Open AccessOptimization and synthesis of multilayer frequency selective surfaces via bioinspired hybrid techniques
Author(s) -
Wirlan Gomes Lima,
Jasmine Araújo,
Fabrício J. B. Barros,
Gervásio P. S. Cavalcante,
Cássio da Cruz Nogueira,
Bruno Castro,
Miércio Cardoso de Alcântara Neto
Publication year - 2020
Language(s) - English
Resource type - Journals
ISSN - 2411-2933
DOI - 10.31686/ijier.vol8.iss5.2371
Subject(s) - hfss , cuckoo search , finite element method , artificial neural network , microwave , cutoff frequency , matlab , computation , materials science , acoustics , selective surface , computer science , algorithm , physics , engineering , structural engineering , optoelectronics , artificial intelligence , particle swarm optimization , telecommunications , microstrip antenna , antenna (radio) , operating system
In this study, two bioinspired computation (BIC) techniques are discussed and applied to the project and synthesis of multilayer frequency selective surfaces (FSS) within the microwave band, specifically for C, X and Ku bands. The proposed BIC techniques consist of combining an artificial, general regression neural network to a genetic algorithm (GA) and a cuckoo search algorithm, respectively. The objective is to find the optimal values of separation between the investigated FSS. Numerical analysis of the electromagnetic properties of the device is made possible with the finite integration method (FIT) and validated through the finite element method (FEM), utilizing both softwares CST Microwave Studio and Ansys HFSS respectively. Thus, the BIC-optimized devices present good phase / angular stability for angles 10°, 20°, 30° and 40°, as well as being polarization independent. The cutoff frequencies to control the operating frequency range of the FSS, referring to transmission coefficient in decibels (dB), were obtained at a threshold of –10dB. Numerical results denote good accordance with measured data.