Robot manipulation through inverse kinematics

摘要

Inverse kinematics of a robot is very essential to find the joint variables that satisfy the desired pose of the robot during its manipulation. This is used in controlling the robot position, animation of the robot, etc. In this paper, step-by-step explanation and comparison of two widely used methods, namely, inverse kinematics and Jacobian inverse methods, for robot manipulation are presented. For this purpose a six degrees-of-freedom wrist-partitioned industrial robot KUKA KR5 Arc was used to illustrate the methods. A novel approach has been proposed for selecting the appropriate set of joint angles among the several inverse kinematic solutions. It is based on weight of each link and manipulability. The comparison of these approaches for linear and circular trajectory is presented. Their advantages, limitations, applications, and computations involved are also highlighted.