Firefly Optimised PID Control for Upper Extremity Rehabilitation Robot

摘要

This paper proposes a meta-heuristic technique to optimize controller parameters for upper extremity rehabilitation robot. The exoskeleton used in this study is a three-degrees-of-freedom system facilitating shoulder and elbow joint movements. The robot exoskeleton is designed for facilitating shoulder abduction/ adduction, extension/ flexion, horizontal external/ internal rotation and elbow extension/ flexion. Zeigler-Nichols is a conventional technique for setting the PID parameters produces large overshoots, high rise-time and settling-time in the system. To address such issues nature-inspired Firefly optimization algorithm for the parameterizing 2-DOF-PID controller has been proposed. A comparative analysis between two algorithms has been done for a 3-DOF upper extremity rehabilitation robot. For analysis, four different objective functions ISE (integral square error), ITSE (integral time square error), IAE (integral absolute error) and ITAE (integral time absolute error) have been used. The results of the analysis showed that Firefly algorithm with ITAE as objective functions performs better in terms of overshoot, rise-time and settling-time.