Distributed Light Brightness Control based on cuSASGP

Created by W.Langdon from gp-bibliography.bib Revision:1.3973

@InProceedings{Ono:2016:CEC,
  author =       "Keiko Ono and Yoshiko Hanada",
  title =        "Distributed Light Brightness Control based on
                 {cuSASGP}",
  booktitle =    "Proceedings of 2016 IEEE Congress on Evolutionary
                 Computation (CEC 2016)",
  year =         "2016",
  editor =       "Yew-Soon Ong",
  pages =        "838--845",
  address =      "Vancouver",
  month =        "24-29 " # jul,
  publisher =    "IEEE Press",
  keywords =     "genetic algorithms, genetic programming, GPU",
  isbn13 =       "978-1-5090-0623-6",
  DOI =          "doi:10.1109/CEC.2016.7743878",
  abstract =     "This paper describes an approach for controlling light
                 luminance using the CUDA-based Self-adaptive
                 Subpopulation Model of Genetic Programming (cuSASGP).
                 The method involves the evolution of a genetic
                 programming lighting control rule for the ceiling
                 lights in an office room to satisfy different
                 brightness requirements at each desk and reduce
                 electric power consumption. Although the lighting
                 control problem has many local minima, cuSASGP uses
                 solution features to construct an appropriate island
                 formation that avoids these local minima. Thus, an
                 approach for controlling light luminance based on
                 cuSASGP could be expected to improve performance in
                 terms of avoiding local minima and genetic diversity.
                 We first define the lighting control problem for
                 ceiling lights, and propose a genetic programming
                 approach. We implement five types of functional nodes
                 and three types of terminal nodes. Moreover, we verify
                 that genetic diversity can be achieved by adopting
                 subpopulation models such as the island method and
                 cuSASGP in the lighting control problem. For two
                 different environments, we demonstrate that the
                 proposed genetic programming approach can optimize an
                 appropriate lighting pattern that satisfies both user
                 requests and energy constraints, and that use of
                 cuSASGP enhance genetic diversity.",
  notes =        "WCCI2016",
}

Genetic Programming entries for Keiko Ono Yoshiko Hanada

Citations