Minimum-effort redundancy resolution of robot manipulators unified by quadratic programming

Abstract

This paper presents the latest result that the minimum-effort redundancy resolution of robot manipulators with joint physical limits is unified into a quadratic-programming (QP) problem formulation with different coefficient matrices and vectors defined for different schemes. Such a general QP formulation is subject to equality, inequality and bound constraints, simultaneously. Motivated by the realtime solution to such robotic inverse-kinematics problems, the standard QP optimization routines and primal-dual neural network based on linear variational inequalities (due to its simple piecewise-linear dynamics and higher computational efficiency) are investigated in this paper. The QP-based unification of robots' redundancy resolution is substantiated by a number of computer-simulations of PUMA560, PA10, and planar arms.