NEURO-FUZZY-GENETIC CONTROLLER DESIGN FOR ROBOTIC MANIPULATORS

Abstract

In this dissertation PID, Fuzzy PID and Neuro-Fuzzy controller for Puma 560 Robotic manipulator is being compared. Efficient Matlab/Simulink Model of manipulator dynamics is designed .Forward and inverse kinematics of Puma 560 is modeled on Matlab/Simulink. PID controller is tuned using different search methods like Genetic Algorithm (GA), Simulated Annealing (SA),Particle Swarm Optimization(PSO) ,Generalized Pattern search (GPS),hybrid GA-GPS and hybrid GA-Nelder-Mead. Fuzzy PID controller is designed and tuned using same search methods as that used for PID controller tuning. ANFIS and Hybrid GA-GPS tuned Radial basis function based Neuro-Fuzzy controller is designed .Performances for all these methods are compared in terms of joint space ITSE and Cartesian space ISE for tracking spiral and butterfly Cartesian space trajectories. Disturbance signal is added to check robustness of controller. 1