基于生物免疫的動(dòng)態(tài)環(huán)境優(yōu)化算法及其應(yīng)用.pdf

1、貴州大學(xué)碩士學(xué)位論文基于生物免疫的動(dòng)態(tài)環(huán)境優(yōu)化算法及其應(yīng)用姓名：錢淑渠申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：運(yùn)籌學(xué)與控制論指導(dǎo)教師：張著洪20070401BiologicalimmunebasedimmuneoptimizationalgorithmsindynamicenvironmentsandtheirapplicationsAbstract刪Ieintelligentalgorithmsarcwidelyappliedtostaticopti

2、mizationproblems。aconsiderablenumberofvaluableachievementshavebeenreposedHoweverItisstudiedrarelytodealwithdynamicoptimizationproblems，inwhichthekeysolvingthisclassofproblemsistodesignintelligentoptimizationtechniquescap

3、ableofstronglytrackingthechangingenvironmentovertimeandachievingreasonabletradeoffbetweenperformanceeffectandefficiencySo，ithasbecomeanimportantresearchtopictodesignmoreadvancedintelligentoptimizationtechniquestocopewith

4、dynamicoptimizationproblemsFromtheangleofintelligentoptimization，recentlythemainresearchworkhasbeenfocosedonmodifyingclassicaIgeneticalgorithmsbutlessprogressTherefore，inthisdissertation，basedonthetheoryofbiologicalimmun

5、esystems，threekindsofimmuneoptimizationalgorithmsindynamicenvironmentsareproposedfordynamicsingle—objectiveoptimization，dynamicmultiobjectiveoptimizationandonlinegreenhousecontr01respectivelyThesealgorithmsareexaminedthr

6、mlghuuRl舐calexperiments，comparativeanalysisandapplicationsThemainworkissummedupasfollows：AAnovelimmuneoptimizationalgorithmindynamicenvironmentsisproposedtodeal謝111dynamicsin鰣eo場(chǎng)ectiveoptimizationproblemsIndesignofthealg

7、oriflun，severaloperatorsareestablished，ie，dynamicevolutionrelyingonantibodylearningantibodyrearrangementdependingongenedrift，dynamicalmemorypoolcomposedofmanymemorysubsetsbuiltupontheimmunemenlorycharacteristicsandthefun

8、ctionofdynamicmaintenanceinwhichthepoolutilizestheaveragelinkagetokeepthoseexcellentmemorycellsandenvironmentalidentifierandgenerationruleofinitialantibodypopulationsrelatedtodynamicsurveillanceThealgorithmpossessessuchp

9、ropertiesasstructuralsimplicityfeasibilityanddynamicalregulationoftheexecutiontimefordifferentenvironmentsExperimentalrasultsandcomparisonillustruteitssuperiorityincludingtheeffectivetradeofrbci3ⅣecnperfornlanCeeffectand

10、efficiencyaswellasthepotentialforcomplexdynamicalhi曲dimensionaloptimizationproblems＆AnonlinegreenhousecontrolimmuneoptimizationalgorithmispropOsedtosolveaclassofclassicalgreenhousecontrolproblemswithdynamicenvironmentsIn

11、thealgorithm，dynamicmemorypoolisdesignedtopreserve山eexcellentantibodi黜fromthepreviousenvironmentsbyusingthedynamicupdatemechanismofmemoryeellsintheimnnnlesystemforreference，whilethesizesofevolvingpopulatioilsareadjusteda

12、ndtheirantibodiesarechosedynamicallyintermsoftheaveragedensitiesofthepopulationsBesides，antibodiespropagatetheirclonesbymeansofthn4raffinitiesassociatedtothegivenantigenThl|oughcomparisonwithseveralevolutionalgorithmsind

13、ynamicenvironments，numericalexperimentsshowthattheproposedalgorithmcantrackstronglychangingenvironmentswithgreatpracticalperspectiveCAdynamicmultiobjectiveinuBaneoptimizationalgorithraisproposedbasedollthecharacteristics

14、ofdynamicmultiobiecti、(eoptimizationandassociatedtosomemetaphorsoftheinlnlunesystemIndesignofthealgorithm，someantibodiesarechosetoparticipateinevolutionthroughsortinglevelselectionwhiletheaffinityofanantibodyispropOrtion

15、altotheaveragedensityofallantibodiesinits‘neighborhood，beingdependentonthepositionoftheantibodyOntheotherhand。eachcloneundergoesmutationwithitsmutationprobabilityconverselyproportionaltotheafIjni“ofitsparentandsuchimmune

16、functionsasimmnnememoryanddynamicmaintenancetogetherwitlItheaveragelinkagemethod，areusedtodesignenvironmentalmemorysetandmemoryp001“11lroughcomparisonwithtwepresentativeevolutionaryalgorithmsandaneighborsearchalgorithm，n