Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.

Evolutionary algorithms, such as genetic algorithms (GAs) [1], evolutionary programming (EP) [2], and evolutionary strategies (ES), have recently received much attention for global optimization of electromagnetic problems [3-5,7]. These evolutionary algorithms are heuristic population-based search p...

Full description

Bibliographic Details
Main Authors: Chellapilla, Kumar., Hoorfar, Ahmad.
Format: Villanova Faculty Authorship
Language:English
Published: 1998
Online Access:http://ezproxy.villanova.edu/login?url=https://digital.library.villanova.edu/Item/vudl:176980
id vudl:176980
record_format vudl
institution Villanova University
collection Digital Library
modeltype_str_mv vudl-system:ResourceCollection
vudl-system:CollectionModel
vudl-system:CoreModel
datastream_str_mv PARENT-LIST-RAW
DC
MEMBER-QUERY
MEMBER-LIST-RAW
LICENSE
THUMBNAIL
STRUCTMAP
LEGACY-METS
PROCESS-MD
PARENT-QUERY
PARENT-LIST
RELS-EXT
AUDIT
AGENTS
hierarchytype
hierarchy_all_parents_str_mv vudl:176907
vudl:172968
vudl:641262
vudl:3
vudl:1
sequence_vudl_176907_str 0000000025
hierarchy_top_id vudl:641262
hierarchy_top_title Villanova Faculty Publications
fedora_parent_id_str_mv vudl:176907
hierarchy_first_parent_id_str vudl:176980
hierarchy_parent_id vudl:176907
hierarchy_parent_title Hoorfar Ahmad
hierarchy_sequence_sort_str 0000000025
hierarchy_sequence 0000000025
spelling Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
Chellapilla, Kumar.
Hoorfar, Ahmad.
Evolutionary algorithms, such as genetic algorithms (GAs) [1], evolutionary programming (EP) [2], and evolutionary strategies (ES), have recently received much attention for global optimization of electromagnetic problems [3-5,7]. These evolutionary algorithms are heuristic population-based search procedures that incorporate random variation and selection. Of the three paradigms GAs are well known to the electromagnetic community. Even though several successful applications have been reported, recent research has identified some inefficiencies in GA performance [6]. This degradation in efficiency is apparent in applications with highly epistatic objective functions, i.e., where the parameters being optimized are highly correlated. On the other hand, EP and ES are more robust to epistatic objective functions and coordinate rotations. EP has been shown to be more efficient than GA on many function optimization problems [2]. The dynamics of GA are explained through the building block hypothesis and the schema theorem [1], which are not fully accepted in the evolutionary computation literature. On the other hand, the convergence theory for EP is well established and EP has been proven to asymptotically converge to the global optimum with probability one, under elitist selection [2]. Further, as will be demonstrated, EP is well suited for optimizing continuous, discrete, and mixed parameter optimization problems. Binary GAs require the parameters to be coded as bits. The selection of the crossover and mutation probabilities is quite arbitrary and they are not adapted during evolution. The selection of the initial values for the strategy parameters for EP and ES are well defined and efficient adaptive and self-adaptive techniques exist for adapting these parameters during evolution. In this work, the capabilities of EP will be demonstrated and contrasted with those obtained using GAs on three challenging electromagnetic optimization problems, namely, the design of optimally thinned linear arrays, aperiodic arrays, and Yagi-Uda antennas. A more complicated problem on the gain optimization of a multilayered microstrip Yagi array will be discussed in a separate paper in this symposium [8].
1998
Villanova Faculty Authorship
vudl:176980
IEEE Transactions on Antennas and Propagation, 1998.
en
dc.title_txt_mv Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
dc.creator_txt_mv Chellapilla, Kumar.
Hoorfar, Ahmad.
dc.description_txt_mv Evolutionary algorithms, such as genetic algorithms (GAs) [1], evolutionary programming (EP) [2], and evolutionary strategies (ES), have recently received much attention for global optimization of electromagnetic problems [3-5,7]. These evolutionary algorithms are heuristic population-based search procedures that incorporate random variation and selection. Of the three paradigms GAs are well known to the electromagnetic community. Even though several successful applications have been reported, recent research has identified some inefficiencies in GA performance [6]. This degradation in efficiency is apparent in applications with highly epistatic objective functions, i.e., where the parameters being optimized are highly correlated. On the other hand, EP and ES are more robust to epistatic objective functions and coordinate rotations. EP has been shown to be more efficient than GA on many function optimization problems [2]. The dynamics of GA are explained through the building block hypothesis and the schema theorem [1], which are not fully accepted in the evolutionary computation literature. On the other hand, the convergence theory for EP is well established and EP has been proven to asymptotically converge to the global optimum with probability one, under elitist selection [2]. Further, as will be demonstrated, EP is well suited for optimizing continuous, discrete, and mixed parameter optimization problems. Binary GAs require the parameters to be coded as bits. The selection of the crossover and mutation probabilities is quite arbitrary and they are not adapted during evolution. The selection of the initial values for the strategy parameters for EP and ES are well defined and efficient adaptive and self-adaptive techniques exist for adapting these parameters during evolution. In this work, the capabilities of EP will be demonstrated and contrasted with those obtained using GAs on three challenging electromagnetic optimization problems, namely, the design of optimally thinned linear arrays, aperiodic arrays, and Yagi-Uda antennas. A more complicated problem on the gain optimization of a multilayered microstrip Yagi array will be discussed in a separate paper in this symposium [8].
dc.date_txt_mv 1998
dc.format_txt_mv Villanova Faculty Authorship
dc.identifier_txt_mv vudl:176980
dc.source_txt_mv IEEE Transactions on Antennas and Propagation, 1998.
dc.language_txt_mv en
author Chellapilla, Kumar.
Hoorfar, Ahmad.
spellingShingle Chellapilla, Kumar.
Hoorfar, Ahmad.
Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
author_facet Chellapilla, Kumar.
Hoorfar, Ahmad.
dc_source_str_mv IEEE Transactions on Antennas and Propagation, 1998.
format Villanova Faculty Authorship
author_sort Chellapilla, Kumar.
dc_date_str 1998
dc_title_str Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
description Evolutionary algorithms, such as genetic algorithms (GAs) [1], evolutionary programming (EP) [2], and evolutionary strategies (ES), have recently received much attention for global optimization of electromagnetic problems [3-5,7]. These evolutionary algorithms are heuristic population-based search procedures that incorporate random variation and selection. Of the three paradigms GAs are well known to the electromagnetic community. Even though several successful applications have been reported, recent research has identified some inefficiencies in GA performance [6]. This degradation in efficiency is apparent in applications with highly epistatic objective functions, i.e., where the parameters being optimized are highly correlated. On the other hand, EP and ES are more robust to epistatic objective functions and coordinate rotations. EP has been shown to be more efficient than GA on many function optimization problems [2]. The dynamics of GA are explained through the building block hypothesis and the schema theorem [1], which are not fully accepted in the evolutionary computation literature. On the other hand, the convergence theory for EP is well established and EP has been proven to asymptotically converge to the global optimum with probability one, under elitist selection [2]. Further, as will be demonstrated, EP is well suited for optimizing continuous, discrete, and mixed parameter optimization problems. Binary GAs require the parameters to be coded as bits. The selection of the crossover and mutation probabilities is quite arbitrary and they are not adapted during evolution. The selection of the initial values for the strategy parameters for EP and ES are well defined and efficient adaptive and self-adaptive techniques exist for adapting these parameters during evolution. In this work, the capabilities of EP will be demonstrated and contrasted with those obtained using GAs on three challenging electromagnetic optimization problems, namely, the design of optimally thinned linear arrays, aperiodic arrays, and Yagi-Uda antennas. A more complicated problem on the gain optimization of a multilayered microstrip Yagi array will be discussed in a separate paper in this symposium [8].
title Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
title_full Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
title_fullStr Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
title_full_unstemmed Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
title_short Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
title_sort evolutionary programming: an efficient alternative to genetic algorithms for electromagnetic optimization problems.
publishDate 1998
normalized_sort_date 1998-01-01T00:00:00Z
language English
collection_title_sort_str evolutionary programming: an efficient alternative to genetic algorithms for electromagnetic optimization problems.
fgs.type_txt_mv http://www.w3.org/ns/ldp#RDFSource
http://fedora.info/definitions/v4/repository#Resource
http://www.w3.org/ns/ldp#BasicContainer
http://www.w3.org/ns/ldp#Resource
http://fedora.info/definitions/v4/repository#Container
http://www.w3.org/ns/ldp#Container
fgs.createdBy_txt_mv fedoraAdmin
fgs.createdDate_txt_mv 2013-01-22T06:05:29.360Z
fgs.state_txt_mv Active
relsext.sortOn_txt_mv title
fgs.lastModifiedDate_txt_mv 2021-04-12T19:02:40.294Z
relsext.hasModel_txt_mv http://hades.library.villanova.edu:8080/rest/vudl-system:ResourceCollection
http://hades.library.villanova.edu:8080/rest/vudl-system:CollectionModel
http://hades.library.villanova.edu:8080/rest/vudl-system:CoreModel
fgs.ownerId_txt_mv diglibEditor
fgs.label_txt_mv Evolutionary Programming: An Efficient Alternative to Genetic Algorithms for Electromagnetic Optimization Problems.
relsext.isMemberOf_txt_mv http://hades.library.villanova.edu:8080/rest/vudl:176907
fgs.lastModifiedBy_txt_mv fedoraAdmin
relsext.itemID_txt_mv oai:digital.library.villanova.edu:vudl:176980
relsext.hasLegacyURL_txt_mv http://digital.library.villanova.edu/Villanova%20Digital%20Collection/Faculty%20Fulltext/Hoorfar%20Ahmad/HoorfarAhmad-b425dcb6-a714-4026-81ec-147f262044ff.xml
relsext.sequence_txt_mv vudl:176907#25
has_order_str no
agent.name_txt_mv Falvey Memorial Library, Villanova University
KHL
license.mdRef_str http://digital.library.villanova.edu/copyright.html
license_str protected
has_thumbnail_str true
THUMBNAIL_contentDigest_digest_str 203c69e18f4f46c81e9892448d2c07cd
first_indexed 2014-01-11T22:32:46Z
last_indexed 2021-04-12T19:12:48Z
_version_ 1755660285929062400
subpages