Disturbance-Observer Based Tracking Control of Industrial SCARA Robot Manipulators

Abstract

This paper presents a new feedback-feedforward tracking control method for serial manipulators in presence of modeling uncertainties, and both known and unknown disturbances. To improve the tracking performance, the proposed scheme is composed of three main control components: feedforward control, disturbance-observer-based control and feedback control. The feedforward action is designed to account for the effects of the reference signal, considered as known disturbance, on the tracking error dynamics. The disturbance-observer-based control law is proposed to deal with the modeling uncertainties as well as unknown disturbances. Using the concepts of V-stability and H∞ control, the feedback component is designed to satisfy some predefined closed-loop specifications, aiming to improve the tracking performance. The effectiveness of the new tracking control method is demonstrated with a multibody model of a 2-DoF SCARA robot manipulator. In particular, a comparison with two classical tracking control methods is performed to emphasize the interest of our method for robotics control.