蟻群算法改進(jìn)及其聚類分析應(yīng)用.pdf

1、哈爾濱理工大學(xué)碩士學(xué)位論文蟻群算法改進(jìn)及其聚類分析應(yīng)用姓名：郭明申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：丁華福20090301哈爾濱理工大學(xué)工學(xué)碩士學(xué)位論文TheImprovementoftheAntColonyAlgorithmandItsUsingonClusteringAnalysisAbstractAntColonyAlgorithmisanewdevelopedbionicoptimizationalgorithmwh

2、ichsimulatetheants’findingfoodprocessBecauseofitssimplemechanism，plusfeedbackparallelizationstronglustinessandexcellentdistributedcomputationalmechanism，ithasshownitsexcellentcapabilityandhugedevelopingpotentialinsolving

3、manycomplicatedoptimizationproblemsDespitethestricttheoreticalbasisofantcolonyalgorithmhasnotyetbeenlaid，andbothathomeandabroadrelatedresearchisstillinthepilotphaseoftheexplorationandpreliminaryapplicationtheantcolonyalg

4、orithm’Sstudyhadbeenfromasingletravelingsalesmanproblempenetratedintothemultiapplicationareas，fromsettlementofone—dimensionalstaticoptimizationproblemdevelopedtosolvethemultidimensionalanddynamiccombinatorialoptimization

5、problems，andfromdiscretedomainwithinthescopeofresearchgraduallyextendedtoacontinuousdomainwithinthescopeofresearchHowevertherearestillsomefaultsinantcolonyalgorithm：asinperformance，thediversityandstabilityofthesolustions

6、searchedbyantcolony孤℃inconflict、^，itllconvergencespeedofthealgorithmThisisbecausetheindividualmovementofantcolonyisrandomAlthoughthroughtheexchangeofinformationtotheevolutiontowardtheoptimalpaththeyCanhardlyfmdaoptimumon

7、efromamassofpathswithinashorttimewhentheproblemscaleislargeenougheyelesslyacceleratetheconvergencespeedwillmaketheants’localsearchandleadtoprematureandstagnationofthealgorithmeasilyInthispaperwefirstintroduceclassicalant

8、colonyalgorithlmandgiveasummaryaboutitscurrentresearchworkThroughtheresearchofthebasicantcolonyalgorithm’Sshortcomings，wedrawonthemanyadvantagesofartificialfishswarmalgorithmAlthoughtherearesomegapsoftheartificialfishind