Robust analysis of a genetic algorithm based optimization method for real-time iterative learning control applications

Optimality based iterative learning control (ILC) schemes have been proved to be quite popular among the researchers. However most of these algorithms cannot cope with nonlinearities and hard constraints in the problem domain. This is a severe limitation since ILC theory is mainly applied to repetitive systems, which can be nonlinear. Thus, in order to overcome this problem a genetic algorithm (GA) based optimization method for ILC systems was introduced in (Hatzikos et al., 2003a) and (Hatzikos et al., 2003b). To our knowledge, this is the first time that a GA controller is introduced into ILC. This approach proved to be able to overcome all the shortcomings mentioned earlier and furthermore the simulation results have indicated satisfactory performance when the algorithm was applied to various dynamical systems. In this paper the idea of applying the proposed GA-ILC algorithm in real-time is investigated and a simulation structure that can be implemented into real-time applications is proposed. Furthermore the robustness of the proposed process is analysed in the existence of model uncertainties and unwanted disturbances. Simulations results are used to illustrate the performance of the proposed GA-ILC scheme.