Quasi-Direct Adaptive Control for Non-Minimum Phase Systems

Abstract

An adaptive control algorithm applicable to non-minimum phase, discrete, linear, tiem invariant systems is presented. This algorithm does not require solving a Bezout identity at each sampling step by incorporating two different identification algorithms. The first one identifies the plant's parameters and provides them to the second one which in turn identifies the controller's parameters. Independent tracking and regulation objectives are achieved in the limit with bounded plant input and output for a range of initial plant parameters estimates.