multi-objective optimization. This paper proposes a multi-objective coordinated trajectory planning method based on deep reinforcement learning algorithm. Firstly, the dynamics model of HST-G is established, in which multiple optimization objectives including operational synchronization and headway are considered. Secondly, by capturing the state of group trains as the state variables, the coordinated trajectory planning problem is modeled as a MDP. And then the PPO algorithm for HST-G trajectory planning is delivered featuring the continuity of control action space. Finally, the global offline and online adjustment collaborative planning simulation of trains is carried out based on real line data. The experimental results show that the PPO-based coordinated trajectory planning improves line capacity utilization compared to traditional mode. The PPO algorithm outperforms the GA algorithm in planning efficiency and group trains punctuality, while also achieving a better balance among running synchronization, ride quality and punctuality.