Joint velocity compensation for kinematically redundant manipulators

This paper concerns an approach for the inverse kinematic resolution of kinematically redundant manipulators when following a desired path. It can be called joint velocity compensation (JVC). Its principle is employing the redundancy to ensure uniform velocity or velocities with desired simple patterns for certain selected joints, while the remaining joints move with variable velocities as required by a given task. As many joints as the order of redundancy can have predefined velocities. From a control point of view it would be easier and more precise to maintain a uniform velocity for a joint rather than one that varies. Compared with other techniques, this approach has the advantages that it conserves the cyclic behaviour of a manipulator in the joint space for a closed path in the task space, less arithmetic operations are involved and the control of the manipulator as a whole becomes easier. Because of the physical and task constraints the choice of these selected velocities are not arbitrary. This paper elaborates the matter on how a desired velocity can be assigned to a joint such that all the constraints are respected and the task is performed as desired. Simulation results of the implementation of this approach to a redundant robot arm with 7-DOF are presented