Numerical Solution Using Nonlinear Least-Squares Method for Inverse Kinematics Calculation of Redundant Manipulators

摘要

In this paper, we present an Inverse Kinematics (IK) algorithm based on the nonlinear least-squares method for redundant manipulators. The Newton-Raphson (NR) method is a common method for IK calculation of redundant manipulators. The NR method, however, causes many problems in terms of joint angle limits, singularity, and solvability. Several methods have therefore been proposed to solve these problems. Most, however, focus only on IK calculation performance when a desired trajectory moves outside of the workspace. A manipulator is required to move continuously, even after a desired trajectory moves outside of the workspace. It is thus also necessary to implement the IK calculation method for bringing a desired trajectory back into the workspace. In this study, we propose a user-friendly method for robotic manipulation that is capable of implementing accurate IK calculation when a desired trajectory must be returned to the workspace.