時(shí)間窗約束下的車輛路徑問題遺傳算法【外文翻譯】

1、外文翻譯原文GeicAlgithmsftheVehicleRoutingProblemwithTimeWindowsMaterialSource:SpecialissueonBioinfmaticsGeicAlgithmsAuth:OlliBrysy1IntroductionVehicleRoutingProblems(VRP)areallaroundusinthesensethatmanyconsumerproductssuchass

2、oftdrinksbeerbreadsnackfoodsgasolinepharmaceuticalsaredeliveredtoretailoutletsbyafleetoftruckswhoseoperationfitsthevehicleroutingmodel.InpracticetheVRPhasbeenrecognizedasoneofthegreatsuccessstiesofoperationsresearchithas

3、beenstudiedwidelysincethelatefifties.Publicservicescanalsotakeadvantageofthesesystemsindertoimprovetheirlogisticschain.Garbagecollectiontowncleaningtakesaneverincreasingpartofthebudgetoflocalauthities.Atypicalvehiclerout

4、ingproblemcanbedescribedastheproblemofdesigningleastcostroutesfromonedepottoasetofgeographicallyscatteredpoints(citiessteswarehousesschoolscustomersetc).Theroutesmustbedesignedinsuchawaythateachpointisvisitedonlyoncebyex

5、actlyonevehicleallroutesstartendatthedepotthetotaldemsofallpointsononeroutemustnotexceedthecapacityofthevehicle.TheVehicleRoutingProblemwithTimeWindows(VRPTW)isageneralizationoftheVRPinvolvingtheaddedcomplexitythateveryc

6、ustomershouldbeservedwithinagiventimewindow.AdditionalcomplexitiesencounteredintheVRPTWarelengthofrouteconstraintarisingfromdepottimewindowscostofwaitingtimewhichisincurredwhenavehiclearrivestooearlyatacustomerlocation.S

7、pecificexamplesofproblemswithtimewindowsincludebankdeliveriespostaldeliveriesindustrialrefusecollectionschoolbusroutingsituationswherethecustomermustprovideaccessverificationpaymentupondeliveryoftheproductservice[Solomon

8、Desrosiers1988].Besidesbeingoneofthemostimptantproblemsofoperationsresearchinpracticaltermsthevehicleroutingproblemisalsooneofthemostdifficultproblemstosolve.Itisquiteclosetooneofthemostfamouscombinatialoptimizationprobl

9、emstheTravelingSalespersonProblem(TSP)whereonlyoneinvehicleroutingcontext.Inallotherapproachesfvehicleroutingproblemwithtimewindowstheencodingissueisdisregarded.3Applicationsfvehicleroutingproblemwithtimewindows[Thangiah

10、1995]describesamethodcalledGIDEONthatassignscustomerstovehiclesbypartitioningthecustomersintosectsbygeicalgithmcustomerswithineachfmedsectareroutedusingthecheapestionmethodof[GoldenStewart1985].Inthenextsteptheroutesarei

11、mprovedusingλexchangesintroducedby[Osman1993].Thetwoprocessesareruniterativelyafinitenumberoftimestoimprovethesolutionquality.Thesearchbeginsbyclusteringcustomerseitheraccdingtotheirpolarcodinateangleromly.Theproposedsea

12、rchstrategyacceptsalsoinfeasibilitiesduringthesearchagainstcertainpenaltyfacts.IntheGIDEONsystemeachchromosomerepresentsasetofpossibleclusteringschemesthefitnessvaluesarebasedoncrespondingroutingcosts.Thecrossoveroperate

13、xchangesaromlyedptionofthebitstringbetweenthechromosomesmutationisusedwithaverylowprobabilitytoromlychangethebitvalues.[PotvinBengio1996]proposeageicalgithmGENEROUSthatdirectlyappliesgeicoperatstosolutionsthusavoidingthe

14、codingissues.Theinitialpopulationiscreatedwithcheapestionheuristicof[Solomon1987]thefitnessvaluesoftheproposedapproacharebasedonthenumberofvehiclestotalroutetime.Theionprocessisstochasticbiasedtowardthebestsolutions.Fthi

15、spurposealinearrankingschemeisused.Duringtherecombinationphasetwoparentsolutionsaremergedintoasingleonesoastoguaranteethefeasibilityofthenewsolution.Twotypesofcrossoveroperatsareusedtomodifyaromlyedroutetoarouteintotheot

16、herparentsolution.Aspecialrepairoperatisthenappliedtotheoffspringtogenerateanewfeasiblesolution.Mutationoperatsareaimedatreducingthenumberofroutes.Finallyindertolocallyoptimizethesolutionmutationoperatbasedonoptexchanges

17、[1976]isincluded.[Bergeretal.1998]proposeamethodbasedonthehybridizationofageicalgithmwithwellknownconstructionheuristics.Theauthsomitthecodingissuesrepresentasolutionbyasetoffeasibleroutes.Theinitialpopulationiscreatedwi