Deterministic Path Planning Algorithms for Manipulator Chains and Mobile Manipulators

Abstract

This paper proposes 2 schemes for calculating paths of manipulators that convey objects in a large space. The one is for a manipulator chain and to relay objects to distant goal positions finds attitudes of manipulator pairs that exchange objects. The other is for a mobile manipulator and calculates paths of the manipulator and the base vehicle simultaneously by regarding the vehicle with nonholonomic constraints as an end of the base arm. Both schemes exploit Backtrack free path planning algorithm (BFA), therefore they are deterministic, i.e. they determine existence of paths and calculate paths without fail if paths exist. Also, their computation volumes are predictable and increase only linearly with numbers of arms.