An efficient numerical approach for calculating the inverse kinematics for robot manipulators

Abstract

SUMMARY This paper compares three numerical methods for obtaining the joint trajectory of a robot manipulator, which causes movement along the desired Cartesian path. The first solves the kinematic equations, which are given in the Jacobian form, while the other two solve the nonlinear kinematic equations directly by using an iterative computational procedure based on the conjugate gradient technique. The computational efficiency of the proposed methods is estimated in terms of the execution time on a VAX 11/750 minicomputer. It is shown that by using the capacity of microcomputers, these methods could be well used in real-time computation.