Cooperative control of high-speed trains for headway regulation: A self-triggered model predictive control based approach. (May 2019)

• Record Type: Journal Article
• Authors: Xun, Jing; Yin, Jiateng; Liu, Ronghui; Liu, Fan; Zhou, Yang; Tang, Tao
• Is Part Of: Transportation research. Volume 102(2019)
• Page Start: 106
• Page End: 120
• Publication Date: 2019-05
• Subjects: Cooperative train control -- Headway adjustment -- High-speed railways -- Self-triggered control; Transportation -- Periodicals; Transportation -- Technological innovations -- Periodicals; 388.011
• DOI: 10.1016/j.trc.2019.02.023
• Languages: English
• ISSNs: 0968-090X

Abstract

Highlights: A multi-train cooperative control model was proposed based on CACC. We improved the rolling horizon approach into a novel self-triggered model predictive control scheme. Train controller output constraints and safe train following distance were considered. The simulation results were analyzed to demonstrate the effectiveness of the proposed approach. Abstract: The advanced train-to-train and train-to-ground communication technologies equipped in high-speed railways have the potential to allow trains to follow each with a steady headway and improve the safety and performance of the railway systems. A key enabler is a train control system that is able to respond to unforeseen disturbances in the system (e.g., incidents, train delays), and to adjust and coordinate the train headways and speeds. This paper proposes a multi-train cooperative control model based on the dynamic features during train longitude movement to adjust train following headway. In particular, our model simultaneously considers several practical constraints, e.g., train controller output constraints, safe train following distance, as well as communication delays and resources. Then, this control problem is solved through a rolling horizon approach by calculating the Riccati equation with Lagrangian multipliers. Due to the practical communication resource constraints and riding comfort requirement, we also improved the rolling horizon approach into a novel self-triggered model predictive control scheme to overcome these issues. Finally, two case studies are given through simulation experiments. The simulation results are analyzed which demonstrate the effectiveness of the proposed approach.