Elite Based Multiobjective Genetic Programming in Nonlinear Systems Identification

Created by W.Langdon from gp-bibliography.bib Revision:1.3872

@Article{Patelli:2010:AECE,
  title =        "Elite Based Multiobjective Genetic Programming in
                 Nonlinear Systems Identification",
  author =       "Alina Patelli and Lavinia Ferariu",
  journal =      "Advances in Electrical and Computer Engineering",
  year =         "2010",
  volume =       "10",
  number =       "1",
  pages =        "94--99",
  month =        feb,
  keywords =     "genetic algorithms, genetic programming, evolutionary
                 algorithms, multiobjective optimization, nonlinear
                 system identification",
  ISSN =         "1582-7445",
  URL =          "http://www.aece.ro/abstractplus.php?year=2010&number=1&article=17",
  broken =       "http://www.doaj.org/doaj?func=openurl\&genre=article\&issn=15827445\&date=2010\&volume=10\&issue=1\&spage=94",
  DOI =          "doi:10.4316/AECE.2010.01017",
  publisher =    "Stefan cel Mare University of Suceava",
  bibsource =    "OAI-PMH server at www.doaj.org",
  oai =          "oai:doaj-articles:7de67791d79585b8b25988d832c27761",
  size =         "6 pages",
  abstract =     "The nonlinear systems identification method described
                 in the paper is based on genetic programming, a robust
                 tool, able to ensure the simultaneous selection of
                 model structure and parameters. The assessment of
                 potential solutions is done via a multiobjective
                 approach, making use of both accuracy and parsimony
                 criteria, in order to encourage the selection of
                 accurate and compact models, characterized by expected
                 good generalization capabilities. The evolutionary
                 process is implemented from an elitist standpoint, and
                 upgraded by means of two original contributions, namely
                 an adaptive niching mechanism and an elite clustering
                 procedure. The authors have also suggested a set of
                 enhancements to aid the genetic operators in
                 effectively exploring the space of possible model
                 structures. In symbiosis with the customized genetic
                 operators, a QR local optimization procedure was
                 integrated within the algorithm. It exploits the
                 nonlinear, linear in parameter form that the working
                 models are generated in, for providing a faster
                 parameter computation. The performances of the proposed
                 methodology were revealed on two applications, of
                 different complexity levels: the identification of a
                 simulated nonlinear system and the identification of an
                 industrial plant.",
}

Genetic Programming entries for Alina Patelli Lavinia Ferariu

Citations