Integration of genetic algorithms and GIS for optimal location search

《Integration of genetic algorithms and GIS for optimal location search》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、InternationalJournalofGeographicalInformationScienceVol.19,No.5,May2005,581–601ResearchArticleIntegrationofgeneticalgorithmsandGISforoptimallocationsearchXIALI{andANTHONYGAR-ONYEH{{SchoolofGeographyandPlanning,SunYat-senUniversity,135WestXingangRd.,Guangzhou510275,P.R.China

2、;e-mail:lixia@graduate.hku.hk;gplx@zsu.edu.cn{CentreofUrbanPlanningandEnvironmentalManagement,TheUniversityofHongKong,PokfulamRoad,HongKongSAR,P.R.China;e-mail:hdxugoy@hkucc.hku.hk(Received24February2004;accepted23September2004)Optimallocationsearchisfrequentlyrequiredinman

3、yurbanapplicationsforsitingoneormorefacilities.However,thesearchmaybecomeverycomplexwhenitinvolvesmultiplesites,variousconstraintsandmultiple-objectives.Theexhaustiveblind(brute-force)searchwithhigh-dimensionalspatialdataisinfeasibleinsolvingoptimizationproblemsbecauseofahu

4、gecombinatorialsolutionspace.Intelligentsearchalgorithmscanhelptoimprovetheperformanceofspatialsearch.ThisstudywilldemonstratethatgeneticalgorithmscanbeusedwithGeographicalInformationsystems(GIS)toeffectivelysolvethespatialdecisionproblemsforoptimallysittingnsitesofafacilit

5、y.DetailedpopulationandtransportationdatafromGISareusedtofacilitatethecalculationoffitnessfunctions.MultipleplanningobjectivesarealsoincorporatedintheGAprogram.ExperimentsindicatethattheproposedmethodhasmuchbetterperformancethansimulatedannealingandGISneighborhoodsearchmeth

6、ods.TheGAmethodisveryconvenientinfindingthesolutionwiththehighestutilityvalue.Keywords:Geneticalgorithms;GIS;Optimallocation;Multipleobjectives;Simulatedannealing1.IntroductionAnoftenencounteredspatialdecisionproblemistosearchforthebestsiteorsitestoaccommodateoneormorefacil

7、itiestogeneratethebestutilityvalues(e.g.themaximumpopulationcoverageandminimumtransportcost).Traditionallocation-allocationmethodsbeforeGISdatawereavailableonlyuserelativelysmalldatasets(Church1999).Thegeneralfacilitylocationproblemanditsvariants,includingmostlocation-alloc

8、ationandp-medianproblems,areknowntobeNP-hardcombinatorialoptimizationproblems.Most