EURAL CONTROLLER OF ROBOTIC MANIPULATOR

Abstract

In this project, the design and simulation of a neural controller for robotic manipulator is carried out. The simulation study is carried out on one-link and two-link manipulators. This project develops a neural controller designing for the proper choice of the transfer function by the using of the back-propagation function “trainlm” for the one link manipulator. It describes the effect of the variation of parameters related to one-link manipulator with the neural controller. The range of the robotic manipulator parameter where a neural controller performs with tracking capability and identification capability is described in detail. This all performance is done on the basis of the coding which is developed by us. MATLAB coding for the neural controller for the two link manipulator is developed. The effects of the learning rate of the training method on the neural controller performance and the learning rate variations are also presented.