学习日志之synthesis and optimization(8)——heuristic优化工具

目录

 

需要知道的一些背景知识

对espresso-heuristic算法过程的Matrix表述(方便用计算机)

Tautology

Containment(基于tautology)

Complementation(基于tautology)


需要知道的一些背景知识

令一个逻辑函数为f(x_{1},x_{2},...,x_{n})

对于其中一个变量x_{i}的cofactor记为:f_{x_{i}}=f(x_{1},x_{2},...,1,...x_{n}) (在f中将x_{i}取为1)

对于其中一个变量{}x_{i}}'的cofactor记为:f_{​{x_{i}}'}=f(x_{1},x_{2},...,0,...x_{n}) (在f中将x_{i}取为0)

Boole's expansion theorem:f(x_{1},x_{2},...,x_{n})=x_{i}f_{x_{i}}+{x_{i}}'f_{​{x_{i}}'} (即用cofactor来表示函数)

函数的单调性(unateness):

f_{x_{i}}\geqslant f_{​{x_{i}}'}称为positive unate;f_{x_{i}}\leqslant f_{​{x_{i}}'}称为negative unate

如果函数中所有的变量都是positive unate或者是negative unate的,那么就称这个函数是positive unate或者是negative unate

boolean函数的导数:\frac{\partial f}{\partial x_{i}}\equiv f_{x_{i}}\oplus f_{​{x_{i}}'}

f函数关于x_{i}的concensus:C_{x_{i}}\equiv f_{x_{i}}\cdot f_{​{x_{i}}'}

f函数关于x_{i}的smoothing:S_{x_{i}}\equiv f_{x_{i}} +f_{​{x_{i}}'}

如何用Matrix来表示一个cover:

例如:f = a’d’ + a’b + ab’ + ac’d

wAAACH5BAEKAAAALAAAAAABAAEAAAICRAEAOw==

对espresso-heuristic算法过程的Matrix表述(方便用计算机)

1. expansion

将方程用cube表示在minterm中先任选一个进行扩展(目的是将其中一个变量升级为don't care,即合并同类项)。先将这两个minterm用Matrix表示出来,单后对表中的每一个位(表中元素的两个位都要)按位或。得到的结果即为扩展的结果。

例:f=ab+bc+ca

画出cube为,选中起始点为ab'c,需要向abc方向扩展

用Matrix表示两个变量,并且将两个变量按位或,就得到了上图中的\gamma。这相当于做了一个并(Union)操作。

2. 接下来要判断这个expansion的方向是不是正确。因为方向是随机选的,可能会扩展到不是minterm的点上。做这一步的时候首先我们前提是已经知道了这个函数的off-set(真值表里找所有为0的项),然后将这个合并的结果与所有off-set中的元素进行按位与操作(和上面一样)可以德奥一个值,遍历这个off-set如果其中有一个结果中没有出现‘00’,那么就可以判断这个扩展的方向是错误的,因为你选到了不包含在方程中的项。

例:上面的例子中,我事先知道了off-set中包含了一个ab'c'。如果计算机向ab'c'扩展,那么扩展的结果根据上一步方法就是ab'。将其和off-set中的所有minterm进行按位与(交(intersection))操作,进行到ab'c'的时候如下图可知这里没有‘00’,所以我们选择的方向是错误的。

Tautology

这个定义为一个逻辑函数不论其变量取什么值,其函数值总为真。其思想是如果一个函数为tautology那么这个函数的两个cofactor都是tautology,因此要知道这个函数是否为tautology如果判断不了,可以由其cofactor来判断。

判断一个函数是否为tautology的原则(用Matrix表示):

  • Matrix有一整行都是‘11’(don't care)=> Tautology
  • 有一整小列为‘0’ => !Tautology
  • 所有的其variable只有一个,其它的都是'11',且这一列没有全‘0’的小列 => Tautology

1. 写出f的Matrix

2. 如果不能判断,将其表达为关于其某个变量的cofactor的形式,如果还不能用上面的结论判断那就再分解,知道可以判断为止。流程大致如下图所示

例:f=ab+ac+ab'c'+a'

Matrix f:

显然判断不出来,所以要写出其cofactor.我们可以先写其关于a的两个cofactor。首先a对于所有变量的Matrix如下,显然上面三个条件都不满足,所以要写出其cofactor表达式

用f的第一行与这个a的一行做交运算(按位与)

这个里面没有‘00’的项就需要进一步计算,将a的complement与对应在f中的行进行并操作(按位或)

这样就得到了关于a 的cofactor的第一行,如果碰见在做交运算时出现‘00’的状况,就不要加到cofactor里去。在这么算完4行之后就可以得到f_{a}的matrix,

不过显然这个玩意也看不出来,于是按照这个继续往下分。流程如下图所示

Containment(基于tautology)

可用于检查函数中是否包含特定的implicant。主要思想是由f_{\alpha }是tautology的=>Implicant \alpha 包含于函数F中

Complementation(基于tautology)

用于计算函数的补函数。显然如果是tautology的话其补函数就是‘0‘了,需要计算的一般不会是tautology的。跟前面的步骤一样,分解算出各个cofactor是否为tautology,对于不是tautology的项按位取反。

例如:f=ab+ac+a'最后算出来的二叉树如下所示

只剩了一个f_{b'}不是tautology的。将这些最底层的全部取反,tautology的项都变成了0,而f_{b'}变成了c'。即取反结果如下所示

Soft computing and nature-inspired computing both play a significant role in developing a better understanding to machine learning. When studied together, they can offer new perspectives on the learning process of machines. The Handbook of Research on Soft Computing and Nature-Inspired Algorithms is an essential source for the latest scholarly research on applications of nature-inspired computing and soft computational systems. Featuring comprehensive coverage on a range of topics and perspectives such as swarm intelligence, speech recognition, and electromagnetic problem solving, this publication is ideally designed for students, researchers, scholars, professionals, and practitioners seeking current research on the advanced workings of intelligence in computing systems. Chapter 1 ApplicationofNatured-InspiredAlgorithmsfortheSolutionofComplexElectromagnetic Problems................................................................................................................................................. 1 Massimo Donelli, University of Trento, Italy Inthelastdecadenature-inspiredOptimizerssuchasgeneticalgorithms(GAs),particleswarm(PSO), antcolony(ACO),honeybees(HB),bacteriafeeding(BFO),firefly(FF),batalgorithm(BTO),invasive weed(IWO)andothersalgorithms,hasbeensuccessfullyadoptedasapowerfuloptimizationtools inseveralareasofappliedengineering,andinparticularforthesolutionofcomplexelectromagnetic problems.Thischapterisaimedatpresentinganoverviewofnatureinspiredoptimizationalgorithms (NIOs)asappliedtothesolutionofcomplexelectromagneticproblemsstartingfromthewell-known geneticalgorithms(GAs)uptorecentcollaborativealgorithmsbasedonsmartswarmsandinspired byswarmofinsects,birdsorflockoffishes.Thefocusofthischapterisontheuseofdifferentkind ofnaturedinspiredoptimizationalgorithmsforthesolutionofcomplexproblems,inparticulartypical microwavedesignproblems,inparticularthedesignandmicrostripantennastructures,thecalibration ofmicrowavesystemsandotherinterestingpracticalapplications.Startingfromadetailedclassification andanalysisofthemostusednaturedinspiredoptimizers(NIOs)thischapterdescribesthenotonly thestructuresofeachNIObutalsothestochasticoperatorsandthephilosophyresponsibleforthe correctevolutionoftheoptimizationprocess.Theoreticaldiscussionsconcernedconvergenceissues, parameterssensitivityanalysisandcomputationalburdenestimationarereportedaswell.Successively abriefreviewonhowdifferentresearchgroupshaveappliedorcustomizeddifferentNIOsapproaches forthesolutionofcomplexpracticalelectromagneticproblemrangingfromindustrialuptobiomedical applications.ItisworthnoticedthatthedevelopmentofCADtoolsbasedonNIOscouldprovidethe engineersanddesignerswithpowerfultoolsthatcanbethesolutiontoreducethetimetomarketof specific devices, (such as modern mobile phones, tablets and other portable devices) and keep the commercialpredominance:sincetheydonotrequireexpertengineersandtheycanstronglyreducethe computationaltimetypicalofthestandardtrialerrorsmethodologies.Suchusefulautomaticdesigntools basedonNIOshavebeentheobjectofresearchsincesomedecadesandtheimportanceofthissubject iswidelyrecognized.Inordertoapplyanaturedinspiredalgorithm,theproblemisusuallyrecastas aglobaloptimizationproblem.Formulatedinsuchaway,theproblemcanbeefficientlyhandledby naturedinspiredoptimizerbydefiningasuitablecostfunction(singleormulti-objective)thatrepresent thedistancebetweentherequirementsandtheobtainedtrialsolution.Thedeviceunderdevelopment  canbeanalyzedwithclassicalnumericalmethodologiessuchasFEM,FDTD,andMoM.Asacommon feature,theseenvironmentsusuallyintegrateanoptimizerandacommercialnumericalsimulator.The chapterendswithopenproblemsanddiscussiononfutureapplications. Chapter 2 AComprehensiveLiteratureReviewonNature-InspiredSoftComputingandAlgorithms:Tabular andGraphicalAnalyses........................................................................................................................ 34 Bilal Ervural, Istanbul Technical University, Turkey Beyzanur Cayir Ervural, Istanbul Technical University, Turkey Cengiz Kahraman, Istanbul Technical University, Turkey SoftComputingtechniquesarecapableofidentifyinguncertaintyindata,determiningimprecisionof knowledge,andanalyzingill-definedcomplexproblems.Thenatureofrealworldproblemsisgenerally complexandtheircommoncharacteristicisuncertaintyowingtothemultidimensionalstructure.Analytical modelsareinsufficientinmanagingallcomplexitytosatisfythedecisionmakers’expectations.Under thisviewpoint,softcomputingprovidessignificantflexibilityandsolutionadvantages.Inthischapter, firstly,themajorsoftcomputingmethodsareclassifiedandsummarized.Thenacomprehensivereviewof eightnatureinspired–softcomputingalgorithmswhicharegeneticalgorithm,particleswarmalgorithm, antcolonyalgorithms,artificialbeecolony,fireflyoptimization,batalgorithm,cuckooalgorithm,and greywolfoptimizeralgorithmarepresentedandanalyzedundersomedeterminedsubjectheadings (classificationtopics)inadetailedway.Thesurveyfindingsaresupportedwithcharts,bargraphsand tablestobemoreunderstandable. Chapter 3 SwarmIntelligenceforElectromagneticProblemSolving................................................................... 69 Luciano Mescia, Politecnico di Bari, Italy Pietro Bia, EmTeSys Srl, Italy Diego Caratelli, The Antenna Company, The Netherlands & Tomsk Polytechnic University, Russia Johan Gielis, University of Antwerp, Belgium ThechapterwilldescribethepotentialoftheswarmintelligenceandinparticularquantumPSO-based algorithm,tosolvecomplicatedelectromagneticproblems.Thistaskisaccomplishedthroughaddressing the design and analysis challenges of some key real-world problems. A detailed definition of the conventionalPSOanditsquantum-inspiredversionarepresentedandcomparedintermsofaccuracyand computationalburden.Sometheoreticaldiscussionsconcerningtheconvergenceissuesandasensitivity analysisontheparametersinfluencingthestochasticprocessarereported. Chapter 4 ParameterSettingsinParticleSwarmOptimization........................................................................... 101 Snehal Mohan Kamalapur, K. K. Wagh Institute of Engineering Education and Research, India Varsha Patil, Matoshree College of Engineering and Research Center, India Theissueofparametersettingofanalgorithmisoneofthemostpromisingareasofresearch.Particle SwarmOptimization(PSO)ispopulationbasedmethod.TheperformanceofPSOissensitivetothe parametersettings.Intheliteratureofevolutionarycomputationtherearetwotypesofparametersettings  - parametertuningandparametercontrol.Staticparametertuningmayleadtopoorperformanceas optimalvaluesofparametersmaybedifferentatdifferentstagesofrun.Thisleadstoparametercontrol. Thischapterhastwo-foldobjectivestoprovideacomprehensivediscussiononparametersettingsandon parametersettingsofPSO.Theobjectivesaretostudyparametertuningandcontrol,togettheinsight ofPSOandimpactofparameterssettingsforparticlesofPSO. Chapter 5 ASurveyofComputationalIntelligenceAlgorithmsandTheirApplications...................................133 Hadj Ahmed Bouarara, Dr. Tahar Moulay University of Saida, Algeria Thischaptersubscribesintheframeworkofananalyticalstudyaboutthecomputationalintelligence algorithms.Thesealgorithmsarenumerousandcanbeclassifiedintwogreatfamilies:evolutionary algorithms(geneticalgorithms,geneticprogramming,evolutionarystrategy,differentialevolutionary, paddyfieldalgorithm)andswarmoptimizationalgorithms(particleswarmoptimisationPSO,antcolony optimization(ACO),bacteriaforagingoptimisation,wolfcolonyalgorithm,fireworksalgorithm,bat algorithm,cockroachescolonyalgorithm,socialspidersalgorithm,cuckoosearchalgorithm,waspswarm optimisation,mosquitooptimisationalgorithm).Wehavedetailedeachalgorithmfollowingastructured organization(theoriginofthealgorithm,theinspirationsource,thesummary,andthegeneralprocess). Thispaperisthefruitofmanyyearsofresearchintheformofsynthesiswhichgroupsthecontributions proposedbyvariousresearchersinthisfield.Itcanbethestartingpointforthedesigningandmodelling newalgorithmsorimprovingexistingalgorithms. Chapter 6 OptimizationofProcessParametersUsingSoftComputingTechniques:ACaseWithWire ElectricalDischargeMachining..........................................................................................................177 Supriyo Roy, Birla Institute of Technology, India Kaushik Kumar, Birla Institute of Technology, India J. Paulo Davim, University of Aveiro, Portugal MachiningofhardmetalsandalloysusingConventionalmachininginvolvesincreaseddemandof time,energyandcost.Itcausestoolwearresultinginlossofqualityoftheproduct.Non-conventional machining,ontheotherhandproducesproductwithminimumtimeandatdesiredlevelofaccuracy.In thepresentstudy,EN19steelwasmachinedusingCNCWireElectricaldischargemachiningwithpredefinedprocessparameters.MaterialRemovalRateandSurfaceroughnesswereconsideredasresponses forthisstudy.Thepresentoptimizationproblemissingleandaswellasmulti-response.Consideringthe complexitiesofthispresentproblem,experimentaldataweregeneratedandtheresultswereanalyzed byusingTaguchi,GreyRelationalAnalysisandWeightedPrincipalComponentAnalysisundersoft computingapproach.Responsesvarianceswiththevariationofprocessparameterswerethoroughly studiedandanalyzed;also‘bestoptimalvalues’wereidentified.Theresultshowsanimprovementin responsesfrommeantooptimalvaluesofprocessparameters.  Chapter 7 AugmentedLagrangeHopfieldNetworkforCombinedEconomicandEmissionDispatchwith FuelConstraint.................................................................................................................................... 221 Vo Ngoc Dieu, Ho Chi Minh City University of Technology, Vietnam Tran The Tung, Ho Chi Minh City University of Technology, Vietnam This chapter proposes an augmented Lagrange Hopfield network (ALHN) for solving combined economicandemissiondispatch(CEED)problemwithfuelconstraint.IntheproposedALHNmethod, theaugmentedLagrangefunctionisdirectlyusedastheenergyfunctionofcontinuousHopfieldneural network(HNN),thusthismethodcanproperlyhandleconstraintsbybothaugmentedLagrangefunction andsigmoidfunctionofcontinuousneuronsintheHNN.Fordealingwiththebi-objectiveeconomic dispatchproblem,theslopeofsigmoidfunctioninHNNisadjustedtofindthePareto-optimalfrontand thenthebestcompromisesolutionfortheproblemwillbedeterminedbyfuzzy-basedmechanism.The proposedmethodhasbeentestedonmanycasesandtheobtainedresultsarecomparedtothosefrom othermethodsavailabletheliterature.Thetestresultshaveshownthattheproposedmethodcanfind goodsolutionscomparedtotheothersforthetestedcases.Therefore,theproposedALHNcouldbea favourableimplementationforsolvingtheCEEDproblemwithfuelconstraint. Chapter 8 SpeakerRecognitionWithNormalandTelephonicAssameseSpeechUsingI-Vectorand Learning-BasedClassifier................................................................................................................... 256 Mridusmita Sharma, Gauhati University, India Rituraj Kaushik, Tezpur University, India Kandarpa Kumar Sarma, Gauhati University, India Speaker recognition is the task of identifying a person by his/her unique identification features or behaviouralcharacteristicsthatareincludedinthespeechutteredbytheperson.Speakerrecognition dealswiththeidentityofthespeaker.Itisabiometricmodalitywhichusesthefeaturesofthespeaker thatisinfluencedbyone’sindividualbehaviouraswellasthecharacteristicsofthevocalcord.Theissue becomesmorecomplexwhenregionallanguagesareconsidered.Here,theauthorsreportthedesignof aspeakerrecognitionsystemusingnormalandtelephonicAssamesespeechfortheircasestudy.Intheir work,theauthorshaveimplementedi-vectorsasfeaturestogenerateanoptimalfeaturesetandhaveused theFeedForwardNeuralNetworkfortherecognitionpurposewhichgivesafairlyhighrecognitionrate. Chapter 9 ANewSVMMethodforRecognizingPolarityofSentimentsinTwitter.......................................... 281 Sanjiban Sekhar Roy, VIT University, India Marenglen Biba, University of New York – Tirana, Albania Rohan Kumar, VIT University, India Rahul Kumar, VIT University, India Pijush Samui, NIT Patna, India Onlinesocialnetworkingplatforms,suchasWeblogs,microblogs,andsocialnetworksareintensively beingutilizeddailytoexpressindividual’sthinking.Thispermitsscientiststocollecthugeamountsof dataandextractsignificantknowledgeregardingthesentimentsofalargenumberofpeopleatascale thatwasessentiallyimpracticalacoupleofyearsback.Therefore,thesedays,sentimentanalysishasthe potentialtolearnsentimentstowardspersons,objectandoccasions.Twitterhasincreasinglybecome  a significantsocialnetworkingplatformwherepeoplepostmessagesofupto140charactersknownas ‘Tweets’.Tweetshavebecomethepreferredmediumforthemarketingsectorasuserscaninstantlyindicate customersuccessorindicatepublicrelationsdisasterfarmorequicklythanawebpageortraditional mediadoes.Inthispaper,wehaveanalyzedtwitterdataandhavepredictedpositiveandnegativetweets withhighaccuracyrateusingsupportvectormachine(SVM). Chapter 10 AutomaticGenerationControlofMulti-AreaInterconnectedPowerSystemsUsingHybrid EvolutionaryAlgorithm...................................................................................................................... 292 Omveer Singh, Maharishi Markandeshwar University, India Anewtechniqueofevaluatingoptimalgainsettingsforfullstatefeedbackcontrollersforautomatic generationcontrol(AGC)problembasedonahybridevolutionaryalgorithms(EA)i.e.geneticalgorithm (GA)-simulatedannealing(SA)isproposedinthischapter.ThehybridEAalgorithmcantakedynamic curveperformanceashardconstraintswhicharepreciselyfollowedinthesolutions.Thisisincontrast tothemodernandsinglehybridevolutionarytechniquewheretheseconstraintsaretreatedassoft/hard constraints.Thistechniquehasbeeninvestigatedonanumberofcasestudiesandgivessatisfactorysolutions. Thistechniqueisalsocomparedwithlinearquadraticregulator(LQR)andGAbasedproportionalintegral (PI)controllers.Thisprovestobeagoodalternativeforoptimalcontroller’sdesign.Thistechniquecan beeasilyenhancedtoincludemorespecificationsviz.settlingtime,risetime,stabilityconstraints,etc. Chapter 11 MathematicalOptimizationbyUsingParticleSwarmOptimization,GeneticAlgorithm,and DifferentialEvolutionandItsSimilarities.......................................................................................... 325 Shailendra Aote, Ramdeobaba College of Engineering and Management, India Mukesh M. Raghuwanshi, Yeshwantrao Chavan College of Engineering, India Tosolvetheproblemsofoptimization,variousmethodsareprovidedindifferentdomain.Evolutionary computing(EC)isoneofthemethodstosolvetheseproblems.MostlyusedECtechniquesareavailable likeParticleSwarmOptimization(PSO),GeneticAlgorithm(GA)andDifferentialEvolution(DE). Thesetechniqueshavedifferentworkingstructurebuttheinnerworkingstructureissame.Different namesandformulaearegivenfordifferenttaskbutultimatelyalldothesame.Herewetriedtofindout thesimilaritiesamongthesetechniquesandgivetheworkingstructureineachstep.Allthestepsare providedwithproperexampleandcodewritteninMATLAB,forbetterunderstanding.Herewestarted ourdiscussionwithintroductionaboutoptimizationandsolutiontooptimizationproblemsbyPSO,GA andDE.Finally,wehavegivenbriefcomparisonofthese. Chapter 12 GA_SVM:AClassificationSystemforDiagnosisofDiabetes.......................................................... 359 Dilip Kumar Choubey, Birla Institute of Technology Mesra, India Sanchita Paul, Birla Institute of Technology Mesra, India Themodernsocietyispronetomanylife-threateningdiseaseswhichifdiagnosisearlycanbeeasily controlled.Theimplementationofadiseasediagnosticsystemhasgainedpopularityovertheyears.The mainaimofthisresearchistoprovideabetterdiagnosisofdiabetes.Therearealreadyseveralexisting methods,whichhavebeenimplementedforthediagnosisofdiabetes.Inthismanuscript,firstly,Polynomial Kernel,RBFKernel,SigmoidFunctionKernel,LinearKernelSVMusedfortheclassificationofPIDD.  SecondlyGAusedasanAttributeselectionmethodandthenusedPolynomialKernel,RBFKernel, SigmoidFunctionKernel,LinearKernelSVMonthatselectedattributesofPIDDforclassification.So, herecomparedtheresultswithandwithoutGAinPIDD,andLinearKernelprovedbetteramongallof thenotedaboveclassificationmethods.ItdirectlyseemsinthepaperthatGAisremovinginsignificant features,reducingthecostandcomputationtimeandimprovingtheaccuracy,ROCofclassification. Theproposedmethodcanbealsousedforotherkindsofmedicaldiseases. Chapter 13 TheInsectsofNature-InspiredComputationalIntelligence............................................................... 398 Sweta Srivastava, B.I.T. Mesra, India Sudip Kumar Sahana, B.I.T. Mesra, India Thedesirablemeritsoftheintelligentcomputationalalgorithmsandtheinitialsuccessinmanydomains haveencouragedresearcherstoworktowardstheadvancementofthesetechniques.Amajorplunge inalgorithmicdevelopmenttosolvetheincreasinglycomplexproblemsturnedoutasbreakthrough towardsthedevelopmentofcomputationalintelligence(CI)techniques.Natureprovedtobeoneofthe greatestsourcesofinspirationfortheseintelligentalgorithms.Inthischapter,computationalintelligence techniquesinspiredbyinsectsarediscussed.Thesetechniquesmakeuseoftheskillsofintelligent agentbymimickinginsectbehaviorsuitablefortherequiredproblem.Thediversitiesinthebehaviorof theinsectfamiliesandsimilaritiesamongthemthatareusedbyresearchersforgeneratingintelligent techniquesarealsodiscussedinthischapter. Chapter 14 Bio-InspiredComputationalIntelligenceandItsApplicationtoSoftwareTesting............................ 429 Abhishek Pandey, UPES Dehradun, India Soumya Banerjee, BIT Mesra, India Bioinspiredalgorithmsarecomputationalprocedureinspiredbytheevolutionaryprocessofnature andswarmintelligencetosolvecomplexengineeringproblems.Intherecenttimesithasgainedmuch popularityintermsofapplicationstodiverseengineeringdisciplines.Nowadaysbioinspiredalgorithms arealsoappliedtooptimizethesoftwaretestingprocess.Inthischapterauthorswilldiscusssomeof thepopularbioinspiredalgorithmsandalsogivestheframeworkofapplicationofthesealgorithmsfor softwaretestingproblemssuchastestcasegeneration,testcaseselection,testcaseprioritization,test caseminimization.Bioinspiredcomputationalalgorithmsincludesgeneticalgorithm(GA),genetic programming (GP), evolutionary strategies (ES), evolutionary programming (EP) and differential evolution(DE)intheevolutionaryalgorithmscategoryandAntcolonyoptimization(ACO),Particle swarmoptimization(PSO),ArtificialBeeColony(ABC),Fireflyalgorithm(FA),Cuckoosearch(CS), Batalgorithm(BA)etc.intheSwarmIntelligencecategory(SI).  Chapter 15 Quantum-InspiredComputationalIntelligenceforEconomicEmissionDispatchProblem.............. 445 Fahad Parvez Mahdi, Universiti Teknologi Petronas, Malaysia Pandian Vasant, Universiti Teknologi Petronas, Malaysia Vish Kallimani, Universiti Teknologi Petronas, Malaysia M. Abdullah-Al-Wadud, King Saud University, Saudi Arabia Junzo Watada, Universiti Teknologi Petronas, Malaysia Economicemissiondispatch(EED)problemsareoneofthemostcrucialproblemsinpowersystems. Growingenergydemand,limitedreservesoffossilfuelandglobalwarmingmakethistopicintothe centerofdiscussionandresearch.Inthischapter,wewilldiscusstheuseandscopeofdifferentquantum inspiredcomputationalintelligence(QCI)methodsforsolvingEEDproblems.Wewillevaluateeach previouslyusedQCImethodsforEEDproblemanddiscusstheirsuperiorityandcredibilityagainst othermethods.WewillalsodiscussthepotentialityofusingotherquantuminspiredCImethodslike quantumbatalgorithm(QBA),quantumcuckoosearch(QCS),andquantumteachingandlearningbased optimization(QTLBO)techniqueforfurtherdevelopmentinthisarea. Chapter 16 IntelligentExpertSystemtoOptimizetheQuartzCrystalMicrobalance(QCM)Characterization Test:IntelligentSystemtoOptimizetheQCMCharacterizationTest............................................... 469 Jose Luis Calvo-Rolle, University of A Coruña, Spain José Luis Casteleiro-Roca, University of A Coruña, Spain María del Carmen Meizoso-López, University of A Coruña, Spain Andrés José Piñón-Pazos, University of A Coruña, Spain Juan Albino Mendez-Perez, Universidad de La Laguna, Spain Thischapterdescribesanapproachtoreducesignificantlythetimeinthefrequencysweeptestofa QuartzCrystalMicrobalance(QCM)characterizationmethodbasedontheresonanceprincipleofpassive components.Onthistest,thespenttimewaslarge,becauseitwasnecessarycarryoutabigfrequency sweepduetothefactthattheresonancefrequencywasunknown.Moreover,thisfrequencysweephas greatstepsandconsequentlylowaccuracy.Then,itwasnecessarytoreducethesweepsanditssteps graduallywiththeaimtoincreasetheaccuracyandtherebybeingabletofindtheexactfrequency.An intelligentexpertsystemwascreatedasasolutiontothedisadvantagedescribedofthemethod.This modelprovidesamuchsmallerfrequencyrangethantheinitiallyemployedwiththeoriginalproposal. Thisfrequencyrangedependsofthecircuitcomponentsofthemethod.Then,thankstothenewapproach oftheQCMcharacterizationisachievedbetteraccuracyandthetesttimeisreducedsignificantly. Chapter 17 OptimizationThroughNature-InspiredSoft-ComputingandAlgorithmonECGProcess................ 489 Goutam Kumar Bose, Haldia Institute of Technology, India Pritam Pain, Haldia Institute of Technology, India Inthepresentresearchworkselectionofsignificantmachiningparametersdependingonnature-inspired algorithmisprepared,duringmachiningalumina-aluminuminterpenetratingphasecompositesthrough electrochemical grinding process. Here during experimentation control parameters like electrolyte concentration(C),voltage(V),depthofcut(D)andelectrolyteflowrate(F)areconsidered.Theresponse dataareinitiallytrainedandtestedapplyingArtificialNeuralNetwork.Theparadoxicalresponseslike  highermaterialremovalrate(MRR),lowersurfaceroughness(Ra),lowerovercut(OC)andlowercutting force(Fc)areaccomplishedindividuallybyemployingCuckooSearchAlgorithm.Amultiresponse optimizationforalltheresponseparametersiscompiledprimarilybyusingGeneticalgorithm.Finally, inordertoachieveasinglesetofparametriccombinationforalltheoutputssimultaneouslyfuzzy basedGreyRelationalAnalysistechniqueisadopted.Thesenature-drivensoftcomputingtechniques corroborateswellduringtheparametricoptimizationofECGprocess. Chapter 18 AnOverviewoftheLastAdvancesandApplicationsofArtificialBeeColonyAlgorithm.............. 520 Airam Expósito Márquez, University of La Laguna, Spain Christopher Expósito-Izquierdo, University of La Laguna, Spain SwarmIntelligenceisdefinedascollectivebehaviorofdecentralizedandself-organizedsystemsofa naturalorartificialnature.Inthelastyearsandtoday,SwarmIntelligencehasproventobeabranchof ArtificialIntelligencethatisabletosolvingefficientlycomplexoptimizationproblems.SomeofwellknownexamplesofSwarmIntelligenceinnaturalsystemsreportedintheliteraturearecolonyofsocial insectssuchasbeesandants,birdflocks,fishschools,etc.Inthisrespect,ArtificialBeeColonyAlgorithm isanatureinspiredmetaheuristic,whichimitatesthehoneybeeforagingbehaviourthatproducesan intelligentsocialbehaviour.ABChasbeenusedsuccessfullytosolveawidevarietyofdiscreteand continuousoptimizationproblems.InordertofurtherenhancethestructureofArtificialBeeColony, thereareavarietyofworksthathavemodifiedandhybridizedtoothertechniquesthestandardversion ofABC.Thisworkpresentsareviewpaperwithasurveyofthemodifications,variantsandapplications oftheArtificialBeeColonyAlgorithm. Chapter 19 ASurveyoftheCuckooSearchandItsApplicationsinReal-WorldOptimizationProblems........... 541 Christopher Expósito-Izquierdo, University of La Laguna, Spain Airam Expósito-Márquez, University of La Laguna, Spain ThechapterathandseekstoprovideageneralsurveyoftheCuckooSearchAlgorithmanditsmost highlightedvariants.TheCuckooSearchAlgorithmisarelativelyrecentnature-inspiredpopulationbasedmeta-heuristicalgorithmthatisbaseduponthelifestyle,egglaying,andbreedingstrategyof somespeciesofcuckoos.Inthiscase,theLévyflightisusedtomovethecuckooswithinthesearch spaceoftheoptimizationproblemtosolveandobtainasuitablebalancebetweendiversificationand intensification.Asdiscussedinthischapter,theCuckooSearchAlgorithmhasbeensuccessfullyapplied toawiderangeofheterogeneousoptimizationproblemsfoundinpracticalapplicationsoverthelast fewyears.Someofthereasonsofitsrelevancearethereducednumberofparameterstoconfigureand itseaseofimplementation.
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值