A Robust Unknown Input Observer Design applied to an Industrial Pilot Plant

This paper presents a robust unknown input observer (UIO) design procedure applied to an industrial pilot plant. The proposed design investigates the possible advantages of a proportional-integral UIO (PI-UIO) to solve the problem of states and disturbances estimation when the industrial plant is subject to parametric uncertainties. Existence conditions to obtain the PI-UIO structure are provided, as well as Linear Matrix Inequalities (LMIs) conditions to ensure the robust stability for the closed-loop system. A practical comparison using a proportional UIO widely used in the literature and the proposed approach is made. Finally, experimental results obtained with the application of the two UIOs strategies are presented.