Low-complexity model-free automatic train operation control with protection constraints

A low-complexity model-free control method is proposed to concurrently satisfy the speed tracking performance constraint and collision avoidance requirements for automatic train operations (ATO). By defining prescribed performance functions, these protection constraints are transformed into either a speed error control problem or a distance error control problem, each with a predefined evolution range. A novel error-based high-gain feedback switch control algorithm is developed specifically for speed and distance control, utilizing solely speed and distance information without relying on models. The designed control method can achieve the prescribed performance speed control when the anterior train is remote and guarantee the safe distance when the anterior train is closed-by. Moreover, recognizing that the closed-loop system formed by the proposed speed/distance protection controller may result in a non-convex solution space, we establish the closed-loop system by Carathéodory function to demonstrate the system stability through three steps: describing the existence of the maximum value and the global solution; proving the control objectives; developing the boundedness and continuity of the control input. The advantages of the proposed control method are validated by testing it on CRH2-A trains in both stochastic and emergency operation environment.