nach oben

2010 | Buch

Kapitel lesen Erstes Kapitel lesen

Differential Evolution in Electromagnetics

herausgegeben von: Anyong Qing, Ching Kwang Lee

Verlag: Springer Berlin Heidelberg

Buchreihe : Adaptation, Learning, and Optimization

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

Differential evolution has proven itself a very simple while very powerful stochastic global optimizer. It has been applied to solve problems in many scientific and engineering fields. This book focuses on applications of differential evolution in electromagnetics to showcase its achievement and capability in solving synthesis and design problems in electromagnetics.Topics covered in this book include:• A comprehensive up-to-date literature survey on differential evolution• A systematic description of differential evolution• A topical review on applications of differential evolution in electromagnetics• Five new application examplesThis book is ideal for electromagnetic researchers and people in differential evolution community. It is also a valuable reference book for researchers and students in the optimization or electrical and electronic engineering field. In addition, managers and engineers in relevant fields will find it a helpful introductory guide.

Inhaltsverzeichnis

Frontmatter

A Literature Survey on Differential Evolution

Motivations

Eliminating Inconsistencies

It has been observed since 2004 that there are many inconsistent or even false claims prevailing in the community of differential evolution [1]. Two measures have been taken to clarify them. The first is a system level parametric study on differential evolution [1]-[4]. The second is the large scale literature survey mentioned here. It is one of the foundation stones of this book.

Anyong Qing

Basics of Differential Evolution

A Short History

Inception

Differential evolution was proposed by K.V. Price and R. Storn in 1995 [1]. At that time, Price was asked to solve the Chebyshev polynomial fitting problem [1]-[5] by Storn [2], [5]. Initially, he tried to solve it by using genetic annealing algorithm [6]. However, although he eventually found the solution to the 5-dimensional Chebyshev polynomial fitting problem by using genetic annealing algorithm, he was frustrated to notice that genetic annealing algorithm fails to fulfill the three requirements for a practical optimization technique: strong global search capability, fast convergence, and user friendliness.

A breakthrough happened when Price came up with an innovative scheme for generating trial parameter vectors. In this scheme, a new parameter vector is generated by adding the weighted difference vector between two population members to a third member. Such a scheme was named as differential mutation and has been well known to be the crucial idea behind the success of differential evolution. The cornerstone for differential evolution was therefore laid.

Price wrapped up his invention with other critical ideas: natural real code, arithmetic operations, mother-child competition and selection, and execution of evolutionary operations in the order of mutation-crossover-selection. Consequently, differential evolution, a very reliable, efficient, robust, and simple evolutionary algorithm was developed.

Anyong Qing

A Retrospective of Differential Evolution in Electromagnetics

Introduction

Coverage

The electromagnetic spectrum extends from below frequencies used for modern radio to gamma radiation at the short-wavelength end, covering wavelengths from thousands of kilometers down to a fraction of the size of an atom. The long wavelength limit is the size of the universe itself, while it is thought that the short wavelength limit is in the vicinity of the Planck length, although in principle the spectrum is infinite and continuous. Radio waves, microwaves, terahertz waves, infrared light, visible light, ultraviolet light, X-rays, and gamma rays are all kinds of electromagnetic waves. However, in this chapter, attention is focused on radio waves and microwaves since other electromagnetic waves are entertained by more specific subjects, for example, optics for light waves.

In addition, electromagnetics is closely related with many other disciplines. Many inter-disciplinary fields have been increasingly created through mutual invasion between such disciplines and electromagnetics. However, in this chapter, an application of differential evolution will not be classified into the electromagnetics category unless it is applied to solve an electromagnetic problem.

This chapter is based purely on the literature survey mentioned in Chapter 1 of this book. Please note that some of the collected publications are not cited here due to concern of language translation accuracy for non-English publications and/or classification accuracy for those publications whose full text is unavailable to this author at this moment.

Similarly, to avoid any potential misleading to readers, partial result for year 2009 is not presented here.

Anyong Qing

Application of Differential Evolution to a Two-Dimensional Inverse Scattering Problem

Introduction

Inverse scattering problems [1]-[2] are of great importance in non-destructive and non-invasive evaluation applications. Typically, the region of investigation is inaccessible and has to be evaluated using different approaches including electromagnetic waves. In such scenarios, the region is illuminated by electromagnetic waves from various directions and the electromagnetic fields scattered by objects in the region are measured at various receivers. The electrical and geometric properties of objects present inside the region are then reconstructed using the measured scattered electromagnetic fields.

Krishna Agarwal, Xudong Chen, Yu Zhong

The Use of Differential Evolution for the Solution of Electromagnetic Inverse Scattering Problems

Abstract

Inspection of penetrable objects by using differential evolution together with a recently proposed iterative multiscaling approach is discussed in this Chapter. Several new results are included concerning the reconstruction of inhomogeneous targets under various imaging conditions.

A. Donelli, A. Massa, G. Oliveri, M. Pastorino, A. Randazzo

Modeling of Electrically Large Equipment with Distributed Dipoles Using Metaheuristic Methods

Introduction

Near-Field to Far-Field Transformation

The reference environment for measurement of radiated emissions is the open area test site (OATS). The far field of a radiating element is of great interest [1], [2] since it is independent of distance. However, in an OATS environment, it is hardly possible to separate interferences generated by the equipment under test (EUT) from external ones. Semi-anechoic chambers have to be used instead. For large radiating elements, the minimal distance [1], [3]-[5] needed to measure the far field is large, especially at low frequencies. Therefore, large costly semi-anechoic chambers which are only affordable to very few big laboratories are required.

Joan Ramon Regué, Miquel Ribó, José Gomila

Application of Differential Evolution to a Multi-Objective Real-World Frequency Assignment Problem

Introduction

Frequency spectrum is one of the scarcest resources for any mobile operator. Frequencies have to be reused throughout the network. Consequently, interferences may occur and some separation constraints may be violated. Frequency assignment problem (FAP) aims to use effectively the available frequency spectrum to minimize interferences by carefully allocating available frequencies to existing base stations [1]. It is a very demanding problem in telecommunications, especially in GSM networks [2], even though it is very time-consuming. It is one of the most fundamental problems in mobile communications planning. A good FAP solution leads to better network quality and increased capacity without sacrificing quality of service (QoS) for all users of the mobile network.

Marisa Silva Maximiano, Miguel A. Vega-Rodríguez, Juan A. Gómez-Pulido, Juan M. Sánchez-Pérez

RNN Based MIMO Channel Prediction

Abstract

In this work, differential evolution (DE) is combined with particle swarm optimization (PSO) and another evolutionary algorithm (EA) to create a novel hybrid PSO-EA-DEPSO algorithm. The alteration between PSO, PSO-EA, and DEPSO provides additional diversity to counteract premature convergence. This hybrid algorithm is then shown to outperform PSO, PSO-EA, and DEPSO when applied to wireless MIMO channel prediction.

Chris Potter

Backmatter

Titel: Differential Evolution in Electromagnetics
herausgegeben von: Anyong Qing
Ching Kwang Lee
Verlag: Springer Berlin Heidelberg
Electronic ISBN: 978-3-642-12869-1
Print ISBN: 978-3-642-12868-4
DOI: https://doi.org/10.1007/978-3-642-12869-1

Premium Partner

Marktübersichten

Die im Laufe eines Jahres in der „adhäsion“ veröffentlichten Marktübersichten helfen Anwendern verschiedenster Branchen, sich einen gezielten Überblick über Lieferantenangebote zu verschaffen.

Zur Marktübersicht