Iterative Learning Control for Periodic Systems Using Model Predictive Methods with Adaptive Sampling Rates

This paper addresses iterative learning control (ILC) for periodic systems using model predictive and optimization methods to redesign trajectories and reject periodic disturbances. Stability and optimality of these optimization methods is analysed and illustrated on simulations. The additional prospects of the optimization formulation (e.g. including energy costs, system identification) referred to the trajectory planning are accentuated. To reduce the calculation effort of the optimization algorithm a variable and adaptive sampling period is introduced. The advantages compared to classical ILC methods especially in consideration of constraints are presented.