Premium
Cooperative particle swarm optimization of passive microwave devices
Author(s) -
Mahanfar Alireza,
Bila Stéphane,
Aubourg Michel,
Verdeyme Serge
Publication year - 2007
Publication title -
international journal of numerical modelling: electronic networks, devices and fields
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.249
H-Index - 30
eISSN - 1099-1204
pISSN - 0894-3370
DOI - 10.1002/jnm.655
Subject(s) - particle swarm optimization , convergence (economics) , mathematical optimization , computer science , multi swarm optimization , filter (signal processing) , population , grid , algorithm , mathematics , demography , geometry , sociology , economics , computer vision , economic growth
Particle swarm optimization (PSO) has lately become very popular in the electromagnetics domain. Although in some instances PSO shows a superior performance compared with other global optimization techniques, it is still computationally more expensive relative to classical gradient techniques. In this paper, a cooperative particle swarm optimization (CPSO) is adopted to achieve a faster convergence compared with the conventional PSO, while maintaining its main feature, which is the capability of finding global optimum. In order to deploy PSO more efficiently, the often neglected effect of the initial population on the overall convergence of PSO is discussed. It is shown that subdividing the space into grid cells and using random distribution within these cells will give the best results in terms of convergence speed. Different boundary conditions are tried on the CPSO algorithm. In order to verify the performance of the proposed algorithm, the algorithm is compared with the conventional PSO using six different objective functions. As a design example, an ultra‐wide‐band filter is designed. The results show a slightly faster convergence compared with the conventional PSO. The designed filter is fabricated and experimental results are also shown. Copyright © 2007 John Wiley & Sons, Ltd.