| Abstract: | In this paper, a robust control scheme to drive a knee‐exoskeleton has been designed to assist people with limited knee movement. Dynamic modeling and parameter identification of the “subject's lower limb‐exoskeleton” has been performed and a sliding mode observer (SMO) developed and integrated in the system's closed loop to provide velocity. Lyapunov's theory allows demonstration of the stability of the system. Experimental tests have been performed with the aid of five voluntary subjects, in sitting position, and during flexion/extension of the knee joint. The performance of the proposed control scheme is compared with those obtained using the classical proportional integrator derivative (PID) controller. These tests track the desired position and velocity trajectories with small tracking errors and evaluate the system's stability and system's robustness against the subject's parameters, variations, and external disturbances. |
