改進(jìn)遺傳算法及其應(yīng)用.pdf

1、華北電力大學(xué)（保定）碩士學(xué)位論文改進(jìn)遺傳算法及其應(yīng)用姓名：申紅蓮申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：應(yīng)用數(shù)學(xué)指導(dǎo)教師：張國(guó)立20071220華北電力大學(xué)碩士學(xué)位論文摘要摘要基于距離測(cè)度的實(shí)數(shù)編碼遺傳模擬退火算法是2004年蔡良偉提出的一種具有全局搜索能力的新算法。本文對(duì)該算法進(jìn)行了改進(jìn)，定義了基于距離測(cè)度和適應(yīng)度的自適應(yīng)變異概率，采用改進(jìn)的交叉策略和模擬退火操作，并在群體趨于一致時(shí)保留最優(yōu)個(gè)體，重新產(chǎn)生其它新個(gè)體。對(duì)帶邊界約束函數(shù)優(yōu)化問(wèn)題進(jìn)行仿真計(jì)

2、算，結(jié)果表明改進(jìn)后算法收斂概率較高，收斂速度較快。針對(duì)TSP問(wèn)題，本文定義了基于距離測(cè)度和適應(yīng)度的的自適應(yīng)交叉概率，采用改進(jìn)的交換啟發(fā)交叉算子，產(chǎn)生不差于父代的個(gè)體，根據(jù)最優(yōu)和次優(yōu)個(gè)體的差異，采用2變換法產(chǎn)生新個(gè)體或者進(jìn)行模擬退火操作。仿真實(shí)驗(yàn)結(jié)果表明新算法收斂概率較高，得到最優(yōu)解的平均代數(shù)要遠(yuǎn)遠(yuǎn)小于其他所比較的算法。關(guān)鍵詞：遺傳算法，模擬退火算法，距離測(cè)度，適應(yīng)度，函數(shù)優(yōu)化ABSTRACTReal—codedgeneticanneal

3、ingalgorithmbasedondistancemeasurementisanewalgorithmfor哲obal—searchingwhichproposedbyCaiLiangWeiin2004Thealgorithmhasbeenimprovedinthispaperitdefinedthemutationprobabilityaccordingtodistancemeasurementandfitness，adopted

4、improvedcrossoverandsimulatedannealingoperationInaddition，whenpopulationtendedtouniform，itreservedthebestindividualandreproducedotherindividualsApplyingittothefunctionoptimizationproblemwithboundaryconstraints，thesimulat

5、edresultsshowthattheimprovedalgorithmismoreeffectiveinrealizingthehighconvergenceprobabilityandrapidevolvingspeedInordertosolveTSPproblemitdefinedthecrossingprobabilitybasedondistancemeasurementandfitness，adoptedimproved

6、exchangeheuristiccrossovcroperatorwhichthenewindividualisnoworsethantheoldone。Accordingtothedifferencebetweenthebestindividualandthebetterindividual，itadoptedthemethodthatusedoubleexchangeproducenewindividualsordosimulat

7、edannealingopermionThesimulatedresultsshowthatthenewalgorithmhashighconvergenceprobabilityandthegenerationnumbersisfarlessthanothercomparativealgorithmsShenHonglian(DepartmentofMathematicsandPhysics)DirectedbyprofZhangGu