ITERATIVE LEARNING CONTROL FOR DISCRETE-TIME-SYSTEMS WITH EXPONENTIALRATE OF CONVERGENCE

An algorithm for iterative learning control is proposed based on an op timisation principle used by other authors to derive gradient-type alg orithms. The new algorithm is a descent algorithm and has potential be nefits which include realisation in terms of Riccati feedback and feed forward components. This realisation also has the advantage of implici tly ensuring automatic step-size selection and hence guaranteeing conv ergence without the need for empirical choice of parameters. The algor ithm achieves a geometric rate of convergence for invertible plants. O ne important feature of the proposed algorithm is the dependence of th e speed of convergence on weight parameters appearing in the norms of the signals chosen for the optimisation problem.