29445.基于msrm的多峰函數(shù)極值求解算法

1、分類號學(xué)校代碼10542密級基于MSRM的多峰函數(shù)極值求解算法MuItiextremumSolutionBasedonMSRM研究生姓名余沛毅指導(dǎo)教師姓名、職稱劉剛副教授學(xué)科專業(yè)研究方向計(jì)算機(jī)技術(shù)人工智能與數(shù)據(jù)挖掘湖南師范大學(xué)學(xué)位評定委員會辦公室二零一四年五月ABSTRACTWhilesolvingcomplexpracticalproblemsofoptimization，weoftenmeetseveraloptimizationp

2、roblemsaboutlocalorglobaloptimumsoffunctions，calledmulti—extremumfunctionoptimizationproblemsInthemathematics，architecture，engineering，machineryandmanyotherpracticalfields，weneedtransformtheproblemsintomulti—extremumfunc

3、tiontosolvethemMulti—extremumfunctiontakesanimportantapplication，SOitgainslotsofattentionsfrommanyresearchersThepreviousmethodsforsolvingmulti——extremumfunctionoptimizationproblemsincludeclassicalnumericalmethod，extensiv

4、elyusedglobalsearchingalgorithm，crowdingnichegeneticalgorithmsandSOonHowever，theevaluationofthesearchingperformanceofthesemethodsmainlyfocusesontheproblemof‘howmanyglobaloptimumscouldbefound’Thatistosay，thesemethodscanon

5、lybeusedinsolvingtheproblemofsearchingtheglobaloptimumpeaks，insteadoffiguringoutthesearchingoflocalpeaksGraspingthewholeinformationoffunctionsinpracticalapplicationstoprovidevariouschoicesorinformationfordecision—makers，