Gong, Feijie (Beijing Jiaotong University), ShangGuan, Wei (Beijing Jiaotong University), Song, Hongyu (Beijing Jiaotong University), Dun, Yichen (Beijing Jiaotong University), Cai, Baigen (Beijing Jiaotong University)

DRL-Based Optimization of LQR for Virtual Coupling Train Interval Control

Scheduled for presentation during the Invited Session "S16b-Control, Communication and Emerging Technologies in Smart Rail Systems" (TH-EA-T16), Thursday, November 20, 2025, 13:50−14:10, Southport 1

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 18, 2025

Abstract

Virtual Coupling (VC) allows trains to operate in formation, remarkably promoting operational efficiency. The inter-train interval control is crucial for the safe operation of VC. To address this, this study proposes a Linear Quadratic Regulator (LQR) interval controller based on DRL optimization. A quadratic optimization model is built using the LQR state feedback control law, and the control output is derived by solving the Riccati equation. To overcome the low accuracy caused by suboptimal LQR parameters, the Deep Deterministic Policy Gradient (DDPG) algorithm optimizes the parameters. An Actor-Critic network tunes control parameters, with LQR parameters as DDPG actions, interval control errors as states, and reward maximization linked to error reduction. Optimized parameters are achieved through iterative learning. This approach is stable while improving the accuracy of the interval control by 68. 5%.