Improving the performance of classical control design by using add-on simple adaptive control

Abstract

The Simple Adaptive Control (SAC) controls an augmented plant that comprises the true plant with Parallel Feedforward . The Almost Strictly Positive Real (ASPR) property of the augmented plant guarantees the stability of the tracking loop and leads to asymptotic following. Prior publications have shown that, based only on the prior knowledge on stabilizability properties of systems (usually available), the Parallel Feedforward Configuration (PFC) allows adaptive control of realistic
systems, even if they are both unstable and non-minimum phase. However, it was commonly thought that the addition of the PFC requires a price when compared with good linear time invariant (LTI) designs that do not use any addition to the plant. The paper shows that the use of a properly designed SAC with PFC, as Add-On to LTI system design, improves the performance. Although SAC doesn't directly control the true plant tracking error, it always gives smaller tracking error and reduced sensitivity to plant disturbance and output measurement noise with respect to the best LTI controller.