Real-Time Optimal Speed Coordination and Scheduling for High-Speed Trains Based on Model Predictive Control

Abstract

In the high-speed train control system, the command information such as allowable running distance, time and speed can be sent by the global system for mobile communications for railways (GSM-R). This paper will propose the framework of real-time train scheduling and control based on model predictive control for the optimal speed set-points of high-speed trains. The rolling optimization process combines the genetic algorithm with the simulation of train operation to evaluate the performance of speed set-points, which can be easily implemented in the parallel computing environment for real-time processing. The conflict resolution at the crossing stations is modeled by and embedded in the combination of various speed set-points which are formed from virtual to simulation speed. The final actual speed of train is engendered based on the movement authority and running time through the system of automatic train protection (ATP). The simulation results demonstrate the favorable performance of proposed method.