基于多粒度的旅行商问题描述及蚁群优化算法.pdfVIP

计算机研究与发展 ISSN 1000—12391CN 11—1777门rP JournalofComputerResearchandDevelopment 47(3)：434-444，2010 基于多粒度的旅行商 问题描述及其蚁群优化算法 冀俊忠 黄 振 刘椿年 代启国 (北京丁业大学计算机学院多媒体与智能软件技术北京市重点实验室 北京 100124) (jjz01@bjut．edu．cn) AnAntColonyAlgorithm Based on Multiple-Grain Representation for theTraveling SalesmanProblems JiJunzhong，HuangZhen，LiuChunnian，andDaiQiguo (BeijingMunicipalKeyLaboratoryofMultimediaandIntelligentSoftwareTechnology，ColiegeofComputerScienceand Technology，BeijingUniversityofTechnology，Beijing100124) Abstract Antcolonyoptimization (ACO)isapopulation—basedmetaheuristictechniquetoeffectively solvecombinationoptimizationproblems．However，itisstillanactiveresearchtopichow tOimprove theperformanceofACO algorithms．Though therearemanyalgorithmstoeffectivelysolvetraveling salesmanproblems(TSPs)，thereisanapplicationbottleneckthattheACO algorithm coststOOmuch timeinorderto getanoptimalsolution．ToimprovethetimeperformanceofACO insolving1arge scaleTSPs，afastalgorithm ispresentedinthispaper．Firstly．anovelmultiple—grainrepresentation modelofTSPsisproposed．Basedonthemodel，anew algorithm forTSPsispresented，whichmainly containssix phases，i．e．a granularity partition algorithm based on density clustering，an ACO algorithm based on the coarse grain，a connection algorithm between two granularities，an ACO algorithm in the same granularity， a fusion algorithm among granularities， and a subsection optim izationalgorithm regardlessofgranularities．Theanalysisofcomputation complexityand the experimentalresultsforlargenumberofTSPsdemonstratethattheproposedalgorithm can greatly improvethespeedofconvergenceincontrasttotheclassicalACO algorithm ，andishighlycompetitive intimeperformancecomparedwithsome1atestelitistACO algorithms． Key words TSP；multiple—grain city model；ACO algorithm ； clustering algorithm ； subsection optimization 摘 要 针对蚁群算法在求解大规模旅