Robust Control Design for a Dynamical System with Hard Nonlinearities : Application to Friction Compensation in a Robot Joint

Abstract

A new methodology to design an optimal state feedback controlled to solve problems mainly caused by the presence of nonlinearities in motion control systems. Usually these nonlinearities are isolated, modeled and compensated by various methods. A theoretical background of the proposed generalized method is developed first. The method rely on a local modeling approach applied to give a simpler description of the system dynamics inside a certain set of operating conditions, then a stabilizing state feedback controller is designed based on Linear Matrix Inequalities and applied to robot joint control to enhance tracking performances. Simulations and experimental results are used to verify the efficiency of the proposed method