Genetic Algorithm Based PD Control and Fuzzy Logic Control of a Two Link Robot

Abstract

This paper presents two schemes for the control of robotic manipulator systems. The first method is based on proportional plus derivative (PD) control with the PD controller gain parameters optimized via Genetic Algorithm (GA). The cost function (or performance index) to be minimized is the time taken by the system to achieve a final desired position. The second scheme is a fuzzy PD controller whose inputs are error and derivative of error. The fuzzy rules have been formulated with common behavior of the system under certain input conditions and defuzzification has been carried out to obtain crisp joint torque values. The configuration considered in the present paper is a two robotic link system.