土木工程外文翻譯

外文翻譯EstimatingFutureHighwayConstructionCostsC.G.Wilmot,M.ASCE,1andG.Cheng,P.E.2Abstract:TheobjectiveofthisresearchwastodevelopamodelthatestimatesfuturehighwayconstructioncostsinLouisiana.Themodeldescribesoverallhighwayconstructioncostintermsofahighwayconstructioncostindex.Theindexisacompositemeasureofthecostofconstructionlabor,materials,andequipment;thecharacteristicsofcontracts;andtheenvironmentinwhichcontractsarelet.Futureconstructioncostsaredescribedintermsofpredictedindexvaluesbasedonforecastsofthepriceofconstructionlabor,materials,andequipmentandtheexpectedcontractcharacteristicsandcontractenvironments.Thecontractcharacteristicsandcontractenvironmentsthatareunderthecontrolofhighwayagencyofficials,canbemanipulatedtoreflectfuturecost-cuttingpolicies.ApplicationofthemodelinforecastingtohighwayconstructioncostsinLouisianashowsthatthemodelcloselyreplicatespastconstructioncostsfortheperiod1984–1997.Whenappliedtoforecastingfuturehighwayconstructioncosts,themodelpredictsthathighwayconstructioncostsinLouisianawilldoublebetween1998and2015.Applyingcost-cuttingpoliciesandassuminginputcostsare20%lessthananticipated,themodelestimateshighwayconstructioncostswillincreaseby75%between1998and2015.Keywords:Highwayconstruction;Costs;Estimation.onStateDepartmentsofTransportationarerequiredtopreparehighwayconstructionprogramsthatdescribetheirplannedconstructionactivityintheshortterm.Thereisusuallyconsiderableinterestintheprogramfromlocalauthorities,politicians,andinterestgroups.Draftprogramsaretypicallypresentedtothepublicandtovariousagenciesatthelocal,regional,state,andfederallevelforcommentandreview.Ultimately,aprogramwillbeapprovedbythestatelegislatureandwillbecometheformalprogramofconstructionofthestateDepartmentofTransportationuntilanewprogramisdevelopedinthenextcycleafewyearslater.Becauseindividualprojectsareofconsiderableimportancetopoliticiansandindividualinterestgroups,itiscommonthatprogressonaconstructionprogramiscloselymonitored.Anydeviationislikelytobequeried,andtheSecretaryofthestateDepartmentofTransportationoraseniorofficialinthedepartmentwilloftenhavetodefendthesituationpubliclyorinthestatelegislature.Thiscanleadtoperceptionsofincompetenceanderosionofsupportfromthelegislatureandthepublic.Topreparereliablehighwayconstructionprograms,roadauthoritiesmusthaveaccurateestimatesoffuturefundingandprojectcosts.Whilefuturefundingisobviouslyneverknownwithagreatdealofcertainty,itisoftentheestimationofprojectcoststhatcauseupsetsintheexecutionofconstructionprograms.Inaccuratecostestimationisonesourceoferror,butanother,theescalationincostofaprojectovertime,isanothersourcedisruptiontotheprogramthatisusuallynotanticipatedandcateredfor.Typically,whenprojectsarecosted,theircostsareestimatedintermsofthecurrentcostoftheproject,andthisestimateisnotadjustedfortheyearinwhichtheprojectisscheduledforimplementation.Thesecostincreasescanbesignificantandare,ofcourse,cumulativeacrossprojects;also,theyriseatanincreasingrateeachyearintothefuture.Estimatingfuturehighwayconstructionisthefocusofthispaper.ThemodeldevelopedinthisstudywasdevelopedwithdatafromtheLouisianaDepartmentofTransportationandDevelopment~DOTD!andisthereforeparticulartothatstate.However,themethodologyemployedcouldbeemployedinotherareas.MeasuringProjectCostsWhenconstructioninthefieldlagsbehindplannedconstructionintheconstructionprogram,itisusuallybecausetheprojectsthathavebeenconstructedhavecostmorethananticipated.Thisisnotrandomvariationofactualcostsaboutestimatedcosts,because,clearly,underestimateswouldcanceloutoverestimatesovertimeinsuchasituation.Rather,itisevidenceofaattheoveralllevel,whichismucheasiertomeasurethanattheindividualprojectlevel.Inthepast,changeinoverallconstructioncostshasbeenmeasuredintermsofconstructionindices.Theseindicesareweightedaveragesofthecostofasetofrepresentativepayitemsovertime.Theyhavebeenusedtodisplaycosttrendsinthepast.However,thereisnoreasonwhycostindicesmustberestrictedtodisplayingpasttrends;theycanalsoportrayfutureoverallcosts,providedtherepresentativepayitemsonwhichtheindexisbasedcanbeforecast.Apredictiveconstructioncostindexwasadoptedinthisstudytodescribethechangeinoverallconstructioncostsinthefuture.Theformulationoftheindexisdescribedlaterinthepaper.PastIncreasesinConstructionCostsWhenthechangeinoverallconstructioncostsinthepastisobserved(asmeasuredbypopularconstructioncostindices),itisapparentthattheychangesignificantlyfromyeartoyearandthatthechangescansometimesbequiteerratic.Thecommonassumptionthatconstructioncostschangewiththerateofinflationcanleadtopoorestimatesoffutureconstructioncost.Toillustrate,theFederalHighwayAdministration’sCompositeBidPriceIndex,anindexofoverallhighwayconstructioncosts,isplottedinFig.1togetherwiththeConsumerPriceIndex(CPI),acommonexpressionofgeneralinflation.TheFHWACBPIfortheentirenationandforLouisianaaloneisplottedinthediagram.Allindiceshavebeennormalizedtoavalueof100in1987forcomparisonpurposes.Fromthediagram,itisclearthathighwayconstructioncostschangeerraticallyandevendisplaydifferentshortandlong-termtrendsfromtothoseoftheCPI.ItisalsoapparentthatconstructioncostchangesaredifferentinLouisianafromthoseinthenationasawhole.Whilenotshownhere,reviewoftheFHWACBPIfromotherstatesshowsthatmanyofthemshowadeviationfromnationalvalues.PastMethodsofForecastingHighwayConstructionCostForecastingfuturehighwayconstructioncostshasbeenachievedinbasicallythreewaysinthepast.First,unitratesofconstructionsuchasdollarspermilebyhighwaytypehavebeenusedtoestimateconstructioncostsintheshortterm.However,thismethodhasgenerallybeenfoundtobeunreliable,becausesiteconditionssuchastopography,insitusoil,landprices,environment,andtrafficloadsvarysufficientlyfromlocationtolocationtomakeaveragepricesinaccurateestimatesofthepriceofindividualprojectsorevenofallprojectsinaparticularyear.Second,extrapolationofpasttrends,ortime-seriesanalysis,hasbeenusedtoforecastfutureoverallconstructioncosts(Koppula1981;Hartgenetal.1997).Typically,constructioncostshavebeencollapsedintheseanalysestoasingleoverallexpressionofconstructioncostsuchastheFHWACBPIortheEngineeringNewsRecord’sBuildingConstructionIndex~ENRBCI!orConstructionCostIndex~ENRCCI!.However,thesetypesofmodelsareusuallyonlyusedforshort-termforecastingduetotheirrelianceonthenotionthatpastconditionsaremaintainedinthefuture.Third,modelshavebeenestablishedthatdescribeconstructioncostsasafunctionoffactorsbelievedtoinfluenceconstructioncosts.Therelationshipbetweenconstructioncostsandthesefactorshavebeenestablishedfrompastrecordsofconstructioncosts.Typically,themodelsestablishedinthismannerhavebeenusedtoestimatethecostofindividualcontracts.Thesemodels,withtheirrelationalstructure,aretheonlymodelsexpectedtoprovidereliablelong-termestimates.Themodeldevelopedinthisstudyisofthistype.ProposedConstructionCostModelItisclearthattherearenumerousfactorsthataffectconstructioncosts.However,itisstrikingthatmostconstructioncostmodelsdevelopedinthepasthaveusedonlyafewofthemanyinfluentialfactorsidentifiedabove.Onereasonforthisisthatinformationisgenerallynotavailableonmanyfactorsindatasetsusedtoestimatemodels.Anotherreasonisthatinformationonthequalitativeconditionssurroundingeachcontractisdifficulttoobtain.Theseareproblemsthatprevailinmostcircumstancesandaredifficulttoovercome.Tomitigateagainsttheeffectofanincompletesetoffactors,twostrategiescanbeemployed.First,itmaybepossibletorepresentsomeoftheabsentfactorsbysurrogatevariablesthatareinthedataset.Forexample,asmentionedearlier,annualbidvolumehasbeenusedinthepastasaninversemeasureofthelevelofcompetitionprevailingintheconstructionindustryatthattime (Herbsman1986).Similarly,thenumberofplanchangeseachyearcanserveasameasureofdesignquality.Second,ifthemodelingofconstructioncostischangedfromestimatingthecostofindividualprojectstoestimatingoverallconstructioncostseachyear,themodelingtaskissimplified.Thisisbecauseitisnolongernecessarytotrytomodelindividualprojectsinwhichconditionsinflatethepriceinonecaseanddeflateitinanother,sincesuchconditionswouldtendtocancelthemselvesoutamongprojectsinthesameyear.Forexample,firmsthatreducetheirbidpricesinanefforttowinaparticularcontractcouldbebalancedoutwithinthesamefiscalyearbythosethatincreasetheirpricesbecausetheyalreadyhaveenoughworkandarenotparticularlyinterestedinwinningthecontract.Similarly,thosefirmswithexpertiseinthetypeofconstructionrequiredwillbebalancedoutbythosewithlowlevelsofexpertiseinthatarea.Thus,itisgenerallymoretolerabletooperatewithfewerrelevantfactorswhenmodelingattheaggregateoroveralllevelthanwhenmodelingatthedisaggregatelevel.Theobjectiveofthisstudyistoestablishamodel,estimatedonhistoricalquantitativedata,thatincorporatesasmanyrelevantvariablesaspossibleandiscapableofestimatingthefutureoverallcostofhighwayconstructiononanannualbasis.ThemodelisintendedtoassesstheimpactofalternativefutureconditionsonhighwayconstructioncostsandassistofficialsoftheLouisianaDOTDtoidentifymanagementpoliciesthatwillhelplimittheincreaseinhighwayconstructioncostsinthestate.Itwasalsotheperceptionofthoseinterviewedthatcontractsletinthefourthquarterofthefiscalyeartendedtoresultinhigherbidprices.Thiswasbecausetherewasatendencyforprojectstoaccumulateinthefourthquarterduetovariousdelays,andtheincreasedvolumeofprojectsresultedindecreasedcompetitionamongcontractors.ModelStructureThemodeldevelopedtopredictoverallhighwayconstructioncostsinthisstudyisbasedonfivesubmodelsofpriceestimation.Eachsubmodelestimatesthepriceofapayitemrepresentativeofcostmodeladominantconstructionarea.Dominantconstructionareaswereidentifiedfrompastexpenditureindifferentareasofhighwayconstruction.FromtheLouisianaDOTDdatafortheperiod1984–1997,itwasfoundthatmorethan50%ofallhighwayconstructionexpenditureoccurredintheareasofasphaltconcretesurfaces,Portlandcementconcretesurfaces,excavationandembankment,structuralsteel,structuralconcrete,andreinforcingsteel.Interestingly,theseconstructionareasareidenticaltothoseusedtoestimatetheFHWACBPI.Thestructuralsteelconstructionareawasnotincludedinthemodeldevelopedinthisstudy,becausemorethan98%ofexpenditureinthisconstructionareawasbidasalumpsumineachcontractwithnorecordoftheamountofsteelincludedinthebid.Thismadecomparisonofthecostofstructuralsteelamongcontractsimpossible.Theotherfiveconstructionareasincludedinthemodelwereallrepresentedbypayitemswhosepriceswereexpressedintermsofrates,whichpermittedcomparisonamongcontracts.AschematicrepresentationoftheoverallmodelwithitsfivesubmodelsisshowninFig.2.Eachsubmodelestimatesthepriceofarepresentativepayitemfromeachofthefivedominantconstructionareas.ThecontributionofeachsubmodeltotheoverallmodelisaccomplishedbycombiningthepricesoftherepresentativepayitemsinanindexsimilartothatoftheFHWACBPI.Inthiscase,becausetheformulationisslightlydifferentfromtheFHWACBPIandisconstructedspecificallytoreflectpastandfutureoverallconstructioncostsinLouisiana,itisnamedtheLouisianaHighwayConstructionIndexandisdefinedasionModelperformanceisideallyvalidatedusingdatanotusedintheestimationofthemodel.Inthiscasenosuchdatawasavailable.Dividingtheexistingdatasetintotwoportionstoestimatethemodelononeportionandusetheotherforvalidationwasnotpractical,giventhelimitedsamplesizeinsomeofthesubmodels.Forexample,theconcretepavementsubmodelhasatotalofonly212observations,andestimatingthesubmodelonthehighlyvariabledataonfewerobservationswouldreducetheaccuracyoftheestimates.Thus,theperformanceofthemodelwasassessedbyobservinghowwellitreproducedobservedconstructioncosts.Usingthesamedataasthatonwhichthemodelwascalibrated,theestimatedandobservedLHCIvaluesfortheperiod1984–1997areshowninFig.3.The95%confidencelimitoftheobservedLHCIisalsoshowninthefiguretoillustratethattheestimatedLHCIvaluesare,forthemostpart,containedwithinthe95%confidencelimitoftheobservedLHCIvalues.ThechisquaredtestofthesimilarityoftheestimatedandobservedLHCIvaluesindicatesthatasignificantdifferencecouldnotbeobservedatthe99%levelofsignificance.InvestigatingthebehavioroftheconstructioncostindexinFig.3revealsinterestingreasonsbehindtheobservedbehavior.Reviewingthedataandobservingitsimpactontheforecaststhroughthemodelallowsananalysttodeterminetheprimarycausesofchangeinconstructioncostsduringcertainperiodsinthepast.Forexample,themaincauseofthedecreaseinconstructioncostsobservedintheperiod1984–1986canbetracedbacktoadeclineinlaborandpetroleumcostsduringthatperiod.Therapidincreaseinconstructioncostsfrom1995to1996wasprimarilyduetoacombinationofrisingpetroleumcostsandanincreasedproportionofsmallercontracts.Thedropinconstructioncostsobservedimmediatelyfollowingthisevent(i.e.,in1997)wasmainlytheconsequenceofanincreaseintheaveragesizeofprojectsfromdurationofprojects.ConclusionsThisstudyhasshownthattheliteratureindicatesthatacomprehensivesetoffactorscontributestothecostofhighwayconstruction.Inthisstudy,themostinfluentialfactorswerefoundtobethecostofthematerial,labor,andequipmentusedinconstructingthefacility.However,alsoaffectconstructioncosts.Inparticular,contractsize,duration,location,andthequarterinwhichthecontractisletwerefoundtohaveasignificantimpactoncontractcost.Bidvolume,bidvolumevariance,numberofplanchanges,andchangesinconstructionpractice,standards,orspecificationsalsomakeasignificantimpactoncontractcosts.Otherfactorsareexpectedtohaveanimpactonconstructioncostsbutwerenotincludedinthisanalysisbecausenodataontheirvalueswereavailable.Themodeldevelopedinthisstudyreproducespastoverallconstructioncostsreasonablyaccuratelyattheaggregatelevel.Predictedoverallconstructioncostsarenotsignificantlydifferentfromobservedcostsatthe99%levelofsignificance.Thisaccuracyislargelytheresultoftheaggregatelevelatwhichconstructioncostsaremeasuredinthisstudy;attheindividualcontractlevel,thesubmodelscaptureonlybetween42and72%ofthevariationinthedata.Itissuspectedthatmuchofthisvariationisduetounobserved,essentiallysubjectivefactorsthatinfluencethebidpricesinindividualcontracts.However,someoftheseidiosyncraticvariationsattheindividualcontractlevelaverageoutintheaggregationprocess.ThismodelcanbeusedbyhighwayofficialsinLouisianatotestalternativecontractmanagementstrategies.Increasingcontractsizes,reducingthedurationofcontracts,reducingbidvolumeandbidvolumevariance,reducingthenumberofplanchanges,andreducingtheproportionofcontractsletinthefourthquarterallservetoreduceoverallconstructioncosts.Highwayofficialscanassesstheimpactofstrategiestheybelieveareachievablebyapplyingthemodel.Mostimportantly,though,themodelcanassistinestimatingfutureconstructioncostsandprovidingthemeanstoproducemorereliableconstructionprograms.AssociateProfessor,LouisianaTransportationResearchCenterandDept.ofCivilandEnvironmentalEngineering,LouisianaStateUniv.,BatonRouge,LA70803-6405.CivilEngineer,GEC,Inc.,9357InterlineAve.,BatonRouge,LA70809.C.G.Wilmot,M.ASCE,andG.Cheng,P.E．EstimatingFutureHighwayConstructionCosts．JOURNALOFCONSTRUCTIONENGINEERINGANDMANAGEMENT?ASCE/MAY/JUNEingindesign’.WorldComputing,Vol.26(1),pp74–76.ISO(1994).ISO10303-1Part1:Overviewandfundamentalprinciples,InternationalOrganizationforStandardization,Geneva,Switzerland.edmodeltosupportdistributedcollaborativedesignofbuildings’.AutomationinConstruction,Vol.7(2–3),pp177–188.LeeH.K.,LeeY.S.,KimK.H.andKimJ.J.(2007).‘Acost-basedinformationmodelforaninteriordesigninalarge-scalehousingproject’,ICCIT07,2007InternationalConferenceonConvergenceInformationTechnology,PosterSessions:Session4.ofConstructionEngineeringandManagement,Vol.123(2),pp113-120.公路建設(shè)造價的未來C.G.Wilmot,M.ASCE,1andG.Cheng,P.E.21聯(lián)合教授，路易斯安娜運輸研究中心和國家環(huán)境工程局，路易斯安娜國立大學2注冊工程師據(jù)公路建設(shè)造價索引，該模型介紹了中所有公路建設(shè)的工程費用。這索引是一份綜合定額，和項目環(huán)境基礎(chǔ)上的預測價格。在公路管理機構(gòu)控制下的項目特征和項目環(huán)境能可根據(jù)未來的成本削減政策來操作處理。公路造價預算模型在路易斯安娜的運用表現(xiàn)出這模型幾乎是1984—1997年時建設(shè)造價模型的重現(xiàn)。未來在運用公路造價預算時，該模型預測路易斯安娜的公路造價將在1998年到2015年間的增加兩倍。在應用削減成本政策和假設(shè)投入的資關(guān)鍵詞：公路建設(shè)，造價，預算立項中成為國家交通部的正式建設(shè)項目。辯護。這會導致使人不相信你的能力，并使來自立法機構(gòu)和公眾的支持受到?jīng)_擊。明顯，預留資金永遠不能確保，通常是對在建設(shè)項目實施中引起混亂的項目造價的評估。不準確的造價意見是一個錯誤來源，但另一方面，不斷過時的工程造價是另一個使項目失敗的項目造價估量當某領(lǐng)域建設(shè)落后于建設(shè)計劃中的進度，通常是因為已經(jīng)建造的項目費超過了預期費用。實際費用相對于預算不是隨意變動的，因為很明顯在這種情況下過低估計與過高估計會相互抵消。而且，這是所有項目全部過低評價的證據(jù)。這樣的好處是能在一般水平估量，這比在特別的項目水平估量容易得多。費用項目的平均造價。他們已經(jīng)習慣于根據(jù)過去造價。但是無法解釋造價索引必須嚴格表現(xiàn)過去趨勢的原因，他們同樣能表現(xiàn)未來一般造價，預備建立索引的典型費用項目，能預測。預算造價索引在本研究中被用于解釋未來在一般建設(shè)造價的變化。索引的簡化公式會在下文中提到。建設(shè)造價的增長歷史當過去一般建設(shè)造價中的變化被注意到了(如被普通建設(shè)造價索引估量)，很明顯這年復一年變化得很顯注而且有時非常沒規(guī)律。通常建設(shè)造價隨通貨膨脹率變化的設(shè)想會對未來建設(shè)造價做出錯誤的判斷。舉個例子，公路管理聯(lián)合會(FHWACBPI)的綜合投標價格索引是關(guān)區(qū)分為對于整個國家和只對路易斯安娜。所有索引已經(jīng)因相似目的在1987年根據(jù)100標準規(guī)范。從表中清楚看到公路建設(shè)造價變化不規(guī)律，甚至呈現(xiàn)趨向于CPI變化的不同長短期。而且很明顯建設(shè)造價變化在路易斯安娜和整個國家是不同的。沒有在此顯示的，其他國家FHWACBPI的評論表明其中一些與國家標準有偏差。過去公路建設(shè)預算方法單位建設(shè)價格已經(jīng)用于短期預算(HartgenandTalvitie1995；Stevens1995)。但是經(jīng)常發(fā)現(xiàn)格對特殊工程，甚至對特殊年份的所有工程進行不準確造價有很大差異(HartgenandTalvitieetal.1997)。通常，建設(shè)造價在僅對一般造價描寫的分析中已失去意義，如FHWACBPI，工程新紀錄之建筑工程定額(ENRBCI)，或建設(shè)造價定額(ENTCCI)。但是，這些模型通常只造價描述為影響造價的功能因素。建設(shè)造價和這些因素之間的關(guān)系已經(jīng)從過去的建設(shè)造價紀于提供不規(guī)律長期判斷的唯一模型。本研究中就是這個模型。地點、工程類型、項目持續(xù)時間和項目規(guī)模對單個項目造價有重大影響。Herbsman(1986)表示，除材料費，人工費和機械費，每年的項目總量(i.e.,theso-calledbidvolume)影響項目價格。這是智慧的提議，價位影響競爭(至少在短期內(nèi))，反過來，這也影響項目價格。Olsen和Epps(1981)表示出價量年復一年的變化影響出價價格。其他人認為政府法令，計劃變動，承造方管理團隊的質(zhì)量，提高竣工安全及時間的項目信息對建筑造價有影響(Koehnetal.1987;ElhagandBoussebaine1998).也有人說質(zhì)量因素，如建造企業(yè)的表里如一，建造者的需求，以及建造者與管理機構(gòu)的關(guān)系都會對建造有很大影響(Fayek1998)。建設(shè)造價模型很顯然有很多因素影響造價。然而令人吃驚的是過去大部分造價模型只用了一小部分重要因關(guān)于每個項目質(zhì)量情況的信息很難獲得，在大多數(shù)情況下成功和很難克服都是問題。例子，如早先提到的，過去每年的出價規(guī)模可作為評價那時建設(shè)產(chǎn)業(yè)競爭力水平的反因子 (Herbsman1986)。同樣的，每年計劃的改變量能評價設(shè)計質(zhì)量。其次，如果建設(shè)造價模型勝的企業(yè)會在同一年以因此提價來達到平衡，因為他們已經(jīng)做了足夠的工作而且不急于求成。少的因素在整體水平比在處于崩潰水平建立模型更好。公路造價增長的警務的管理。變化的信息。項目書及其變動和每年計劃改變的總數(shù)都是從原始資料中獲得的。次是面談，另兩次通過電話會議交談。面試官有一份問題單去控制面試，但回答人也可選擇回答范圍。這會確認實施中的變化，如更堅硬的道路設(shè)計能容納更多交通量，欄桿下鋪磚便于除草，注意沖蝕控制，欄桿末端處理變化，增加臨時交通線的用處。不管怎樣，因為這些些會談，這對以后瀝青水泥的出價增長立即負起了責任。變化包括聚合改性瀝青的介紹和提高材料種類測試程序的應用。這些變化是所期待的，同時也推廣使提高瀝青水泥的造價大約會談洞察到項目承包在四分之一財務年中會使報價更高。因為有一個趨勢是在四分之一年中會積累各種障礙，工程量的提高會減少承造者間的競爭。關(guān)于過去和未來建設(shè)資金的資料是從多種來源獲得的。貿(mào)易局的經(jīng)濟研究處公布了工業(yè)雇傭和收入的歷史記錄，用來評價每個工業(yè)部門的工人人均工資。這個信息用來確定1980-1997人薪資提供信息。1987年過去和預測的工人薪資水平被編成了一份反映每100勞動力價值的統(tǒng)計表。1984-2015年間公路建設(shè)器材和材料價格資料來源于一家信息資源公司(DRI)。見表一典型支付項目混合資源中的替代變量建設(shè)設(shè)備建設(shè)機器路基材料建設(shè)用沙/碎石/石料AA級水泥水泥原料和相關(guān)產(chǎn)品瀝青混凝土精制石油產(chǎn)品和建設(shè)用沙/碎石/石料變形強型鋼模型結(jié)構(gòu)混凝土加強木棍和碳評價子系統(tǒng)組成。每個子模型評價一個建設(shè)優(yōu)勢娜DOTD1984-1997年間的資料可以發(fā)現(xiàn)超過50%的公路建設(shè)費用用于瀝青混凝土表面，波特蘭水泥混凝土表面，坑道和路基，鋼結(jié)構(gòu)，混凝土結(jié)構(gòu)和固型鋼。有趣的是，這些建設(shè)地FHWACBPI的，因為法比較。模型中其他五項都由支付項目表示，其價格用比率表示更易比較。每個子模型的建立是為了降低當年描述各種建設(shè)投資的可變支付項目價格，考慮中的項目特征和那時良好的項目環(huán)境使項目很順利。在模型中嘗試盡可能多的獨立的各類因素使一般模型中影響建設(shè)造價的因素盡可能綜合和敏感。所示。每個子模型從五個優(yōu)勢建設(shè)區(qū)域評價一個選擇性支付項目的價格。整體模型的每個子模型的貢獻是整合了與FHWACBPI的索引表相似的索引表中的可選擇性支付項目的價格。在這里，因為這形式和FHWACBPI有細小的差別而且是有效性型一共只有212項觀測值，信息多樣但觀測值相對較少會降低對模型評價的精確度。所以模型執(zhí)行以觀察它怎樣重新觀測建設(shè)造價來評估。Fig3用模型測量出來的資料，1984-1997年間評價和觀察的LHCI值如Fig3所示。LHCI的信用限制。LHCI評價和觀測值相似性測驗表明顯著的不同點在99%不能被注意到。預先計劃的影響，這使分析家找到了在過去一些時期中建設(shè)造價改變的主要原因。舉個例子，觀察1984-1986年間建設(shè)造價下降的主要原因要追溯到那段時期勞動力和石油價格的下降。后建設(shè)造價的下跌(i.e.,in1997)主要原因是1996年承包的項目平均規(guī)模的增大和項目平本研究已經(jīng)表明論文說明了公路建設(shè)造價的一個系列綜合原因。研究表明最有影響力的因素響工程造價。特別要指出，項目規(guī)模，持續(xù)時間，地點和項目投資對建設(shè)造價的影響也是非計說明書的改動也能起到很大影響。其他會產(chǎn)生影響的因素不包括在本分析中因為缺乏相關(guān)可利用的數(shù)據(jù)。確地重現(xiàn)了過去普通建設(shè)造價。建設(shè)預算在99%的重要性上與決算相差無幾。相較于個別項目，子模型只能獲得資料中42%—72%的變化程度的情在整體過程中被平均。這模型判斷出到2015年如果過去的趨勢被推斷或通貨膨脹率作為評價未來建設(shè)造價依據(jù)，那么路易斯安娜的公路建設(shè)造價增長會比預期更快。根據(jù)模型預測，即使經(jīng)濟情況比預期更舊可能會流行，。建設(shè)造價中工程增加的主要原因是石油生產(chǎn)和建設(shè)機器價格的預期增長?？赡軙^緩慢。謝本論文中運用的研究成果來自于路易斯安娜運輸研究中心。向他們和來自路易斯安娜運輸發(fā)展局工作人員的幫助表示衷心的感謝。來源：C.G.Wilmot,M.ASCE,andG.Cheng,P.E．EstimatingFutureHighwayConstructionCosts．JOURNALOFCONSTRUCTIONENGINEERINGANDMANAGEMENT?ASCE/MAY/JUNEUsingWeibullAnalysisforEvaluationofCostandSchedulePerformanceKhaledM.Nassar1;HordurG.Gunnarsson2;andMohamedY.Hegab,P.E.3Largeamountsofmoneyarelosteachyearintheconstructionindustrybecauseofpoorscheduleandcostcontrol.Fewcontractorsspecifyandfollowsystematicschedulemonitoringpractices.Traditionally,theearnedvaluemethodEVMisusedtocontrolandmonitorscheduleperformanceusingthescheduleandcostperformanceindiceswhichcomparethebudgetedcostofworkperformedtowhatwasoriginallyscheduledorwhatisactuallyexpended.Thispaperpresentsastatisticalapproach,namelyWeibullanalysis,toevaluatestochasticallythescheduleperformanceofconstructionordesignprojects.TheapproachcanbeusedinconjunctionwiththeEVMtoenhancetheevaluationandcontrolofscheduleperformance.Weibullanalysisisacommonmethodforfailureanalysisandreliabilityengineeringusedinawiderangeofapplications.Inthispaper,theapplicabilityofWeibullanalysisforevaluatingandcomparingthereliabilityofthescheduleperformanceofmultipleprojectsispresented.Thevariousstepsintheanalysisarediscussedalongwithanexampleinwhichtwoprojectsareanalyzedandcompared.TheauthorsconcludethatWeibullanalysishasseveraladvantagesandprovidesarelativelyrobustandeffectivemethodforconstructionmanagerstobettercontrolandmonitortheirprojects.Constructionprojectsareseldom“on-schedule”allthetime.Theamountofworkperformedontheprojectusuallyfluctuatesfromoneperiodtotheother.Infact,accordingtothe“80/20”rule.suggestedbythe18thCenturyeconomist,Pareto80%oftheeffortwillbeexpendedduring20%oftheprojectduration.Thiscreatesasignificantproblemintryingtomonitortheactualperformanceoftheprojectscheduleandspecificallytryingtodecidewhetherornottheprojectcanfinishontime.,thereliabilityoftheprojectscheduleperformance.TheearnedvaluemethodEVMisoftenusedasaprojectcontroltechniquetoprovideaquantitativemeasureofscheduleperformance.IntheEVM,thescheduleandcostperformanceindicesC/SPIareusedforconstantmonitoringoftheproject’scostandschedulebasedonanoriginalcriticalpathmethodCPMscheduleChang2001.Unfortunately,CPMschedulingingeneralcreatesanunrealisticexpectationregardingprojectscheduleperformance.Startingaprojectonedayandexpectingthattheprojectwillfinishonanexactdatesometwoorthreeyearsinthefutureisunrealistic.Therefore,thereisaneedtoextendscheduleanalysistechniques,liketheEVM,toprovidemeansforprobabilisticallyanalyzingthescheduleperformanceandmeasuringtherisksinvolved.Inothermanagementdomainslikesoftwareprojectsandbusinessmanagement,toolsfromtheareaofreliability,availability,andmaintainabilityRAMengineeringarenowusedtopredictandavoidsuddensystemfailures.RAMengineeringisemergingasthenewestdisciplineinproductdevelopmentduetonewaccurate,quantitativemethods.Weibullanalysisisforemostamongthesetools.Weibullanalysisisawidelyusedmethodinthefieldofmanufacturingandindustrialengineering.Ithasmainlyservedforsimulatingfailureratesorlifetimeofcertainproducts.Byusingthisparticularanalysisonecanmakepredictionsaboutproduct’slife,comparethereliabilityofcompetingproductdesigns,andmanagesparepartsinventoriestonameafewcommonindustrialapplicationsDorner1997,1999.TheobjectiveofthisstudyistodiscussandpresenttheapplicabilityofWeibullanalysisforevaluatingscheduleperformanceusingcostandperformanceindicesC/SPIandtoprovideconstructionmanagerswithaneffectivetoolforstochasticallyevaluatingtheC/SPI.AnumberofprobabilistictechniqueswerecheckedagainstthedatausingAnderson-DarlingmeasureADvalueandtheWeibulldistributionwasfoundtobethebestrepresentativeforevaluatingscheduleperformanceandcostperformancehavingthelowestADvalue.Inthispaper,anexampleisprovided,inwhichtwoprojectsarecomparedandanalyzed.ThefirstpartofthispaperbrieflycoverstheEVMandperformanceindices.ThesecondpartdiscussestheWeibulldistribution,it