Design of an optimal combination of feedback and iterative learning controllers

This paper describes how an optimal learning controller can be designed to work with an existing feedback controller so that the combination of feedback and iterative learning control is optimal. Optimality is defined as minimizing a cost function that dictates how the tracking error is reduced from repetition to repetition. The combination strategy leverages the inherent advantage of feedback control to handle unexpected disturbance and noise, and the advantage of iterative learning control to correct for repeating error that the feedback controller fails to eliminate, and any other unknown repeating disturbance in the system.