Train control algorithm based on generalized predictive control

Abstract

The train running process is a complicated process, and the traditional control algorithm is difficult to adapt to the changing running conditions. GPC has the characteristics of multi-step prediction, rolling optimization and feedback correction, and it is a kind of control algorithm which is easy to be realized. In this paper, based on the original GPC, stair-like control scheme is used and the predicted future changes are considered as feedback. In this way, the matrix inversion operation is avoided and the overshoot is suppressed. The effectiveness of the improved GPC algorithm is proved by the simulation.