Inverse-free D1G1 solution to acceleration-level inverse kinematics of redundant robot manipulators

Abstract

The forward kinematics (FK) problem for robotic manipulators has a mathematically analytical solution, while the inverse kinematics (IK) problem is computationally expensive, and the closed-form expression of the solution does not always exist. In recent years, many approaches have been proposed to solve such a complicated IK problem, such as the traditional pseudo-inverse based solution. However, there are many limitations when using the pseudo-inverse based solution. Therefore, how to effectively avoid the Jacobian pseudo-inversion is very important and meaningful. Motivated by this, an important branch of dynamics methods, i.e., the gradient dynamics (GD) method, is thus applied to developing a novel inverse-free D1G1 solution at the joint-acceleration level for solving the IK problem of redundant robot manipulators. Furthermore, simulation results based on a three-link redundant robot manipulator substantiate the effectiveness and accuracy of the proposed inverse-free D1G1 solution.