Linear Matrix Inequality Based Data-driven Disturbance Observer Design: Application to a Non-Minimum Phase Motion Stage

Abstract

Disturbance observers are widely used in the industrial field to mitigate the influence of disturbance, and the trial-and-error method is commonly employed in their design. Herein, a fully parameterized bandwidth-maximized disturbance observer problem was formulated into a frequency response data-based convex optimization problem by transforming the original constraints into their convex form. Furthermore, a comparison with the trial-and-error method verified the feasibility of the proposed method.