Finite-Iteration Adaptive ILC for Automatic Operation of High-Speed Trains

Abstract

Iterative learning control (ILC) is a typical method for automatic high-speed train (HST) control. However, the existing ILC methods require that the operation/iteration number of the HST must approach infinity so as to guarantee perfect tracking. This requirement is unrealistic for practical automatic train control. In this paper, a new finite-iteration adaptive ILC (FIAILC) method is proposed and a new framework of composite energy function-based finite-iteration convergence theory is given for the first time. The proposed FIAILC can make the tracking error not only converge to zero as the iteration number go to infinity, but converge to an arbitrarily predefined tracking control precision after a finite number of iterations. This finite iteration number can be obtained theoretically according to the predefined tracking control precision as well as the tunable gain and initial values of the proposed controller. The result is further extended to the HST operation system with speed constraint and a constrained FIAILC is designed accordingly. Moreover, three simulation cases on a practical train operation system similar to China Railway High-speed (CRH)-3 HST are given to show the effectiveness of the proposed FIAILCs.