Intelligent control for lower limb rehabilitation system

Abstract

This paper proposes the design and control of passive and active motion of lower limb rehabilitation system. The dynamic model of the rehabilitation machine is firstly derived for further controller design. Then, an intelligent sliding-mode control (ISMC) system which involved recurrent Hermite neural network (RHNN) estimator to estimate the unknown external disturbance and uncertainty is proposed to track the angular position and velocity in the passive motion. In the active motion, the position-based impedance control is implemented to achieve human-machine compliance.