Adaptive Tracking Control With Disturbance Rejection for Cable-Driven Exoskeleton Robot With Compliant Actuators

Abstract

The cable-driven upper-limb exoskeleton robot with compliant actuators presents notable advantages when applied to rehabilitation training, yet introduces challenges in position tracking control. Traditional control methods inadequately address hysteresis issues caused by cable drives and the performance degradation resulting from compliant actuators. To solve these common issues in such robotic systems, this study introduces a unified dual closed-loop control strategy that aims to enhance system tracking performance. In the outer loop, an adaptive hysteresis compensator is intricately designed to address position errors induced by cable-driven motion from a kinematic perspective. Simultaneously, the inner loop employs a high-order controller utilizing the backstepping method to control the dynamics of the robotic system with flexible joints effectively. Additionally, a dedicated high-order super-twisting observer (STO) is integrated to estimate unknown dynamic models and external disturbance, enabling the generation of disturbance rejection items for the inner-loop controller design. Experimental results validate the efficacy of the proposed control strategy, demonstrating commendable tracking performance when compared with alternative methods.