Advertisement

Parallel robot motion planning in a dynamic environment

  • E. -G. Talbi
  • P. Bessière
  • J. M. Ahuactzin
  • E. Mazer
Posters
Part of the Lecture Notes in Computer Science book series (LNCS, volume 634)

Abstract

In order to achieve fast motion planning for a robot we have chosen the genetic algorithms for the following reasons:
  • they are well adapted to search for solutions in high dimensionality search space. The algorithm can be used without reduction of its efficiency for arms with more than six degree of freedom,

  • they are very tolerant to the form of the function to optimize, for instance these functions do not need to be neither differentiable or continuous. They make no assumptions about the problem space that they are searching. We are using them to solve other optimization problems: graph partitioning, quadratic assignment, ...

  • they are easy to implement on massively parallel distributed memory architectures. The parallel algorithm proposed achieve near-linear speed-up.

References

  1. [1]
    D.E.Goldberg, “Genetic algorithms in search, optimization, and machine learning”, Addison-Wesley, 1989 Google Scholar
  2. [2]
    E-G.Talbi, T.Muntean, “A parallel genetic algorithm for process-processors mapping”. Int. Conf. on High Speed Computing 11, Montpellier, M.Durand and F.El Dabaghi (Editors),Elsevier Science Pub., North-Holland, pp.71–82, Oct 1991.Google Scholar
  3. [3]
    J-C.Latombe, “Robot motion planning”, Ed. Kluwer Academic Publisher, 1991.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1992

Authors and Affiliations

  • E. -G. Talbi
    • 1
  • P. Bessière
    • 1
  • J. M. Ahuactzin
    • 2
  • E. Mazer
    • 2
  1. 1.LGI/IMAGInstitut National Polytechnique de GrenobleGrenoble CedexFrance
  2. 2.LIFIA/IMAGInstitut National Polytechnique de GrenobleGrenoble CedexFrance

Personalised recommendations