A minimax approach to robust repetitive learning control

This paper investigates a robust repetitive control scheme based on a minimax problem formulation. As non-robust repetitive control schemes often lead to decreased suppression or amplification of the disturbing signal if the frequency of the disturbance is not exactly known or slowly varying, robust design approaches have to be used in such cases. Results of the new approach are compared to existing approaches and the trade-off between cancellation peformance and robustness is pointed out.