Iterative learning control for variable setpoints, applied to a motion system

Iterative Learning Control (ILC) is a known technique for improving the performance of systems or processes that operate repetitively over a fixed time interval. ILC generates a feedforward signal effective for providing good tracking control. However, there still exist a number of problems which hinder extensions of ILC schemes. The major obstacle is perhaps the requirement that the trajectory (or repetitive disturbance) must be strictly repeatable over operations. ILC has also liability to deal with stochastic effects. This paper presents the design and the implementation of a time-frequency adaptive ILC that is applicable for motion systems which execute the same kind of repetitive tasks. For the motion system, we show that the adaptive algorithm we propose leads to design one (learned) feed-forward signal suitable for different setpoints. We demonstrate that, when implementing time-frequency adaptive ILC, very good time performance (tracking errors) is obtained. The proposed algorithm converges faster than standard ILC. With time-frequency adaptive ILC noise amplification is reduced.