Minimum distances for robot task simulation

Abstract

SUMMARY Minimum distance algorithms allow users of robot simulation programs to maneuver manipulation arms around and between workspace obstacles. Additionally, they can be used to generate configuration maps for path planning of the manipulator “point” through a more abstract configuration space. This paper summarizes an algorithm for determining minimum distances between two polyhedral elements. Examples of configuration space maps and Cartesian stepping techniques demonstrate algorithmic utility for robot path planning. A number of accelerating strategies which depend on a heirarchical spatial representation of manipulator and workspace elements maintain reasonable CPU times for the simulation user.