Research of the automatic train stop control based on adaptive generalized predictive control

Precise train stop is a key technology of automatic train operation. This study presents an adaptive generalized predictive controller ensuring the train stop exactly at a station. The controller is designed by considering the delay characteristics of the braking system and the real-time constraint conditions of train operation. It takes the desired speed and distance on the line as the control targets to obtain the optimal control variable, which is calculated by real-time quadratic optimization. Concurrently the online identification of the model parameters could improve the performance of the predictive controller. We carry out simulations to prove the effectiveness of the proposed controller and consider the running resistance and the model parameters uncertainty exist. It shows the controller can achieve the high precise train stop under the good comfort of the passengers.