ADAPTIVE CONTROL OF ROBOT ARM so A NEURAL NETWORK APPROACH

Abstract

The precise control of robot manipulator to track the desired trajectory is a very tedious job and almost unachievable with the help of conventional controller. This task is achievable to a certain limit with the help of adaptive controllers but these also have their own limitations of assuming that the systems parameters being controlled change relatively slowly. Many algorithms have been proposed from time to time to minimize this deficiencies. Interfacing of neural network with the robot manipulator is one of the means of getting the rapid convergence of actual trajectory to the desired trajectory. In our work, we have tried to implement the neural network in the adaptive controller with the help of a neural network to control the robot arm. Results have been compared with the conventional P.D. controllers. A comparative study of neural network based controller without adaptive control and a neural network based controller with adaptive control is also given here