A new pseudoinverse for manipulator collision avoidance

Abstract

This paper presents a control system architecture for robot manipulator suitable for on line path planning in a cluttered environment. The proposed technique fits for both redundant and non redundant manipulators. We use potential fields in order to control the joint velocities to make the manipulator escape from obstacles. We also propose a new weighted pseudoinverse matrix that improves the manipulator capability to find feasible paths moving around obstacles and passing through narrow corridors without calculating the manipulator mass matrix. The proposed solution is tested on two manipulators: a 7 degree-of-freedom (DOF) planar manipulator and a 7 degree-of-freedom spatial manipulator both moving in a cluttered environment. © 2011 IFAC.