1、碩士學(xué)位論文港口集裝箱吞吐量預(yù)測模型研究ResearchonThePredictionModelofSeaportContainerThroughput作者姓名：翟益盔學(xué)科、專業(yè)：籃理型堂皇王理學(xué)號：2Q31lQ鯉指導(dǎo)教師：韭置氈麴援完成日期：2QQ：12大連理工大學(xué)DalianUniversityofTechnology港口集裝箱吞吐量預(yù)測模型研究ResearchonThePredictionModelofSeaportContain

2、erThroughputAbstractTheseaportcontainerthroughputisinfluencedbymanyfactorssuchasmarket，economyandnon—economyfackorsSoitisacomplicatednonlinearsystemMeanwhilethereissomeinnerruleoverthethroughputvariationinthesystemChina’

3、Sforeigntradeandcontainertransportationiscontinuingthetrendofrapidgrowthalongwiththehighdevelopmentofcountry’SeconomyThegreytheorycanbewellusedtoforecasttheincreasedtimeseriesbecauseofitsowncharacteristicsHowever，sometim

4、es，theforecastingresultofgreytheoryisnotsatisfiedSo，inthepaper，theArtificialNeuralNetworkisusedtoamendedthemodelofGM(1，1)inorderteachthehighforecastingprecisionInthepaper，theforecastmodelofremainingamendedGM(1，1)basedonA

5、rtificalNeuralNetworkisproposedbytheanalysisofthecomponmentofseaportcontainerthroughput，andthenthemodelisusedtoforecasttheseaportcontainerthroughputSOastoconfirmthevalidityoftheforecastmodelFirstly，thepapergivesacomplete

6、introductionoftherecentforecastingmethodsandthestudiesnowadaysofcontainerthroughputhomeandabroadInaddition，itanalyzestheelementsthatinfluencethecontainerthrol】【ghputandpointsouttheseaportcontainerthoughputthatisanonlinea

7、rsystemSecondly，inthepaper，itstatesthecharacteristicsofthemodelofGM(I，1)andArtificialNeuralNetworkConsquently，thealgorithmofremainingamendedGM(1，1)basedonBPArtificialNeuralNetworkisproposedbecauseofthedefectsofthemodelof

8、GM(1，1)Thirdly，theforecastmodelofremainingamendedGM—NN(1，1)isproposedonthebasedofnonlinearthimerseriesforecastingthroryandthealgorithmofremainingamendedGM(1，1)basedonArtificialNeuralNetworkThentheseaportcontainerthroughp

9、utforecastingsystemisdesignedanddevelopedMeanwhileitdiscussestheparamentsoftheforecastmodelFinally，inthepaper，itconfirmstheparamentsoftheforecastmodelThenthevalidityoftheforecastmodelistestedbytheDalianportcontainerthrou

10、ghputforecastexperimentTheDalianportcontainerthroughputsin20062010arealsoforecastedbyusingtheforecastingsystemAndthenitgivesabriefanalysisupontheforecastingresultsKeyWords：TheModelofGM(1，1)；ArtificialNeuralNetwork(ANN)；T