Robust Control of a Mechatronic Exoskeleton for Motion Rehabilitation

Abstract

Locomotion rehabilitation through robotic exoskeletons requires grave precision and accuracy to realize fruitful results because repetitive movements along specified trajectories are involved. Disturbances, noises and uncertainties impact the rehabilitation process, which results in delayed recovery and unsolicited outcomes. In this context, it is necessary to devise a control strategy for exoskeletons, that can provide effective disturbance rejection, noise reduction and uncertainty compensation. To accomplish this purpose, a new approach for modeling, simulation and robust control of a rehabilitation exoskeleton is presented in this research. First, a mechatronic exoskeleton is proposed, and the mathematical model is obtained using the bond graph modeling technique. Second, the exoskeleton control is formulated as a robust optimization problem and algorithm is used to design a controller. Experimental results verify the effectiveness of the proposed controller for robust response. Modeling through bond graph is a new approach for robotic exoskeletons, which was also presented in our previous research but was limited to; modeling and simulation of an exoskeleton system without including the human leg dynamics. This research is an extension of our previous work and presents modeling and simulation of a practical exoskeleton system (with incorporated humaPn leg dynamics) and implements a robust control strategy for desired response of the mechatronic exoskeleton system.