資源描述:
《基于遺傳退火算法的數(shù)據(jù)庫(kù)多連接查詢(xún)優(yōu)化研究與應(yīng)用.pdf》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在教育資源-天天文庫(kù)。
1、分類(lèi)號(hào)UDC:密級(jí)編號(hào)基于遺傳退火算法的數(shù)據(jù)庫(kù)多連接查詢(xún)優(yōu)化研究與應(yīng)用ResearchandApplicationOHMulti-JoinQueryOptimizationofDatabaseBasedonGeneticAlgorithmandSimulatedAnnealing學(xué)位授予單位及代碼:量壹理王鑫堂(!Q!S§2學(xué)科專(zhuān)業(yè)名稱(chēng)及代碼:讓篁扭塾鮭皇堡途(QSl2絲)研究方向:塑堡莊曩筮申請(qǐng)學(xué)位級(jí)別:亟±指導(dǎo)教師:堂塹虹副塾量研究生:塞亙塑論文起止時(shí)間:?200711m2009.4摘要當(dāng)今的信息技術(shù)時(shí)代,數(shù)據(jù)庫(kù)
2、已經(jīng)成為管理信息和挖掘信息巨大潛能的基本和必需工具。隨著時(shí)間的推移,數(shù)據(jù)庫(kù)中的數(shù)據(jù)量日益增加,從海量數(shù)據(jù)中查詢(xún)出滿(mǎn)足用戶(hù)條件的數(shù)據(jù)就要耗費(fèi)大量時(shí)間,為了把數(shù)據(jù)庫(kù)的性能維持在可接受的水平上,眾多學(xué)者展開(kāi)了對(duì)查詢(xún)優(yōu)化技術(shù)的研究。而數(shù)據(jù)查詢(xún)中最復(fù)雜的是多表連接查詢(xún),這極大的影響著數(shù)據(jù)查詢(xún)效率,因此多連接查詢(xún)優(yōu)化是數(shù)據(jù)庫(kù)優(yōu)化的關(guān)鍵問(wèn)題之一。本文根據(jù)數(shù)據(jù)庫(kù)多連接查詢(xún)優(yōu)化的特點(diǎn),將側(cè)重于全局搜索的遺傳算法與側(cè)重于局部搜索的模擬退火算法相結(jié)合,提出了一種基于改進(jìn)的遺傳退火混合算法的查詢(xún)優(yōu)化策略。先從一組隨機(jī)產(chǎn)生的初始種群開(kāi)始全局最
3、優(yōu)解的搜索,通過(guò)選擇、交叉、變異等遺傳操作產(chǎn)生新一代種群,然后對(duì)新個(gè)體進(jìn)行模擬退火操作,將結(jié)果作為下一代種群中的個(gè)體。如此反復(fù)迭代進(jìn)行,到滿(mǎn)足最終條件為止。仿真實(shí)驗(yàn)驗(yàn)證了該算法的有效性。關(guān)鍵詞:數(shù)據(jù)庫(kù)多連接查詢(xún)優(yōu)化連接樹(shù)遺傳算法模擬退火ABSTRACTNowadays,theinformationtechnologyisdevelopingveryfast.Databasehasbeenabasicandnecessaryinstrumentformanagingorexcavminggreatnesspotenti
4、alofinformation.Wimthetimepassing,theamountofdatagoesincreasingly,toquerythedatawhichadapttousersrequirementistime—consuming.Asaresult,numbersofscholarsdevelopedtheresearchonqueryoptimizationtechniquesinordertomaintaintheperformance.Theproblemofmulti-joinqueryi
5、sverycomplicated,itlargelyinfluencestheefficiencyofdataquery,SOoptimizationofmulti-joinqueryisoneofthekeyproblems.Thealgorithmofmulti-joinqueryoptimizationbasedongeneticalgorithmwithoverallsearchabilityandsimulatedannealingwithlocalsearchabilityareproposedbycom
6、biningwithcharacteristicofmulti-joinoptimization.Startinganoptimal.solution.searchtotheoverallsituationinagroupofinitialpopulation,whichisrandomselected.Anewgenerationofpopulationwillbeproducedaftertheselectionstrategy,crossoverandmutation.Andthenthesimulatedan
7、nealingisappliedtothosenewpopulations,andtheresultisusedastheunitofthenextgenerationpopulation.Theaboveprocessisoperatedrepeatedlyanditerative,untiltheresultmeetsthefinalqualification.Thesimulationexperimenthasproveditsefficiency.Keywords:databasemulti-joinquer
8、yoptimizationjointreegeneticalgorithmsimulatedannealingⅡ長(zhǎng)春理工大學(xué)碩士學(xué)位論文原創(chuàng)性聲明本人鄭重聲明:所呈交的碩士學(xué)位論文,《基于遺傳退火算法的數(shù)據(jù)庫(kù)多連接查詢(xún)優(yōu)化研究與應(yīng)用》是本人在指導(dǎo)教師的指導(dǎo)下,獨(dú)立進(jìn)行研究工作所取得的成果。除文中已經(jīng)注明引用的內(nèi)容外,本論文不包含任何其他個(gè)人或集