Redundancy optimization of hyper redundant robots based on movability and assistability

Abstract

This paper describes a new method to utilize redundancy of hyper redundant robots. In order to use redundancy effectively, the contribution of all actuators for the achievement of the target motions and capability for unpredictable target motions in the future should be considered. From this point of view, two evaluation indices, the movability index of each joint and the assistability index for the target motions are proposed. These indices are formulated with the angle and amplitude distribution of the column vectors of Jacobian matrices of each joint, including the end-effector. Then joint input to optimize these indices simultaneously are derived based on the improved gradient projection method. Motion control simulations and experiments with a planar 10R serial manipulator revealed the effectiveness of the proposed method.