A Novel Strategy with Multiple Models to Improve Performance of Adaptive Iterative Learning Control

This paper presents a new strategy for Adaptive Iterative Learning Control (AILC), focusing on monotonic convergence of the tracking error. First, we use parameter estimation algorithms to construct a model of the system. Next, we use these estimates to choose design parameters for a time-varying gain in the control update law without violating monotonic convergence of tracking error. Further, we show the performance improvement in tracking error and control energy with multiple parameter estimation models. The efficacy of the proposed strategy is demonstrated using simulation examples.