Model reference adaptive control for multivariable nonminimum-phase systems using a periodic scheme

A model reference adaptive control problem for multi-input multi-output discrete-time systems of nonminimum phase is considered using the periodic control method. Appropriate periodic feedback gains for attaining an exact model matching are characterized in terms of the past input and output signals through the inverse representation of the periodic control system. This representation is then used to directly identify the appropriate periodic feedback gains. Employing a modified least-square adaptation scheme, the global asymptotic stability of the proposed model reference adaptive periodic controller is established. These results can be directly applied to an adaptive output tracking problem