Dynamic-Growing Fuzzy-Neural Acceleration-Based Impedance Controller for a Lower Limb Rehabilitation Robot

Abstract

Impedance control of a lower limb rehabilitation robot is studied here, using dynamic-growing fuzzy-neural controller in the inner position control loop. Impedance control of rehabilitation robots is important due to their direct contact with patients' body. Impedance control refers to control of the mechanical impedance instead of control the position, speed, or contact forces of a robot separately. Here, a combination of intelligent position controller and acceleration-based impedance controller is proposed and applied to a lower limb rehabilitation robot. The inner loop position controller is an adaptive intelligent self-organizing fuzzy neural network, which holds the calculation optimal. Thus, the resulting proposed structure is an adaptive intelligent model-free controller with swift calculation. The other most important merits of the proposed control scheme are simplicity and interpretability, practicality of the approach, and self-organizing optimality. Finally, several simulation and experimental implementations confirm the efficiency of the proposed impedance control method.