Publication detail

Waveguide Synthesis by Genetic Algoritms with Multiple Crossover

JILKOVÁ, J. RAIDA, Z.

Original Title

Waveguide Synthesis by Genetic Algoritms with Multiple Crossover

English Title

Waveguide Synthesis by Genetic Algoritms with Multiple Crossover

Type

conference paper

Language

en

Original Abstract

In this paper the waveguide design using of the genetic algorithms as global optimization technique is described. The influence of the multipoint crossover and mutation operators on the convergence is experimentally examined and discussed. The cross section shape of waveguides is optimized with respect to the bandwidth of single mode to be maximized. The influence of singlepoint crossover and multipoint crossover operators on the waveguide synthesis was investigated on the H-shaped waveguide simulation (2 degrees of freedom) and on the comb-like waveguide (2, 3 and 4 degrees of freedom). The waveguide models were simulated in Comsol Multiphysics. The optimization algorithm was implemented in MATLAB.

English abstract

In this paper the waveguide design using of the genetic algorithms as global optimization technique is described. The influence of the multipoint crossover and mutation operators on the convergence is experimentally examined and discussed. The cross section shape of waveguides is optimized with respect to the bandwidth of single mode to be maximized. The influence of singlepoint crossover and multipoint crossover operators on the waveguide synthesis was investigated on the H-shaped waveguide simulation (2 degrees of freedom) and on the comb-like waveguide (2, 3 and 4 degrees of freedom). The waveguide models were simulated in Comsol Multiphysics. The optimization algorithm was implemented in MATLAB.

Keywords

Optimization, evolutionary algorithm, multiple operators, mutation, crossover, waveguide.

RIV year

2008

Released

24.07.2008

Publisher

Springer-Verlag Berlin Heidelberg

Location

Heidelberg, Germany

ISBN

978-3-540-85856-0

Book

Evolvable Systems: From Biology to Hardware

Edition

LNCS

Edition number

5216

Pages from

420

Pages to

424

Pages count

4

BibTex


@inproceedings{BUT26942,
  author="Jana {Olivová} and Zbyněk {Raida}",
  title="Waveguide Synthesis by Genetic Algoritms with Multiple Crossover",
  annote="In this paper the waveguide design using of the genetic algorithms as global optimization technique is described. The influence of the multipoint crossover and mutation operators on the convergence is experimentally examined and discussed. The cross section shape of waveguides is optimized with respect to the bandwidth of single mode to be maximized. The influence of singlepoint crossover and multipoint crossover operators on the waveguide synthesis was investigated on the H-shaped waveguide simulation (2 degrees of freedom) and on the comb-like waveguide (2, 3 and 4 degrees of freedom). The waveguide models were simulated in Comsol Multiphysics. The optimization algorithm was implemented in MATLAB.",
  address="Springer-Verlag Berlin Heidelberg",
  booktitle="Evolvable Systems: From Biology to Hardware",
  chapter="26942",
  edition="LNCS",
  howpublished="electronic, physical medium",
  institution="Springer-Verlag Berlin Heidelberg",
  year="2008",
  month="july",
  pages="420--424",
  publisher="Springer-Verlag Berlin Heidelberg",
  type="conference paper"
}