Fuzzy PID controller for upper limb rehabilitation robotic system

Abstract

The aim of this study is to design a reliable and stable control for upper limb rehabilitation robot for post stroke patents. Position control is a key to perform an effective rehabilitation robotic exercise. Rehabilitation robots are coupled with humans which offer ascent to human-robot interaction dynamics and other parametric uncertainties such as environmental contacts and sensory noise. Model based controllers are not easy to implement and are unable to provide precise control in human-exoskeleton interaction because above mentioned factors don't permit to account exact dynamic model. The efficient control system is required to attain the optimal control performances which takes care the system stability and accuracy. Fuzzy Logic (FL) control has ability and it can provide accurate position/force control with the presence of model uncertainties. This paper presents a comparison between a conventional PID controller and online tuning of PID controller based on fuzzy inference system for extension/flexion movements of shoulder joint.