Sliding Control Algorithm with a Neural Network for Exoskeleton Electric Drive System

The article presents a sliding control algorithm with a neural network to eliminate vague parameters and external disturbances in studying the dynamic modeling for the lower extremities of the exoskeleton. The sliding control is used to achieve rapid convergence over the final time and suppress the disturbance. A neural network appraiser is integrated into the control system for evaluating the parameters and model of the control object. The timid system is able to respond to a change in the external disturbances and the parameters of the dynamic modeling of the exoskeleton. The stability of the proposed regulator and the effectiveness of tracking the control system are established by Lyapunov's theorem of stability.