On The Computational Efficiency Of Four Numerical Techniques For Solving The Practical Inverse Kinematics Problem Of Redundant Manipulators

Abstract

In this paper we compare the computational efficiency of four methods for solving the practical inverse kinematics problem of redundant manipulators. Two methods use recursive approaches for sequential determination of feasible joint rates while the other two methods derive complete joint rate solutions at first and then check the joint kinematical constraints. The methods were implemented on a 386 microcomputer. Issues of improving efficiency by changing velocity reference frames and workspace decomposition were examined.