Feedback Control Topology of n-DOF Robotic Manipulator and Optimal Positioning of End-Effector using PSO

Abstract

In this work, optimal control of n-degrees of freedom (DOF) robotic manipulator, using particle swarm optimization (PSO), is proposed. Based on an arbitrary model of robotic manipulator, final position of end-effector was computed using four Denavit-Hartenberg (DH) parameters, associated with each link by forward kinematics. A feedback control topology was implemented that reduced the squared sum error, between the desired end-effector position and obtained end-effector position, in each iteration of PSO, within 3 dimensional cartesian space. The PSO provided optimized values of link parameters i.e. joint angles for revolute joints and link lengths for prismatic joints, so that the end-effector could be perfectly positioned at desired coordinate. This algorithm was successfully implemented in MATLAB/Simulink and tested on several arbitrary robotic manipulator with n-DOF system and it provided satisfactory result on controlling and positioning the end-effector.