Adaptive iterative observer based on integral backstepping control for upper extremity exoskelton robot

Abstract

Assisted control seeks to help subjects to perform physical movements of the body, which they cannot do by themselves. Passive rehabilitation therapy is very important and vital after the stroke accident. In this functioning mode, the subject is completely passive during the movement. The robot brings the injured upper arm of the patient to perform repetitive therapeutic exercises. In this case, the subject's force and the uncertainties caused by repetitive motion, such as mechanical and actuator fatigue, are considered as external disturbances that negatively influence the performance of the exoskeleton robot. To ensure the stability, robustness, and accuracy of the robot, a robust iterative observer based on nonlinear integral backstepping control was implemented with designed exercises performed by subjects. Experimental results show the effectiveness of the proposed control to deal with the external force and repetitive/periodic uncertainties.