案例賽獲獎(jiǎng)年44348f

TeamTeamControlofficeuseofficeuse Problem 數(shù)學(xué)建模資料，請(qǐng)搜索淘 ：學(xué)神資料 2016MCM/ICMSummaryTowardsAHopefulTheworldiswitnessingthelargestrefugeecrisissincethehorrorsofWorldWarII.Modelingrefugeeimmigration,whichiscrucialtotackletheproblem,isanintricateissuethatshouldembracethesophisticationofsocialinterrelatedsystems,andtakeintotheconsiderationofrefugeecrisisonlocalconditions.Concerningthestructureofrefugeecrisisandtheroutesofmigration,withtheofficialdatain2014,weconstructarefugeemigrationmodelandbuildafeedbacksystemusingnetworkysismethodologyandCellularautomatontomakeprecisesimulationaimingtohelpfigureoutasetofefficientInthefirstce,weestablishasetofmetricstoconsiderthedeterminantfactorsinrefugeemigrationssothatwedefineourmeasuresandindexes,afterwhichwesetthestartpointsofsixgivenroutesassixnodes,andchoose14countrieswheremostrefugeesgathertobethenodesinAfricaandCentralEast.WiththeassumptionthattherefugeesmigratenearerandnearertoEurope,wedividethenodesinto4layersbasedonthedistancesfromnodetonodeusingclusterysis.Inthatway,therefugeesmigratewithintheAfterthat,weassumethatrefugeesgetlimitedinformationandinitiallybuildarandommigrationmodeltodeterminethemigratingfactorsbetween2nodes.AndbysimulationwegetresultindicatingthemainroutesreachingEurope,butthenumericaldataisinconsistentwithrealdata.Hence,weadjustourpresumptionsandreviseourmodel.xt,inspiredbyGravityModel,weyzethefactorsthataffectthemigrationofrefugeesandintegrateemintoacomprehensiveattractionindex.Bycollectingandcalculatingthestatistics,wefigureouttheeightbetweentwonodesandtheratiosofpopulationdistributionatthesixstart.07:0.101:0.41:0.369:0.05),whosecorrelationcoefficientwithrealdataR=0.98.Sowegoonwithrevisedodeltounraveltheoptimalflowdistributionunderdifferentconditionsandthemeasurementofweightomnodetonodebackedbysingle-periodsimulation.Inaddition,weexpandthescaleofnodes.Giventhatthefeedbackinformationofrefugeesindifferentperiodsandthe umcapacityineachnode,wesimulatethemigrationprogressofrefugeeswithCellularAutomatonandC++,andtheratiosof3nodesare0.0362:0.537:0.427,R=1.Therefore,wegettoknowabouttheinfluencesofernmentandnon-ernmentorganizationsonrefugeemigration.Asforthe,weattachsignificantimportancetotheempathythatthereceivingpopulationofeachcountrymustfitthepresentrefugeecondition.Byscalingupourmodel,wefindthattheroutesgetsaturatedexceptforroutesinNorthAfricaandCentralEast,whichmaytriggerthedetentionofrefugeesandeventuallyleadtoillegalimmigration.Meanwhile,weworkouttherelationsbetweenthestabilityandcapacityofnodesbybuildingaCobwebModel.Wefindthattherefugeeflowstendtobemoreandmorestableinnodeswithbiggercapacity.Sowealsoproposetostabilizetheflows,whichcontributestobetterresourceallocationandaidfromGOandNGO.Finally,wetestthesensitivityofourmodelandconcludethestrengthsandweakness.Themodelisquitereliableinsmallscalebutstillneedsadvancementforlargerandmoreprecisesimulation.TeamTeampagePAGE1of Our Therelationshipamongthe6 ysisoneach Anoptimal TheStrategywithCobweb Theoutbreakofthe Terrorist HowcanourberesilienttoExogenous societies.Oneinevery122peopleintheworldiscurrentlyeitherarefugee,internallydiscedorseekingasyluminEuropedisproportionay,aswar,persecutionandpovertycontinuetodrivepeopleawayfromtheirhomes[2].FromthechartswecanseethattheasylumapplicantsinEuropeancountriesareincreasingstrikinglyoverthepast5years,leavingpressingconcern.trekthroughMacedonia,Serbia,CroatiaandHungaryaswellastakingperilousvoyagesovertheMediterraneanandAegeanseasintheirdesperationtofleeravageandreachsafety.,thusthetruenumberofarrivalsisalotsmallerthanthatoforiginals.AccordingtotheIOM(theInternationalOrganizationforMigration),morethan3,695migrantsarereportedtohavediedtryingtomakethecrossingin2015[4].Besides,amongthosewhomakewaytotheirdestinations,onlyascarceproportionofrequestsaregranted(32%in2014)[5]Ithasbecomestrikinglycrucialtodealwellwiththesophisticatedrefugeemigration,whichisasocialprocesswellstructuredandorganizedasasocio-spatialsystem[6,7].OurWefirstelaboratereasonablemetricsonvariousfactorsthataffectthemovementofrefugeeswithstatisticsandgraphsfromFRANandUNHCROnlinedatabase,thenwedesignourownevaluationsystemtomeasurethefactorsthathavebeentakenintoconsiderationusingclusteringmethodology.isvalidatedbydatafromEurostatMediaand2014StatisticalYearbook.onthescalability.SafetySafetycomesthefirstintermsofmigration.TherecentsurgeofpopularinterestinandincreasingpublicawarenessofmigrantdeathsintheMediterraneanhasturnedthequestionofroutine-relateddeathsintorareinthepastbuthavebeenincreasinglyheardinthelastfewyears.Thisriseappearstohavecoincidedwiththegrowthintraffickingandsmuggling,sotoothewideruseofguns.Theincreasedinstabilityonthewaytohavencountriesmakemanymigrantspanickedandthreatened.Ontheotherside,thewarconditionin

(fromceitocej)iscalculatedbyourmodelintheTheenvironmentalacceptanceCijisundoubtedlysignificantasrefugeeswhomweassumetobesensiblewillcomparetheenvironmentofthedestinationwiththatoftheorigindistricti.Here,theenvironmentacceptanceisnotmerelyreferredtolivingenvironmentorumcapacityincountryjbutalsoincludethreemainhealth-care,referralsystems,specializedhealthservices,psychosocial-medicalunitsandchildhealthsupportareallinneedtocopewithwoundandtraumainadesireddestination.viveon.Educationandcommunicationssystems,commercialinstallationsareneededforthefunctioningofacommunity.Manyindicatorssuchassocialwell-beingofpeople,jobopportunityandsoonareconsidered.AseconomistAmartyaSenpointsout,”economicgrowthisoneaspectoftheprocessofeconomicdevelopment.”Inourmodel,wetakeGDPasthemainindicatorofeconomiccondition.languages,customs,ceremoniesetc.varyfromcetoce,whichmightbeaburdenforcommunicationbetweenmigrantsandlocalpeople,thusculturalacceptancelactatingmothers,theelderly,disabledandpeoplelookforaceofenvironmentalacceptance,whichhastoremaininunrest.Totransportfromceitocej,oneneedstobalanceacomprehensivetransportationcostwithregardoftime,distanceandtrafficconditionrelatedtothetypeoftransportationetc..ThusatransportationindexConviisdefined.Thetyofrefugeeswhogetgrantedislimitedinanentrypoint,thustheproportionofrepatriatesorrejecteesarerelativelytoohigh,thenrefugeesarelesslikelytotakeonaroutineheadforthatnode.Therefore,wealsogiveanotationofBtrepresentingtheratioofrepatriationinacertainperiodTinordertobettersolvetherefugeecrisisproblem.tyofiWepremisethatthetyofrefugeesallowedtopassthroughanentrypointisaconstant,inthisway,thetotalnumberofpeoplewhogetgrantedisdeterminedbythenumberofentrypoints(nodeinmodel).Pointjistheithentrypointincej.Evidently,thelargerthenumberi,thelargerpassprobabilitywillbeforacertaingroup,whichaffectsthechoiceofrefugees.iTakingallthoseparametersandmeasuresintoaccount,wecanstepfurthertodefineasetofnotationsindesigningamathematicalmodel.PPProbabilitytomigratefromnodeitonodejofrefugeesinyearRatioofregionalacceptanceofgroupi cejinyeariRoutinechosenbygroupiinyearIndexofsafetydead cei cejinyearRatioofrefugees ceiinyeariIndexoftheresourcegiventogroupiinyearNumberofEntrypointsce ceice cei cejinperiodNTi ceiinperiodRatioofrefugeessentbackinperiodOffiiP ceicompared ceceiceiceceRatioofofficialacceptancecej(mainlycustoms)forgroupNo.ientrypointceMT cejtoattractgroupiinperiod

page4ofDistanceonland ceiceDistanceontheseaceicececeRefugeesmigrateingroups,andeachgroupisaunitwithfixednumberofindividuals,whoseoftentakeonaroutinechosenbyaleaderwithinagroup,andmigratetogetherwithfamilyorcommunity;besides,themigrationofrefugeesisgenerallyregional.Thusitispragmatictomakesuchanassumption.Themigrationofrefugeesisirreversibleandtherewillbenoidealshieldbeforearrivingtheirdestinations.Thatistosay,leaveouttheconditionthataunitgobackfromwhereithasreachedto,andtakenoaccountofthosewhostayhalfway.Refugeesaresensiblewhocanadjusttheirroutinesaswellastheweightvalueaccordingtodifferentconditionsatcurrentnode.Refugeesareabletojudgereasonablyonverifiedroutineswithchangingfactorsanddecideontheprobabilitybasedontheweightvalueofcertainroutine.Themigrationofrefugeesisprogressedstepbystep.Notalltherefugeeslocateoncoastlines,thustheroutinesheadforcoastaresubroutinesofmain6routines.Themortalityisaconstantonlandandanotheronthesea.Onthethreeroutinesacrossthewest,centralandeastMediterraneanOcean,therefugeesmigrateontheseawhileonotherroutinesonland,andthetwotransportdifferwitheachotherinriskcoefficient.ThosewhogettheirimmigrationgrantedarenolongerrefugeeswhilethosewhoenterEuropebutaresentbackforalackofentrancepermitarenewrefugees.Migrationoccursquarterly.FRONTEX-FRANQuarterlyprovidesaregularoverviewofirregularmi-grationattheEUexternalborders,basedontheirregularmigrationdataexchangedamongMemberStateTtopictureoutanirregularmigrationatEUlevelandtodiscussonthetrendsandpatterns.Thetyofpassedrefugeesareconstantpernodeperperiod.Thenodesvaryfrom cetoce,butthecapacityforrefugeesofeachnodeissamesothatthetotalcapacityforrefugeesin ceonlydecidedbythetyofTeamTeampagePAGE10ofThetyofentrypoints,distanceonlandandonthesea,thetimeneededtomigratewithofofficialacceptanceandthatofenvironmentalacceptanceallremainasconstants.Thisisbasisofourmodel,theparameterslistedabovearenotsensitivetotime,ifnotso,theproblemwouldpossiblybeadeadend.Giventhosepreconditions,wecansetouttoconstructourmodelsoastofigureoutasafeandefficientroutinefortherefugees.NodedefinitionAn shareogoustrait,andwhosetraitisassumedtobethesameregardlessofthenuancewithinthearea.Andthemetricsandassumptionsapplytothosenodes.andnation,etc.The6routinesthatrefugeestakelocatebetweennodeandnode.Eachnodehastwostates,whicharesaturatedandunsaturated.Thedistinctionbetweenthetwostatesisthatthesaturatednodeshavepassedthetippoint(threshold)andcannotreceiveanymorerefugees.Peoplecannotstayinsaturatednodes.Theseissueshavebeentakenintoconsiderationinthenodalweightysis.selectedandrankedthetop13countriesinAfricaandMiddleEastwheremostrefugeesstarttheirmigration,asisshowninFigure2,andwechoosethemasrepresentingnodestomakeweightysis.Twothingsshouldbenotedthattheresourcestatisticswechooseareallfrom2014.Foronething,mostofthedatain2015isincompleteandforanother,thelatestcompleteofficialstatisticsisuptofirsthalf2015,sowedecidetouse2014data,whichhasbeenverifiedofficiallybyIOM,UNHCRandEurostatetc..AnotheristhatasUgandaandKenya,PakistanandAfghanistan,LebanonandJordanareveryclosetoeachotheronInotherwords,wehavealtogether11nodesinourinitialmodel,asismarkedoutinFigure3.MigratoryFromAlgeriaandMoroccotoSpain,FranceandFromTunisiaandLibyatoItalyandFromTurkeytotheEuropeanUnionviaGreece,southernBulgariaorCircularRouteFromAlbaniatoGiventhesixmainmigratingroutesmappedoutonBBCwebsite,weassumethoseroutesarethefinalroutesthatthe11nodeseventuallyheadfor.AndwechoosewesternMediterranean,centralMediterranean,EasternMediterranean,EastBalkan,Albania,andborderlandbetweenLithuaniaandPolandasthestartingpointsofthose6routes,asisshowninFigure3. Figure2:Top13Refugee Figure3:RoutesandWepresumethattherefugeesinAfricatendtofleepersecution,soastosay,withintheemigrationcentersinacountry,therefugeesalwaysheadfortheregionsorcenterswithinacircleareaclosertoEurope.WecollectedroutedistancedatafromFRANandICMPD(seeinfigure4),thenwedividethe11nodesinto3layersbasedonclusteringmethodologyprinciples,inwhichtheinitiallayerincludes3nodes(Pak-andSouthSudan);theterminallayerincludes4nodes(SouthSudan,Lebanon/Jordan,ChadandIraq).Refugeesmigratelayerbylayer,eventuallytheygettothe6startpointsofthemainroutesreachingtobutitshouldbenotedthatasEBrouteisfarawayfromthelocationofrefugeesinourmodel,anditisverydifficultforthemtotakeonEBrouteandenterEurope,soweleaveoutEBrouteandit’snodesaswell.Theformulais:5d=1Xx,i=1,2,5 Thenwemakeamatrixwithallthe

D 4630406034702840288035203222141023702333 Theunitsareintegratedintoawholecluster,whoseheightvalueareallzero,that Hi 4630406034702840288035203222141023702333 SetheightvalueasK1,andpickouttheunits,formingthefirst H1 46304060 SetheightvalueasK2,andpickouttheunits,formingthesecond H2 284028803520 SetheightvalueasK3,andpickouttheunits,formingthethird H3 141023702333Thusthe11nodesareclassifiedinto3Figure4:NetworkofSupposethatinaperiodcircleTofanode,theoriginaltyofrefugeesstartingoutfromthatnodeNTN

,andtheformulacomes

jNT=Nj

Xi NXiThatis,thetotalrefugeesstartingoutfromanodeistheoriginaltyplusthetyofrefugeesfromtheformerlayer.Basedonthenetworkofnodes,wesimulatetherefugeeflowsdirectlyusingwiththeassumptionthattheweightvaluesofmigratingtowardsthenodesinthenextlayerareallthesameforrefugeeslocatedinthenodeswithinthefirstlayer.Besides,theatthenodeswithinthelastlayertherefugeesonlychoosethreeroutineswherethetransportationanddistanceareoptimumsinsteadoftakingonalongandtiringjourney.Thestatisticsareshownbelow.Figure5:Distancesfromthenodesto6startFigure6:Initial R 1.0000 0.6543 P 1.00000.1586AfteryzeRandP.Wefindthatp>0.05andR=0.6543.Theresultofinitialmodelisinconsistentwithourdata,soweadjustedtheroutesusingweightedapproach,thenweyzetheweightedvaluewhichisdemonstratedaspercentage.Thenearerto1thepercentageis,thebettertheresultis.Therefore,ourinitialmodeldoesn’tworkwell,andwerightawaymodifyit.WepresumethatthetherefugeesontheroutetoGreeceincludethosewhohavealreadyenteredGreece.AccordingtorealdataonIOMandEurostat,thetotalrefugeesonroutetoGreeceshouldhaveaddedwiththoseonotherrouteswhichcombineintothemainGreeceroute.R5=a1R1+a2R2+a3R3+Fromthedatabefore,wecanestimatethenumberofrefugeesflowingintoWB1GravitymodeloriginatesfromtheGravityTheory,indicatingtheattractionbetweentworelatedsubjects.Thebasicformulais:

=PiPj,(a>DaDaIistheinctionindicator,PiPjistheproductofpopulationsoftwocities,representingthemassproductoftwocities,Disthedistancebetweentwocities.Thismodeliswidelyappliedtomeasuretheeconomicandfinancialattraction.I.S.Lowryalsodevelopedaregressionmodel(Lowery,1996)whichshowsthatjobopportunityattractspeopletomigratefromonecetoanother.Nowwearetomeasuretheweightforthenodesinspiredbythosemodelstohelprefugeesmakebetterdecisionontopeople.Tocalculatetheextentofimmigrationbetweenthetwoareas,weapplythemetricsdevelopedabove,Wetakethelogarithmonbothsidesoftheequation,herecomestheexpression:lnMij=lnPi+lnPj?lnDa+Ka,Thedistancedatafromlayertolayerindifferentnodesisillustratedasbelow,towhichwerefertocalculateFigure7:DistanceLayer1toLayer2 Figure8:DistanceLayer2toLayer3NowaddthevariablesderivedfromthemetricsandnotatedabovetoMij,wegetMt=f(St,Convt,Bt,Cij, St:Safety

(St=1(

+1x 1routecrosses10routedoesn’tcross(2x 1routecrosses(20routedoesn’tcross*c1isdangercoefficient,normallyrefersmortalityonlandroutesandc2isonMortalityWithFRANandIOM(FatalJourneysTrackingLiveslostinMigration,2014),wecollectthedeathsdataon6routesofrefugees,herewegivethemortalityonthesea,whichisc2Figure9:MortalityontheAndc1isfiguredoutinasame

=*c3isadjustingiBt:repatriationi

Trij

threekindsofiThisindicatorisdeterminedbytheratiobetweentherepatriatedrefugeesandtotalmigratingrefugeesofacountryinacertainyeart,thelargertheratiois,thelagerBtis.ii iiPjEntrypointi

Cijt=c4Re(refugeesofeachnodeisaconstant.Onceanodepassesthetippoint(threshold),itreachesasaturatedstategrantedmigrantsisthesameforeachnode.SettheumasM,thentheprobabilityforagrouptobereceivedcanbesimplifiedasaBernoulliDistribution:(P pn>0n<Ifniunitsofrefugeespassthroughthenode,P(x=ni)

ni<alnMt=lnSt+lnConvt?lnBt+a*Mijistheweightindexforarefugeegroup’sattractiontoajiNT=N +XNT(1?MoTjiFigure10:RevisednetworkofAgainweusecalculationtoseehowtherefugeescanmigratefromnodetonode,layerbylayer.AndFigure11and12isthepropagationofrefugeesfromnodetonode,theyaxesstandsfortheproportionofpopulationthatonelayerreceivefromtheformerlayer.Figure11:PropagationL1toL2 Figure12:PropagationL2toL3AndFigure13isthepropagationofrefugeesfromthethirdlayertothestartpointsheadforEurope.Figure13:TheratioofrefugeesmigratingfromthethirdlayertothestartComparingthetwogroupsofdata,wefindthattheCorrelationRequalsto0.9914andPequalsto0.00001<0.05.Sowethinkthisgroupofdataisvalid.wefoundthatthistimethemodelcopeswellwithrealdata,showninFigure14.Theresultisdemonstratedbelow,line1istherealdata,line2isrevisedmodelingresult,andline3istheinitialresult.Theylabelisratiobetweenrefugeesatanodeandtotalnumber,xlabelis6routes.Evidently,line1andline2coincidewitheachotherwell.Sowethinkthattheimprovedgravitymodelcancalculatetheweightbetweenthenodesmoreclosetotherealcondition.Andnextwesearchforanoptimalrouteofmigration.Figure14:TheresultsandAnoptimalAimedreducethecostinthethirdlayer,andinthiscasewecalculatetheweightofroutes.WeightoftheimportanceofeachaimInmulti-objectivenning,wesetthesafetyindexas?1,anddistanceindexas?2,so?2(1?St)=tative Thenwemaketativeysisoftheattitudesofreceivingcountriesaccordingtothe 1.52.05.04.75Andthedistanceofeachroute0000000000000000000000Finally,allthefactorsrelatedtoroutenningcanbeseen

η= Withtheysisandourrevisedmodelwecanchoosetheweightoflastfournodesinthethirdlayertoconfirmthebestroute.(1)?1=?2=(2)?1=0.4,?2=(3)?1=0.6,?2=(4)?1=0.3,?2=(5)?1=0.7,?2=Iftakingsafetyfirst,andmeetingthepoliciesofeverycountry,thenchoosethelargest?1=0.7andwefindoutthebestdistributionofrefugeesoneachrouteasFigure15Figure15:?1=0.7,?2=0.3theratiosofrefugeeson6Figure16:?1=0.3,?2=0.7theratiosofrefugeeson6(Line1isrealdataandline2isthemodelingresult.Wefindthatinthiscase,theratiosofrefugeesonthesixroutesare0.20,0.17,0.30,0.31,0.01,0.01,indicatingthatwhatevertheconditionis,therewillstillbehugenumberofpeopletravellingthroughthisroutetotheirfinalnIntheverybeginningwehaveassumedthattheflowofrefugeesoccursperiodically.Sosupposethattherearenperiodseveryyear,andthemigrationflowisacommunityofcertainnumberofrefugees.Thuseachquarterthemigrationpopulationis1%,andthemigratingpeopleinfluencetheremainingones.nThemodelisshownasFigure17:DynamicAsisseeninthefigure,node1,2,3,4representstheareawhererefugeesgatherandstartmigrationinAfricaandCentralEast,andwecangettheweightbetweendifferentnodeswithourmodel.SupposethatduringthefirstLinaperiod.Weaddacoefficientδ=L

Therefore,inthefirstperiodthemigrationpopulationatthefirstnodeis1Nt0,andthemigrationntowardsotherthreenodesis1Nt0w1i,(i=1,2,3),nowtherefugeepopulationfromthefirstnodetowardsnk2nNiwi2+4Nt1k2nNiwi2+4Sointhesecondperiodwecanestimatetherefugeepopulationatthesecondnode.Butduetothechangingfactorsandindexes,theweightmatchangefromnode1tonode2asthepopulation,capacityofrefugeesandtheaverageavailableresourcemaychange.Wedividethosechangingfactorsintotwocategories:Non-paroxysmalChangeincapacityand2Evidently,theresourcesandcapabilityofanodeislimited,theoutsandinsofrefugeesaffecttheweightofothernodeswherethepeopleheadforthisnodeinnextperiod.Takenode1and2asanexample,thepopulationofnode2inperiodt0isNt0,andnowasthechangeofpopulationleadstothechangeofinflowweightofothernodes,wesetthecomparativecoefficientasπ,theexpressionis:2NNtπNt02Sotheweightafterchange

w0=πWWt1= i2

i2nSupposethattherearesaturatedandunsaturatedstatesateachnode,andthesaturatedreceivingpopulationisNL,iftherefugeesaremorethanNLafter1ofthepopulationhavemigratedout,thenodecloses,nallowingout.That

ninninkLn?1Ntj=n?1(n?1Ntj?1+XninninkL

Ntj?1)>NInthisway,wecanforeseethechangeofpopulationatanodeinthenextperiodwhenitisneartipBasedontheformula,wesimulatetheprocessofmigrationwithC++onlywithnon-paroxysmalfactors,andwemaptheprocessoutwithJavaScript,fromlefttoright,uptodown,wecanseethechangingpathsofrefugeesfromAfricaandCentralEasttoEuropewithinsamespaceoftimeandthechangeinpopulation,whichisindicatedbychangeofsizeofthecirclearea.WesupposethatthefinaldestinationofEMandCMrouteisItaly,theoneofWMrouteisSpainandtheoneofWBisGreeceandotherEuropeancountries.Bycertainalgorithmanddatacomparisonwegettheratiosas(0.00362,0.537,0.427),andtherealdatais(0.013,0.545,0.422),throughcomputercalculationwefindthecorrelationcoefficientvalueisR=1.0000(Environmental,militaryfactorsallmakesense.Thechangeinsafetyindexcanbeadjustedbyaddingwarfactorβandenvironmentalfactorγ(

0nowaronthewarontheInthiscase,thesafetyindexis1?St=(x2c1+cx+

Hereweneglecttheintensityandscaleofthewar,andc4isrelatedtothemortalityinwar.Thedefinitionγij

fair Thisalsoaffectsthesafety 1?

=(x2c1+xc+βcTheinfluenceofGOand

1 ijThefundsandpoliciesofernmentscanaffectthecapacityandresourcesofanode.NGOalsoprovideshumanistichelpandmedical-healthaid.iAsfornodei,theinfluenceofGOpoliciesismeasuredwhenNprefugeesmigrateinoroutandthereareinitiallyNtjrefugeeswhomeettheconditionthat:in?1(Ntj+N)>N Iftheconditionisnotmet,thepopulationmigratingintoanodeimpactsotherNtj=Ntj+Thusthecomparativecoefficientis:π

NNNNiAsforNGOandGOinvolvementwhichmightmitigatetherefugeecrisis,reducingtherefugeepopulationwithNNGOunchanged,sotheweightofnodeschangeaccordingly.Andtheactualpopulationateachnodeisdefinedas:Ntj=Ntj? Otheralgorithmsaresamealike,wealsoassumethattheinfluenceoncapacityofrefugeesfromlocal-NFigure19:SimulationofRefugeeMigrationonlywithaccidentalfactorsTheleftfigure:?N=20%,βij=0,γij=cNNfine,thenthemigrationflowwillchangeby15.9%,indicatingthatthechangeincapacitycanchangethemigrationinflows.Therightfigure:?N=?20%,βij=1,γij=0NTheratioofthreenodesare(0.0374,0.562,0.4006),sowhenthecapacityofGreecedecreasesby20%,andthereareontheroutes,themigrationflowwillchangeby6.18%,theflowsstillincreasebecausethereareothermigrantscomingfromFigure20:SimulationofRefugeeMigrationonaLargerAccordingtotheresult,wefindthattherewillbesomanyofrefugeesfromCentralEastflowingintowhilefewheadingSowefindthatourmodelcannotillustratetheflowoftherefugeeswell.TheHimalayasBecauseoftheHimalayasmountainsstandingattheborderof,itwillbemuchharderfortherefugeestocomeinto.Sowhenweighingthevalueofthesenodes,weshouldtakethetopographyintoTheStrategywithCobwebEnlargethecapacityandincreasetheresourcesinreceivingWeexpandthecrisistoalargescale(byafactorof10),andfindthatthatourmodelcan’tpredicttheflowoftheactualmodeling,butinvain.Manyrefugeesmaystayaroundforthelimitedresources,leadingextrarefugeeinflow.Sowhenthescaleexpandsup,therestrictingfactorswehavelistedcanbeuseless,asthenodesalwaysstaysaturated,resultingindetentionandtheEnlargethecapacityaccordingtocurrentsituationtostabilizethepopulationofmigrationModel,wefindthatchangeofcapacityinaperiodwillmakethepopulationchangeateachnode.AlargecapacityofanodewillactuallymaketheflowofrefugeesthroughthenodestablewhilearelativelysmallcapacityofthenodemightFigure21:CobwebModel1Figure22:CobwebModel2NP=AFigure21andFigure22presentstherelationshipbetweenthepassratePandthepassrefugeepopulationWhensimulatingthemigrationprogresswithourmodel,wefindthatsomeofthenodesmaybecomesaturatedwhileothernot,soenhancinginternationalcooperationandpartnership,distributingrefugeesafterreceivingthemmaybebetweenthenodesinvolvedinthedisease.Asaresult,therefugeeswillnotonlyconsiderdistance,theeconomicconditionsandtheattitudetowardsrefugeesofitsneighbouringcountriesaswellastheinter-r