Design and performance analysis of a transparent force control strategy for an exoskeleton

Abstract

Exoskeleton robots are developed for human rehabilitation and power augmentation. Human and robot interaction is an important issue involved with these robots. In this paper, the interaction force reduction is considered leading to the transparency of a lower limb exoskeleton during swing phase of walking. Achieving this goal, a robust Lyapunov based motion control method has been developed. The desired reference signals for motion control are generated using a direct force control approach. Robot accelerations are estimated by an observer to be employed in the control loop. An adaptation procedure has been proposed for the bound estimation of uncertainties. As part of assumptions, the interaction between the robot and human is modeled by a spring-damper set which is reasonably accurate for swing phase. The prototyped exoskeleton parameters have been used for more accurate robot modeling. Finally, the simulation results are presented, indicating good performance of the proposed robot control strategy for a transparent human tracking.