Paper:
Language: English:

Object Reconfiguration with Dextrous Robot Agents

Gábor Vass^, Béla Lantos^, and Shahram Payandeh^**

^*Department of Control Engineering and Information Technology, Budapest University of Technology and Economics
H-1117 Magyar Tudosok krt. 2, Budapest, Hungary
^**Experimental Robotics Laboratory, School of Engineering Science, Simon Fraser University
8888 University Drive, Burnaby, British Columbia, Canada V5A 1S6

Received: September 22, 2005

Accepted: October 28, 2005

Keywords: robot, object reconfiguration, simulated annealing

Journal ref: Journal of Advanced Computational Intelligence and Intelligent Informatics, Vol.10, No.2 pp. 234-240, 2006

Abstract

This paper addresses an object manipulation planning algorithm for dextrous robot systems consisting a multifingered hand and a robotic manipulator. A method has been developed for object reconfiguration design.
The result is a new algorithm using artificial intelligence based on simulated annealing and A* search. The upper level of the manipulation system, the global planner generates the motion of the object. The lower level, the local planner deals with the motion of the agents relative to the object and the design of the contact forces. The local planner is based on simulated annealing, thus the the local minima can be avoided in the energy function of the motion with high probability.Application of the algorithm has been discussed for three robot arms.

[Notice]
* "Preview" is the first 2 pages of the article. You don't need the registration.
* To read the PDF file you will then need to download and install the Adobe Reader.
Adobe Reader is free and available for download here: