Feedback Error Learning by means of the Pre-filter State

Abstract

This paper proposes a scheme for the Feedback Error Learning (FEL). Conventional FEL schemes apply a pre-filter to the reference signal, in order to compensate the delay (relative degree) of the plant. In this case, however, the pre-filter parameters, which is known, have to be learned as well as the plant parameters. The proposed method improves this redundancy by using the pre-filter state. Simulation results illustrate the effectiveness of the proposed method.