Table Of ContentIntroduction to Genetic Algorithms
S.N.Sivanandam S.N.Deepa
·
Introduction to Genetic
Algorithms
With 193 Figures and 13 Tables
Authors
S.N.Sivanandam S.N.Deepa
ProfessorandHead Ph.DScholarDept.ofComputerScience
Dept.ofComputerScienceandEngineering andEngineering
PSGCollegeofTechnology PSGCollegeofTechnology
Coimbatore-641004 Coimbatore-641004
TN,India TN,India
LibraryofCongressControlNumber:2007930221
ISBN978-3-540-73189-4SpringerBerlinHeidelbergNewYork
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,broadcasting,
reproductiononmicrofilmorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,
1965,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violations
areliableforprosecutionundertheGermanCopyrightLaw.
SpringerisapartofSpringerScience+BusinessMedia
springer.com
c Springer-VerlagBerlinHeidelberg2008
⃝
Theuseofgeneral descriptive names,registered names,trademarks, etc. inthis publication does not
imply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotective
lawsandregulationsandthereforefreeforgeneraluse.
Typesetting:IntegraSoftwareServicesPvt.Ltd.,India
Coverdesign:ErichKirchner,Heidelberg
Printedonacid-freepaper SPIN:12053230 89/3180/Integra 5 4 3 2 1 0
Preface
Theoriginofevolutionaryalgorithmswasanattempttomimicsomeoftheprocesses
taking place in natural evolution. Although the details of biological evolution are
notcompletelyunderstood(evennowadays),thereexistsomepointssupportedby
strongexperimentalevidence:
Evolutionisaprocessoperatingoverchromosomesratherthanoverorganisms.
•
Theformerareorganictoolsencodingthestructureofalivingbeing,i.e.,acrea-
tureis“built”decodingasetofchromosomes.
Naturalselectionisthemechanismthatrelateschromosomeswiththeefficiency
•
oftheentitytheyrepresent,thusallowingthatefficientorganismwhichiswell-
adaptedtotheenvironmenttoreproducemoreoftenthanthosewhicharenot.
Theevolutionaryprocesstakesplaceduringthereproductionstage.Thereexists
•
a large number of reproductivemechanisms in Nature. Most common onesare
mutation (that causes the chromosomes of offspring to be different to those of
the parents)and recombination(thatcombinesthe chromosomesof the parents
toproducetheoffspring).
Based upon the features above,the three mentioned models of evolutionarycom-
putingwereindependently(andalmostsimultaneously)developed.
An EvolutionaryAlgorithm(EA) is an iterative and stochastic processthatop-
erates on a set of individuals (population). Each individual represents a potential
solution to the problembeing solved.This solution is obtainedby meansof a en-
coding/decodingmechanism. Initially, the population is randomly generated (per-
haps with the help of a constructionheuristic). Everyindividualin the population
isassigned,bymeansofafitnessfunction,ameasureofitsgoodnesswithrespect
to the problem under consideration.This value is the quantitative informationthe
algorithmusestoguidethesearch.
Amongthe evolutionarytechniques,the geneticalgorithms(GAs) are themost
extendedgroupofmethodsrepresentingtheapplicationofevolutionarytools.They
rely on the use of a selection, crossover and mutation operators. Replacement is
usuallybygenerationsofnewindividuals.
IntuitivelyaGAproceedsbycreatingsuccessivegenerationsofbetterandbetter
individuals by applying very simple operations. The search is only guided by the
fitness value associated to every individual in the population. This value is used
to rank individuals depending on their relative suitability for the problem being
v
vi Preface
solved. The problem is the fitness function that for every individual is encharged
ofassigningthefitnessvalue.
The location of this kind of techniques with respect to other deterministic and
non-deterministicproceduresisshowninthefollowingtree.Thisfigurebelowout-
linesthesituationofnaturaltechniquesamongotherwell-knownsearchprocedures.
Combinations of EAs with Hill-Climbing algorithms are very powerful. Ge-
neticalgorithmsintensivelyusingsuchlocalsearchmechanismaretermedMemetic
Algorithms. Also parallelmodelsincrease the extensionandquality of the search.
TheEAsexplorationcomparesquitewellagainsttherestofsearchtechniquesfor
a similar search effort. Exploitation is a more difficult goal in EAs but nowadays
manysolutionsexistforEAstorefinesolutions.
Genetic algorithms are currently the most prominent and widely used compu-
tational models of evolution in artificial-life systems. These decentralized models
provideabasisforunderstandingmanyothersystemsandphenomenaintheworld.
ResearchesonGAsinalifegiveillustrativeexamplesinwhichthegeneticalgorithm
isusedtostudyhowlearningandevolutioninteract,andtomodelecosystems,im-
munesystem,cognitivesystems,andsocialsystems.
About the Book
This book is meant for a wide range of readers, who wishes to learn the basic
conceptsofGeneticAlgorithms.Itcanalsobemeantforprogrammers,researchers
andmanagementexpertswhoseworkisbasedonoptimizationtechniques.Theba-
sicconceptsofGeneticAlgorithmsaredealtindetailwiththerelevantinformation
and knowledge available for understanding the optimization process. The various
operators involved for Genetic Algorithm operation are explained with examples.
Theadvancedoperatorsandthevariousclassificationshavebeendiscussedinlucid
manner,sothatastartercanunderstandtheconceptswithaminimaleffort.
ThesolutionstospecificproblemsaresolvedusingMATLAB7.0andthesolu-
tionsaregiven.TheMATLABGAtoolboxhasalsobeenincludedforeasyreference
ofthe readersso thattheycan havehandsonworkingwith variousGA functions.
ApartfromMATLABsolutions,certainproblemsarealsosolvedusingCandC
++
andthesolutionsaregiven.
Thebookisdesignedtogiveabroadin-depthknowledgeonGeneticAlgorithm.
This book can be used as a handbookand a guide for students of all engineering
disciplines, management sector, operational research area, computer applications,
andforvariousprofessionalswhoworkinOptimizationarea.
GeneticAlgorithms,atpresent,is a hottopic amongacademicians,researchers
andprogramdevelopers.Duetowhich,thisbookisnotonlyforstudents,butalso
forawiderangeofresearchersanddeveloperswhoworkinthisfield.Thisbookcan
be used as a ready referenceguidefor Genetic Algorithmresearch scholars.Most
oftheoperators,classificationsandapplicationsforawidevarietyofareascovered
herefulfillsasanadvancedacademictextbook.
To conclude,we hope that the reader will find this book a helpfulguide and a
valuablesourceofinformationaboutGeneticAlgorithmconceptsfortheirseveral
practicalapplications.
1 Organizationofthe Book
The book contains11 chaptersaltogether. It starts with the introductionto Evolu-
tionaryComputing.Thevariousapplicationcasestudiesarealsodiscussed.
Thechaptersareorganizedasfollows:
vii
viii AbouttheBook
Chapter1givesanintroductiontoEvolutionarycomputing,itsdevelopmentand
•
itsfeatures.
Chapter 2 enhancesthe growthof Genetic Algorithmsand its comparisonwith
•
otherconventionaloptimizationtechniques.Alsothebasicsimplegeneticalgo-
rithmwithitsadvantagesandlimitationsarediscussed.
Thevariousterminologiesandthebasicoperatorsinvolvedingeneticalgorithm
•
aredealtinChap.3.Fewexampleproblems,enablingthereaderstounderstand
thebasicgeneticalgorithmoperationarealsoincluded.
Chapter 4 discusses the advanced operatorsand techniquesinvolvedin genetic
•
algorithm.
Thedifferentclassificationsofgeneticalgorithmareprovidedin Chap.5.Each
•
oftheclassificationsis discussedwiththeiroperatorsandmodeofoperationto
achieveoptimizedsolution.
Chapter6givesabriefintroductiontogeneticprogramming.Thestepsinvolved
•
and characteristics of genetic programming with its applications are described
here.
Chapter 7 discusses on variousgenetic algorithm optimizationproblemswhich
•
includes fuzzy optimization, multi objective optimization, combinatorial opti-
mization,schedulingproblemsandsoon.
TheimplementationofgeneticalgorithmusingMATLABisdiscussedinChap.8.
•
Thetoolboxfunctionsandsimulatedresultstospecificproblemsareprovidedin
thischapter.
Chapter 9 gives the implementation of genetic algorithm concept using C and
•
C .Theimplementationisperformedforfewbenchmarkproblems.
++
Theapplicationofgeneticalgorithminvariousemergingfieldsalongwithcase
•
studiesisgiveninChapter10.
Chapter 11 gives a brief introduction to particle swarm optimization and ant
•
colonyoptimization.
TheBibliographyisgivenattheendforthereadyreferenceofreaders.
2 SalientFeatures ofthe Book
Thesalientfeaturesofthebookinclude:
DetailedexplanationofGeneticAlgorithmconcepts
•
NumerousGeneticAlgorithmOptimizationProblems
•
StudyonvarioustypesofGeneticAlgorithms
•
ImplementationofOptimizationproblemusingCandC
• ++
SimulatedsolutionsforGeneticAlgorithmproblemsusingMATLAB7.0
•
BriefdescriptiononthebasicsofGeneticProgramming
•
ApplicationcasestudiesonGeneticAlgorithmonemergingfields
•
S.N. Sivanandam completed his B.E (Electrical and Electronics Engineering) in
1964fromGovernmentCollegeofTechnology,CoimbatoreandM.Sc(Engineering)
AbouttheBook ix
in Power System in 1966 from PSG College of Technology, Coimbatore. He
acquiredPhDinControlSystemsin1982fromMadrasUniversity.Hehasreceived
Best Teacher Award in the year 2001 and Dhakshina Murthy Award for Teaching
ExcellencefromPSGCollegeofTechnology.HereceivedTheCITATIONforbest
teachingandtechnicalcontributionintheYear2002,GovernmentCollegeofTech-
nology,Coimbatore.Hehasa totalteachingexperience(UGandPG) of41years.
Thetotalnumberofundergraduateandpostgraduateprojectsguidedbyhimforboth
Computer Science and Engineering and Electrical and Electronics Engineering is
around 600. He is currently working as a Professor and Head Computer Science
andEngineeringDepartment,PSGCollegeofTechnology,Coimbatore[fromJune
2000]. He has been identified as an outstanding person in the field of Computer
Science and Engineering in MARQUIS “Who’s Who”, October 2003 issue, New
providence,NewJersey,USA.Hehasalsobeenidentifiedasanoutstandingperson
inthefieldofComputationalScienceandEngineeringin“Who’sWho”,December
2005 issue, Saxe-CoburgPublications, United Kingdom.He has been placed as a
VIPmemberinthecontinentalWHO’sWHORegistryofnationalBusinessLeaders,
Inc.33WestHawthorneAvenueValleyStream,NY11580,Aug24,2006.
S.N.Sivanandamhaspublished12 books.He hasdeliveredaround150special
lectures of different specialization in Summer/Winter school and also in various
Engineeringcolleges.He hasguidedandcoguided30Ph.D researchworksandat
present9Ph.Dresearchscholarsareworkingunderhim.Thetotalnumberoftech-
nicalpublicationsinInternational/Nationaljournals/Conferencesisaround700.He
hasalsoreceivedCertificateofMerit2005–2006forhispaperfromTheInstitution
of Engineers (India). He has chaired 7 International conferences and 30 National
conferences.He is a memberof variousprofessionalbodieslike IE (India),ISTE,
CSI,ACSandSSI.HeisatechnicaladvisorforvariousreputedindustriesandEn-
gineeringInstitutions.HisresearchareasincludeModelingandSimulation,Neural
networks , Fuzzy Systems and Genetic Algorithm, Pattern Recognition, Multi di-
mensionalsystemanalysis,LinearandNonlinearcontrolsystem,SignalandImage
processing,ControlSystem, Power system, Numericalmethods,Parallel Comput-
ing,DataMiningandDatabaseSecurity.
S.N.DeepahascompletedherB.EDegreefromGovernmentCollegeofTechnol-
ogy,Coimbatore,1999and M.E DegreefromPSG College of Technology,Coim-
batore, 2004. She was a gold medalist in her B.E Degree Programme. She has
receivedG.DMemorialAwardintheyear1997andBestOutgoingStudentAward
fromPSGCollegeofTechnology,2004.HerM.EThesiswonNationalAwardfrom
the Indian Society of Technical Education and L&T, 2004. She has published 5
booksandpapersinInternationalandNationalJournals.Herresearchareasinclude
Neural Network, Fuzzy Logic, Genetic Algorithm, Digital Control, Adaptive and
Non-linearControl.
Acknowledgement
TheauthorsarealwaysthankfultotheAlmightyforperseveranceandachievements.
They wish to thank Shri G. Rangaswamy, Managing Trustee, PSG Institutions,
ShriC.R.Swaminathan,ChiefExecutive;andDr.R.Rudramoorthy,Principal,PSG
CollegeofTechnology,Coimbatore,fortheirwhole-heartedcooperationandgreat
encouragementgiveninthissuccessfulendeavor.Theyalsowishtothankthestaff
membersofcomputerscienceandengineeringfortheircooperation.Deepawishes
tothankherhusbandAnand,daughterNivethithaandparentsfortheirsupport.
xi