Adaptive learning control for non-minimum phase linear systems

The aim of this paper is to show the existence of a local solution to the output tracking problem for uncertain linear systems in which the output reference signal is periodic with known period. An output-feedback adaptive learning control is designed which relies on the Fourier approximation theory and on two adaptive observers. It guarantees, under suitable conditions, exponential output tracking and exponential estimation (with sufficiently high precision) of the: i) constant system parameters; ii) periodic input reference signal. No minimum phase assumption is required. Simulation results illustrate the effectiveness of the proposed approach.