Model Reference Adaptive Control via Online Virtual Reference Feedback Tuning under Unmatched Controller Structures

The study considers model reference adaptive control for a single-input, single-output discrete-time system using an idea of the virtual reference feedback tuning (VRFT). The standard VRFT resolves a model reference control problem using one-shot experimental input-output data instead of a mathematical controlled plant model. The study provides an online tuning method for controller parameter to follow a prescribed reference model output for the closed-loop output. The main feature in the study is to achieve asymptotically tracking to step-type reference signals under unmatched controller structures. The effectiveness of the proposed method is shown through a numerical example.