AFDM Approach for Experience Inclusion in Learning Controllers

In this paper a new method of experience inclusion in iterative learning controllers (ILC) is proposed. Approximate fuzzy data model (AFDM) technique has been adopted for the process of initial input selection. Instead of zero initial input assumption as in most of the ILC algorithms, in this paper the idea of using past trajectory tracking experiences in the selection of initial input for tracking a new trajectory tracking task has been highlighted. Performance of the proposed AFDM based ILC approach, on initial error reduction and error convergence issues are proved. Comparison with existing local learning technique on the selection of initial input for ILC algorithm proves the efficacy of the proposed AFDM based method