Robust Iterative Learning Control on Finite Time Intervals
Abstract
This paper considers the problem of designing optimal robust Iterative Learning Control (ILC) algorithms for LTI processes. Given a multiplicative uncertainty model of the plant, learning operators are designed to minimize the ultimate tracking error and convergence rate, while guaranteeing robust convergence. The optimal learning operators are shown to be noncausal. After the controller is optimized in the frequency domain, a finite-time implementation of the algorithm is shown to achieve the same performance and robustness.
