Optimal Iterative Learning Control for Discrete Linear Time-Varying Systems with Varying Trial Lengths

Abstract

In this study, an optimal iterative learning control scheme is designed for discrete linear time-varying systems with varying trial lengths. Since the trial lengths are different from iteration to iteration, the theoretical information is used to compensate the absent section at the current iteration. In order to obtain the fast convergence speed, an iteration performance index is to maximize the declining quantity of the tracking error of two adjacent iterations, and the argument is the iteration-time-varying learning gain vector. The bigger the difference value, the faster the convergence speed. Furthermore, the optimal iterative learning control scheme is adaptive to the tracking error, which can guarantee the convergence of the tracking error. Numerical simulations are shown to verify the effectiveness of the proposed scheme.