INTELLIGENT CONTROLOF A 5-DOF REDUNDANT ROBOT

Abstract

The objective of this thesis work is to develop an algorithm for solving redundancy resolution and vision control for a redundant robot by optimizing the joint space trajectory using Intelligent control and compare with conventional control technique. Singularity avoidance, Joint limits, and Obstacle avoidance (or Collision– avoidance) are considered as the optimization objective. The problems in redundant manipulator kinematics and control is Inverse Kinematics. This thesis, Self-Organizing Map (SOM) one of the intelligent control to solved the Inverse Kinematics and compared with different traditional approach. The control problems of a redundant robot can be approaches through the dividing the task into sub-task, first priority is to track a trajectory then Singularity avoidance, Joint limits, and Obstacle avoidance is secondary sub-task. The redundancy-resolution algorithm with SOM network generates a smooth minimum joint angle space motion and it is able to optimized effectively trajectory of the joint space. In this thesis paper simulations and experimental have been conducted on 5-DOF Spatial Robotic Manipulator shows effectiveness of the proposed method that is SOM with Weighted Least Norm (WLN) Matrix approach. WLN matrix penalized the joint angle space motion for achieving the desired secondary task. This thesis paper shown that SOM network with WLN Matrix one of the intelligent redundancyresolution scheme and compared with different traditional redundancy resolution approach (Task Priority and the Weighted Least Norm). In this thesis work a visual servoing hardware setup is also designed for a 5-DOF robotic manipulator and addressed the problem of tracking objective in real time.