Maneuvering redundant manipulators in cluttered environments

Abstract

This paper presents a path planning approach for kinematically redundant manipulators working in cluttered environments. The focus of this paper is on the challenges of spatial manipulators with large number of degrees of freedom (dof's). For obstacle avoidance in three-dimensional workspaces, a penalizing scheme is proposed. Based on variational calculus, this approach involves the formulation of a functional, which is minimized to get an optimal path. The minimization of the objective function results in a boundary value problem (BVP), which is solved using relaxation method. A variety of examples are presented to demonstrate the approach and to show its efficiency in providing good quality paths. The paper also includes systematic procedures for initial path generation and the inverse kinematics for serial redundant manipulators. These schemes reduce the preprocessing task to a great extent and add to the efficacy of the proposed method.