A New Scheme for Robust Control of Uncertain Series Elastic Actuator System

Abstract

Series elastic actuator (SEA) is widely used in new generation of robotic systems, particularly rehabilitation robots. The existence of noise and disturbance in the model of most industrial systems is inevitable, where SEA model is also not an exception. Presence of disturbance and uncertainty leads to deviation of the response of system from desired inputs. Kalman filter is a practical method to identify the model and also filtration of noisy data. The approach of this paper is to improve the efficiency of uncertain SEAs in control engineering aspects. Hence, a robust control design including combination of unscented Kalman filter (UKF) and sliding mode control (SMC) is developed for linear force-controlled SEA system. Simulation results show improved performance of the proposed controller to track desired force.