資源描述:
《基于客流激發(fā)能級模型的鐵路客流短期預(yù)測研究》由會員上傳分享,免費在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。
1、北京交通大學(xué)碩士學(xué)位論文基于客流激發(fā)能級模型的鐵路客流短期預(yù)測研究姓名:馬彥祥申請學(xué)位級別:碩士專業(yè):系統(tǒng)工程指導(dǎo)教師:劉軍20080501ABSTRA(了ABSTRACTRailwaypassengersflowforecastingisallimportantbasisofpassengertrafficorganizingandaneffectivewayforthepassengertrafficenterprisetograspthetrendsanddemandoftransportationmanetdevelopment.Thispaperfocusesonshort-term
2、passengervolumeforecastingproblembetweentwocities.Theobjectiveistoinvestigateappropriateforecastingmethodsinordertoguaranteetheprecision.Thispaperbeginstheresearchworkfromobservingthefluctuationcharacteristicsofshort·termpassengerrailwayvolumebetweenorigination—cityanddestination.city,andfindssomesp
3、ecialpropertiesofthevolumechange.Onthisbasis,thepapergivesaconclusionthatmethodsbasedoncombiningtimeseriesanalyzingandqualitativeapproachshouldbeusedinshort—termpassengervolumeforecasting.Short-termRailwayPassengerFlowForecastingmethodBasedonPassengerFlowExcitedLevelsModelisenlightenedbyBoer’Satomic
4、.modelinphysics.Themodelincludetwokeyparts,theoneisevaluationlevelsandtheincrementinexcitedstates.theotheroneistimeseriesforecastingingroundstates.Inthestudyofevaluationlevelsandincrementinexcitedstates。thefactorsofpassengerflowtransitionisanalyzedatfirst,thenFuzzycomprehensiveevaluationisappliedtoe
5、valuatethesefactors.Toovercometheshortageofhistoricaldata,theincrementoflearningsamplesalegotbyclusteringanalysisthetimeseriesdatafromTicketsalerecord.IIlthestudyoftimeseriesforecastingingroundstates,themethodforrecognitionandprocessingsingularvaluesisproposed,thenLS—SVMisappliedtoforecast.Inorderto
6、improveefficiencyandgeneralizationabilityofforecasting,geneticalgorithmisappliedtoselectparametersofLS.SVM.Finally,themethodisappliedtotheactualpassengerflowvolumedailydataofGuangzhoutoWuhan.Themodelandmethodsareeffectivebythetestofexample.KEYWORDS:railwaypassengertransport;short—termforecasting;Lea
7、stsquaressupportvectormachine;PassengerflowexcitedlevelsCLASSNO:學(xué)位論文版權(quán)使用授權(quán)書本學(xué)位論文作者完全了解北京交通大學(xué)有關(guān)保留、使用學(xué)位論文的規(guī)定。特授權(quán)北京交通大學(xué)可以將學(xué)位論文的全部或部分內(nèi)容編入有關(guān)數(shù)據(jù)庫進行檢索,并采用影印、縮印或掃描等復(fù)制手段保存、匯編以供查閱和借閱。同意學(xué)校向國家有關(guān)部門或機構(gòu)送交論文的復(fù)印件和磁盤。(保