Adaptive Impedance Control for Rehabilitation Robotics based on an EMG-driven Muscle Force Model

Guido Gómez Peña¹; Boris Sullcahuaman Jauregui²; Wilian Miranda dos Santos²; Adriano A. G. Siqueira²

¹ eesc - usp; ² eesc-usp

doi:10.20906/CPS/COB-2015-2216

Resumo

This paper deals with adaptive impedance control of rehabilitation robotic systems based on an EMG-driven patient's activity model. The patient's torque and stiffness are estimated using a simplified and optimized model of muscle forces based on electromyographic (EMG) signals. The proposed adaptive control strategy considers these patient's activity measures to compute the stiffness parameter of the robot's impedance control, such that the patient participation on the therapy is always encouraged. Experimental results of trials with healthy subjects are presented, considering the EMG signals of selected muscles acting at the joint knee during flexion and extension movements defined by a visually-guided task and using a variable impedance knee orthosis.

Palavras-chave: adaptive impedance control; lower limb rehabilitation; rehabilitation robotics; impedance control; electromyography