Learning strictly positive real linear systems with uncertain parameters and unknown input disturbances

Abstract

A learning control scheme is derived for the learning of strictly positive real (SPR) linear systems. By using the SPR lemma, it is shown that the exponential rate of convergence of the learning system can be set arbitrarily fast by choosing the learning control gain without the use of any normed inequality condition. Copyright (C) 1996 Elsevier Science Ltd.