OPTIMAL CONTROLLER FOR TRAJECTORY TRACKING OF PUMA 560

Abstract

Robots are mechanical devices which can be programmed to perform a large variety of tasks. In recent years robots have been applied in fields from painting to complicated surgeries. PUMA 560 is an important class of robotic manipulator which is applied in varies fields like painting, material handling, welding and many more industrial applications. The basic problem in controlling robotic manipulators is to make them follow a desired trajectory. Robotic manipulators are highly coupled and nonlinear system which makes their control difficult. In this work PUMA 560 has been modeled in MATLAB and its Forward Kinematics and Inverse Kinematics equations are derived. Offline tuning of PD controller for PUMA 560 is done using Genetic Algorithm (GA), Simulated Annealing (SA) and Particle Swarm Optimization (PSO) is done. Moreover hybrid techniques GA-SA and GA-PSO have also been used to tune PD controller and a new switching criteria has been proposed to improve the convergence percent of the hybrid techniques GA-SA and GA-PSO. To overcome nonlinearities of the system due to parameter variation and environmental disturbances a Fuzzy controller has been implemented for PUMA 560. Proper choices of membership functions play a key role in determine the performance of Fuzzy controller. In order to overcome this problem an adaptive Fuzzy controller has been designed which adjusts its membership functions online. A robust controller using Sliding Mode Controller technique has been designed to control PUMA 560. A new distance dependent boundary function has been proposed to reduce the chattering in SMC controllers. The advantages of Fuzzy controllers have also been incorporated by using an adaptive Fuzzy SMC. A SMC with ANFIS has also been proposed to control Robotic manipulator. The performance of the proposed SMC has been verified by comparing its performance on a quarter car model with a simple SMC. Butterfly and Oval trajectories are used as reference trajectories. ITSE is used to compare performance of various controllers. A disturbance signal is added to check the robustness of the controllers.