Real-Coded Genetic Algorithm with Differential Evolution Operator for Terahertz Quasi-Optical Power Divider/Combiner Design

This paper describes a modified real-coded genetic algorithm with a differential evolution operator (RGA-DE) for a shaped reflector of terahertz (THz) quasi-optical power divider/combiner. The real-coded genetic algorithm (RGA) is more convenient to define the reflector surfaces as compared to other...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied Computational Electromagnetics Society journal 2017-10, Vol.32 (10), p.888
Hauptverfasser: Zhang, Fan, Song, Kaijun, Fan, Yong
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper describes a modified real-coded genetic algorithm with a differential evolution operator (RGA-DE) for a shaped reflector of terahertz (THz) quasi-optical power divider/combiner. The real-coded genetic algorithm (RGA) is more convenient to define the reflector surfaces as compared to other coding methods. Interpolation method is applied to smooth the sophisticated surfaces. A better searching ability is obtained in a small number of individuals by using the differential evolution operator instead of the conventional crossover operator. So a bionics algorithm combining real-coded genetic algorithm and differential evolution (DE) algorithm has been presented to design the THz quasi-optical multiple-way holographic power-combining circuits. The RGA-DE has been utilized to optimize the coupling coefficient kappa of the field distribution to get close to 1. Two real-world examples operating at 380 GHz and 450 GHz respectively are described. The computed results are verified by using the commercial software FEKO Suite.
ISSN:1054-4887
1943-5711