Back to Search Start Over

A Model Predictive Control System for Virtual Coupling

Authors :
Xiaolin Luo
Hongjie Liu
Xiwang Guo
Ming Chai
Tao Tang
Source :
2021 33rd Chinese Control and Decision Conference (CCDC).
Publication Year :
2021
Publisher :
IEEE, 2021.

Abstract

Virtual Coupling (VC) is a novel train operation mode which makes trains move synchronously as a platoon via Train-to-Train (T2T) communication and they can be seen as a single train if they have same route ahead. This method dramatically enhances track capacity by decreasing headway between consecutive trains. How to realize VC operation is a hot topic recently. This paper proposes a model predictive control (MPC) system for VC to track reference driving profile or predecessor's trajectory, i.e., driving as fast as they can, to a certain extent. Finite-time horizon prediction horizon is expanded to Quasi-infinite-time prediction horizon by penalizing the last state error. Input constraints (acceleration and jerk constraints) and output constraints (safety headway constraints) are considered. This system also guarantees asymptotical stability and string stability. The results show that each train in platoon can automatically adjust its dynamic state as anticipation under safety.

Details

Database :
OpenAIRE
Journal :
2021 33rd Chinese Control and Decision Conference (CCDC)
Accession number :
edsair.doi...........c42698811c553c339ceb5e32f61a89c8