A model predictive control strategy with switching cost functions for cooperative operation of trains

The cooperative control of trains is proposed as an innovative method for further improving operation efficiency. Model predictive control (MPC) has been widely discussed for multiple trains because it can handle the challenges posed by the cooperative control problem, such as complex constraints. In real situations, multiple objectives, such as comfort and safety, must be considered when controlling multiple trains with MPC, and the total objective may change during operation, affecting control performance. In this paper, a distributed structure based on switching cost function model predictive control (ScMPC) for multiple trains in a switching situation is given, where the cost functions of the train control problem change with the variable demand of cooperative operation. Furthermore, the feasibility of the proposed method and stability of the closed-loop system are proved to guarantee the stable operation of the controlled trains. Finally, the control method’s effectiveness is verified. Three kinds of cost functions are given, and their control performance is compared to show the effect of different weights and the advantage of ScMPC.