Efficient Sub-Optimal Inverse Kinematic Solution for Redundant Manipulators

Abstract

An efficient algorithm for Inverse Kinematic computation is presented. This algorithm is sub-optimal compared to conventional forms of inverse kinematic solutions, but it solves the problem for redundant high degrees of freedom manipulators efficiently. The presented algorithm relies on a deep understanding of motion planning in human arm. It was observed that the human arm, as a redundant manipulator, starts the motion by alignment the forearm towards the target and let the following joints flow as the hand is moving towards the target. Therefore, the torque sent to each joint actuator is proportional to the instantaneous contribution of that joint in driving the end-effector towards the target. This methodology can be used for any degrees of freedom and any manipulator geometry as long as the forward kinematics is provided. Simulation studies and experimental results are provided to verify the similarity of the resulting motion of this algorithm with natural human arm motion.