Optimization of a proportional derivative (PD) fuzzy controller using the particle swarm optimization (PSO) technique for a 3DOF robot manipulator

Abstract

This paper deals with the optimization of a proportional derivative (PD) fuzzy controller using the particle swarm optimization (PSO) technique. More precisely, we conduct a comparative study to access the performance of the PD fuzzy controller by optimizing its gain and its structural parameters separately, the optimization will done offline; thus leading two distinct fuzzy controllers. The two controllers are applied to control a 3DOF PUMA560 robot manipulator. The structure of the proposed controller is composed of a symmetric fuzzy set ranging between [-1 1] with triangular symmetric membership function. It is shown that the fuzzy controller with optimized gain performs better, compared to the one with optimized parameters, in ideal conditions without noise. Additionally, it is easy and simple to program and implement since there are only a few gain variables to optimize. Also, the execution time is much lower than that required by the optimized parameter fuzzy controller. On the other hand, the optimized parameter controller performs better in the presence of noise in some joints.