船公司背景投资人的集装箱码头项目决策模型构建及应用

上海交通大学硕士学位论文船公司背景投资人的集装箱码头项目决策模型构建及应用姓名：秦翡申请学位级别：硕士专业：项目管理指导教师：苏强;方萌20081001IBPVisualBasicAccessIIBPIIIFORMATION&APPLICATIONOFACONTAINERTERMINALPROJECTS’DECISION-MAKINGMODELFORINVESTORSWITHSHIPPINGCOMPANY’SBACKGROUNDABSTRACTWiththeupgradingdevelopmentoftheglobaleconomy,thedevelopmentofthecontainertransportationhasbecomeasanimportantindicatortothedevelopmentofdifferentcountries’economies.Asacruciallinkinthechainofthecontainertransportation,thecontainerterminalhascometobethefocusofdifferentcountriestoinvestanddevelop.Theinvestmentdoorstothisindustryhavebeenbroadlyopenedwhichalluresmanyinvestorsactivelyenterintothisindustry,formerlymonopolizedbylocalauthoritieswithsteadyreturns.Amongthem,theinvestorswiththebackgroundofshippingcompaniesaregrowingtobeamainforceintheindustry.Thisthesisfocusesonthestudyofthefactorsthataffecttheinvestmentdecision-makingtotheprojectofthecontainerterminalfromthestandoftheinvestorswithbackgroundofshippingcompaniesandafterthensetsupamodeltoassisttheinvestmentdecision-makingBasedontheanalysisandstudyonthebackgroundofthecontainerterminalindustry,andtheanalysisandstudyontheparticularpurposesofthecontainerterminalinvestorswithbackgroundofshippingcompanies,thethesishasfoundouttheevaluatorsinthreecategories,whicharethesurroundings,theeconomy,andtheindustry-supporting.Underthethreecategories,therearesevenmainevaluators,whicharethecontainervolumeABSTRACTIVoftheport,theeconomyofthehinterland,theIRR,theprofitratiotothecapital,theaccumulatedearningbeforetax,thesupportingratiototherelevantindustriesinthegroup,theparticipatingratiooftheinvestor,thatmainlyaffectthedecision-makingtotheprojectofcontainerterminalfromthestandoftheinvestorswithbackgroundofshippingcompanies.Afterthen,withtheapplicationofthetheoryofBPneuralnetworkinthesystemofartificialneuralnetwork,fromthesesevenmainevaluators,thethesishassetuptheframeofadecision-makingmodelofcontainerterminalsfortheinvestorswithbackgroundofshippingcompanies.Basedonthisframe,withtheutilizationoftwosoftwareplatforms,VisualBasicandAccess,aprogramwithfriendlyinterfaceandeasyoperationhasbeencompiledout.Thethesisusesthedataofthesamplecontainerterminalprojectsfromaterminalinvestmentcompanywiththebackgroundofshippingcompaniesasthedatabasetothemodelandputthedatabaseintothemodelforprimarytrainingandvalidatingthemodel.Aftertimesoftrialcomputationandcomparisononthethreeparameters,thenumberofthenodesofthehiddenlayer,themomentumfactorandthelearningrate,thatimpacttherunningeffectofthemodel,thethesishaschosenouttherespectivelyappropriatevalueofthesethreeparameterswhichmakestherelativelylesserrorsfromthenetworkofthemodelandrelativelyfewertimesofcalculation,andthenarelativelyoptimizedstructureofthemodelcomesintoform.Aftertrainingandvalidatingtherelativelyoptimizeddecision-makingmodelbythedataofthesampleprojects,theweighsoftherespectivelayer’snodesinthemodelhavebeendecidedandarelativelystableandusabledecision-makingmodelfortheinvestmentofcontainerterminalprojectscomesout,whichcouldthenbeappliedasanassistanttotheinvestorswithbackgroundofshippingcompaniesfordecision-makingofsomeothercontainerterminalsinfuture.Finally,thethesisteststheapplicationofthedecision-makingmodelbyanothersampleprojectfromtheterminalinvestmentcompanywiththebackgroundofshippingcompanies.ThesampleprojectfortestingisjustABSTRACTVunderconsiderationofinvestmentdecision-making.KEYWORDS:investorswithbackgroundofshippingcompanies,containerterminals,investmentdecision-making,BPneuralnetworkABSTRACT2311.11.1.121.1.2,NPVIRRAHP3BP1.2:BPBPBPVisualBasicAccess42.12.1.1-1613000TEU3502210000TEU10000TEU52.1.2135%1518202.1.310152062.1.4GDPGDP25562010322010GDP8-10%15%2010GDP2211.2TEUGDP2.1.5199948.2339.8-52007200415.58172.22.2.12005200710.48%AP82.2.2AP2001ICTSI5050%25252000453%2004AP802.2.39HPHPSADPWSSAAP2.2.42005774335%16%10200714.61%,36.71%12005TEU1518013%2A.P.404010.1%32403010.1%4336709.2%5426103.7%612103%78702.2%2047051%Drewry22007(TEU)1663013.62%2589012.10%343308.89%439808.17%5A.P.31406.45%67201.48%73000.62%2499051.32%Drewry,11A.P123.1131Fig1Thesystemoftheevaluatorsthataffectthedecision-makingofcontainerterminalprojectsinvestment3.2143.2.13.2.23153GDPGDPTEU1986102758.85%639.41%19871205911.58%6910.01%19881504311.28%9537.37%1989169924.06%11723.59%1990186683.84%15633.57%1991217819.18%21738.97%19922692314.24%27727.65%19933533413.96%38037.13%19944819813.08%50733.34%19956079410.92%66430.90%19967117710.01%80321.02%1997789739.30%107734.03%1998844027.83%131622.23%1999896777.62%173331.70%2000992158.43%226330.61%20011096558.30%266617.77%20021203339.08%372135.40%200313582310.03%486730.80%200415987810.09%616026.60%200518386810.43%756422.80%200621087111.09%936123.80%200724661911.40%1140022.30%16GDP3.33.3.1IRR∑=−=+−ntttIRRCOCI00)1()(3.1CICOtCOCI)(−t17201515151515153.3.2,,:%100×=3.2,1518153.3.3−−=3.353.4193.4.1123.4.2100%20,214.1ArtificialNeuralNetworksANNsNNsConnectionistModelNaturalNeuralNetwork4.1.1,222Fig2Thesketchmapoftheartificialneuron(perceptron)2jix(i=1,2,…,m)jy=−=∑=)(1jjjmiiijjsfyxwsθ4.1jθijwij()xfS∑==miiijjxws04.2jw0