Kalman filter based iterative learning control for discrete time MIMO systems

A new iterative learning control (ILC) algorithm for linear, time invariant, multi-input multi-output, dynamical systems in discrete time is introduced. The gain of the well known P-type ILC algorithm is updated, in every iteration, by solving a discrete Riccati equation that requires the knowledge of the first Markov parameter matrix CB of the dynamical system. The proposed Kalman Filter based learning algorithm is implementable block-wise which is similar to the, recently introduced, Luenberger observer based ILC algorithm.