In cities with expanding public transportation systems, Autonomous Rail Rapid Transit (ART) combines the benefits of traditional rail and bus systems, offering a cost-effective, quickly implementable, and efficient transit solution. This study investigates managing ART train operations under virtual coupling, focusing on decoupling operations at signalized intersections and addressing the challenges posed by dynamic passenger demand with uneven distribution. A comprehensive traffic model for ART trains is introduced, incorporating decoupling operations, signal constraints, and dynamic passenger flow. The decoupling operation process at intersections is integrated, enabling the determination of exact signal waiting times for each train unit under virtual coupling. A Deep Q Network algorithm is then developed to determine optimal train operation strategies with discretized actions related to running speed, allowing adaptation to varying operational scenarios, minimizing passenger travel time, and ensuring punctuality. The model is validated based on simulation experiments using real-world data. And the results indicate that the proposed approach effectively reduces passenger travel time and improves punctuality. The integration of virtual coupling with advanced rescheduling strategies enhances the efficiency of ART systems by optimizing train operations at intersections and dynamically adjusting to passenger demand.
Virtual Coupling Rescheduling for Autonomous Rail Rapid Transit on Y-Type Lines with Signalized Intersections
2025-05-09
2578487 byte
Conference paper
Electronic Resource
English
Simulation-optimization framework for train rescheduling in rapid rail transit
Taylor & Francis Verlag | 2021
|Engineering Index Backfile | 1968
|