An Iterative Learning Approach for Train Trajectory Tracking Control1

Abstract

AbstractThis work presents an iterative learning control (ILC) based automatic train operation (ATO) algorithm to address trajectory tracking problem. The train motion dynamics is first described by a modified discrete model with position as its independent variable, since train motion dynamics repeats along position axis more exactly. ILC method is combined with error feedback to achieve trajectory tracking. Meanwhile, the case with input constraints is considered. Rigorous theoretical analysis confirms that proposed algorithm can guarantee the asymptotic convergence of train speed to desired profile along iteration axis. Its effectiveness is further verified through case studies with intensive simulations.