A solution to the velocity control of redundant robots

Abstract

Abstract Flexible and dexterous manipulation of modern industrial applications requires the use of robots with extra (more than six) axes of motion (degrees of freedom). Thus the kinematic and control considerations of redundant robots are currently of increasing interest to both theorists and practitioners. For this type of robots, a closed-form solution of the inverse kinematic problem is not possible, and the classical approach is to use the pseudoinverse of the Jacobian matrix together with an extra criterion function. This paper presents an iterative technique for velocity control, which is suitable for redundant robots with a maximum of 11 d.o.f. This technique, which splits the Jacobian inversion problem into two subproblems (wrist position and wrist orientation), provides an exact solution for the arm tip position and an approximate solution for the wrist orientation. The convergence of the procedure is examined and a simulated example is included.