LPV state-feedback control of a robotic manipulator via LMI optimization

Abstract

This paper presents an LPV model-based controller synthesis case study on a 2-DOF robotic manipulator. A polytopic LPV formulation was chosen with two scheduling variables. An <tex>$\boldsymbol{H}_{\infty}$</tex> control strategy was used for the reference tracking objective which was implemented through LMI optimization and was compared to a nominal <tex>$\boldsymbol{H}_{\infty}$</tex> controller and feedback linearization extended with PD reference tracking criterion. The solution was tested on a real-life robotic arm located at BME Robotics Laboratory. Numerical evaluation was based on a joint measure of performance and robustness.