Iterative Learning Control for a Type of Modified Smith Predictor

Abstract

This article proposes a novel iterative learning control (ILC) design for a type of modified Smith predictor, in particular, to control a single-input single-output unstable plant or integral process with a time delay. Frequency domain techniques are applied to synthesize the learning control law, and a sufficient condition is given to ensure robust convergence of the tracking error. Robustness of the system is studied, considering a multiplicative uncertainty. Moreover, the impact of the load disturbance over successive iterations is investigated as well. To this end, a numerical example is given to demonstrate the efficacy of the proposed approach.