Task-Based Mass Optimization of Reconfigurable Robotic Manipulator Systems

Abstract

This work develops a method for implementing task-based mass optimization of modular, reconfigurable manipulators. Link and joint modules are selected from a library of potential parts and assembled into serial manipulator configurations. A genetic algorithm is used to search over the potential set of combinations to find mass-minimized solutions. To facilitate the automatic evaluation required by the genetic algorithm, Denavit-Hartenberg parameters are automatically generated from module combinations. Reconfigurable manipulators are shown to be lighter than fixed-topology manipulators, demonstrating the potential utility of reconfigurable robotics technology for mass reduction in space robots.