土地利用與交通需求 PPT課件版 L3 Trip Generation

TripGeneration

出行生成楊超土地利用與交通需求分析同濟(jì)大學(xué)email:tongjiyc@2WhatisTransportation?TransportationinvolvesthemovementofpeopleandgoodsChangeinspace,time,and“utility”UtilityisameasureofsatisfactionorusefulnessTransportationisthereforegenerallynotanendinitself,butoccurstosatisfydemandforsomeactivitiesatthedestination.Transportationisa“derived”demand3tripgenerationtripdistributionmodesplittripassignmentTijmrTiTijTijmStep1.Decisiontotravel:forecastthenumberoftripsthatwillbemadeStep2.Choiceofdestination:determinewheretripswillgoStep3.Choiceoftravelmode:predicthowtripswillbedividedamongavailablemodesStep4.Choiceofroute:predicttheroutesthattripswilltake,resultingintrafficforecastsforthehighwaysystemInputs:LanduseandsocioeconomicprojectionsTransportationsystemArtificialfeedbackTraditionalTrip-basedModel:4-StepTravelDemandModelor{Flows&LevelofService}4出行產(chǎn)生4出行吸引5出行分布6方式劃分7交通分配（路徑選擇）89ThreeZoneExample104-StepProcedureExampleTripGenerationTripDistributionModeChoiceRouteChoiceOi,Dj Tij Tijm TijmrTAZOi123T4766110223TAZDj123T45908822340Tij1123Dj1030545231832901946588Oi4766110223ModeT32mcarbuswalkTotal

2015540PathT32carrR1R2R3Total512320fromiTrafficAnalysisZonetripproductiontripattractiontoj我國(guó)目前城市交通規(guī)劃技術(shù)規(guī)范解決：要不要建、什么時(shí)候建、建多大規(guī)模、建在什么地方等問題通過規(guī)劃，預(yù)先知道建設(shè)項(xiàng)目實(shí)施后的效果，避免盲目建設(shè)11住房和城鄉(xiāng)建設(shè)部2010年3月發(fā)布“城市綜合交通體系規(guī)劃編制辦法”

城市綜合交通體系規(guī)劃是城市總體規(guī)劃的重要組成部分城市綜合交通體系規(guī)劃應(yīng)當(dāng)與城市總體規(guī)劃同步編制，相互反饋與協(xié)調(diào)不同層次交通規(guī)劃的重點(diǎn)交通綜合網(wǎng)絡(luò)規(guī)劃/整體交通規(guī)劃交通專項(xiàng)規(guī)劃/分區(qū)交通專項(xiàng)規(guī)劃城市交通發(fā)展目標(biāo)和水平城市交通方式和交通結(jié)構(gòu)重大設(shè)施的選址和用地規(guī)模實(shí)施規(guī)劃的重要技術(shù)經(jīng)濟(jì)對(duì)策城市交通各子系統(tǒng)的發(fā)展目標(biāo)，具體規(guī)劃方案局部地區(qū)交通特征分析，提出改善處理的方案大型單體建筑項(xiàng)目實(shí)施對(duì)周邊交通系統(tǒng)的影響各種運(yùn)輸系統(tǒng)的布局，各種交通的銜接方式大型交通樞紐的分布與用地范圍道路網(wǎng)絡(luò)布局、道路等級(jí)和功能劃分規(guī)劃方案技術(shù)經(jīng)濟(jì)評(píng)估，提出分期建設(shè)建議局部地區(qū)交通改善規(guī)劃/交通影響分析/交通設(shè)施規(guī)劃設(shè)計(jì)交通發(fā)展戰(zhàn)略規(guī)劃側(cè)重于城市交通發(fā)展方向的研究側(cè)重于配合城市總體規(guī)劃的研究側(cè)重于具體實(shí)施側(cè)重于各子系統(tǒng)的自身分析重要設(shè)施運(yùn)行功能分析與方案設(shè)計(jì)12上海市域交通小區(qū)劃分市域309個(gè)小區(qū)中心城263個(gè)小區(qū)上海市域交通小區(qū)劃分進(jìn)出口19處，其中高速公路3條隨著上海1966規(guī)劃的提出和落實(shí)、第四次綜合交通調(diào)查的開展，交通模型重新劃分小區(qū)：全市約4000個(gè)小區(qū)，中心城約800個(gè)，郊區(qū)約2000個(gè)，其它為特殊點(diǎn)（機(jī)場(chǎng)、火車站、碼頭、地鐵站等）和進(jìn)出口每個(gè)小區(qū)面積（km2/個(gè)）中心城（約640km2)：1.25郊區(qū)（約5700km2)：2.8如何劃分小區(qū)一般劃分小區(qū)有幾個(gè)層次：行政區(qū)域?yàn)榇髤^(qū)、再根據(jù)街道、居委、村鎮(zhèn)等劃分中區(qū)、最后是小區(qū)根據(jù)自然條件和物理分割，結(jié)合行政邊界確定小區(qū)邊界，同時(shí)考慮現(xiàn)狀及規(guī)劃的高速公路、快速路作為邊界小區(qū)的規(guī)?？蓞⒖碱愃瞥鞘校M量保持小區(qū)出行量的接近水平注意特殊點(diǎn)、進(jìn)出口的處理如何劃分小區(qū)如何劃分小區(qū)小區(qū)形心以小區(qū)形狀的形心為基礎(chǔ)，參考小區(qū)內(nèi)用地的特征，將主要的交通產(chǎn)生、吸引點(diǎn)的地塊作為形心位置小區(qū)連接線根據(jù)主要的交通出行方向確定大致方向（注意出行主流向），小區(qū)連接線一端為小區(qū)形心，另一端優(yōu)先選擇該方向的支路的出入口，盡量不連接到主干道小區(qū)連接線數(shù)量以4條為宜交通模型分區(qū)新趨勢(shì)中觀模型：杭州核心區(qū)14km2范圍劃分小區(qū)243個(gè)，平均小區(qū)面積約0.057km2小區(qū)用地分類參考城市規(guī)劃規(guī)范、控制性詳細(xì)規(guī)劃以及類似城市（北京、深圳、上海）用地分類情況同時(shí)參考《北京市交通出行率指標(biāo)》，確定用地分類原則用地分類備注住宅分為兩類，即高檔住宅，一般住宅工廠中觀范圍內(nèi)工廠定為高新研發(fā)產(chǎn)業(yè)用地；外圍區(qū)域分三類，即高新研發(fā)產(chǎn)業(yè)用地、一般工業(yè)用地、制造業(yè)工業(yè)用地商業(yè)文化娛樂辦公與金融學(xué)校分三類：大專院校，中學(xué)，小學(xué)幼兒園醫(yī)院旅館分兩類：三星級(jí)及以上旅館，一般旅館特殊吸引點(diǎn)商業(yè)中心，大型醫(yī)院，政府辦公用地（包括省、市、區(qū)三級(jí)），車站，機(jī)場(chǎng)，風(fēng)景點(diǎn)，倉庫貨場(chǎng)，停車場(chǎng)，大型體育設(shè)施其他包括水域，綠地，市政公用設(shè)施總計(jì)九大類Step1:TripGeneration21TripGenerationAnalysis1ststepinthe4-steptravelforecastingprocessRelatetheintensityoftripmakingtolanduseandsocioeconomiccharacteristicsProvideameasureoftravelfrequency,yieldingatotalvolumeoftripsdefinedinagivenregionatagiventimeLand-use&socio-economiccharacteristicZoneiDependingontheactivitiesinthezone,itcanproduceand/orattracttrips.22TripClassificationbyTripPurposeTypically,aminimumofthreecategoriesareused:HBW-HomeBaseWorktrips(Residential).HBO-HomeBaseOthertrips(Residential)NHB-Non-HomeBasetrips(Non-Residential)Truck,transit,andtaxitrips(ifany)aretreatedseparately.Rcationaltrips.NHBtripsbyspecialgenerators.23TripClassificationbyTripPurpose:Example基本定義出行：從O-D的單向移動(dòng)HB：O/D是家NHB：O和D都不是家出行產(chǎn)生：HB的家或NHB的起點(diǎn)出行吸引：HB的非家端或NHB的終點(diǎn)出行生成：產(chǎn)生和吸引的總和25ModelingTechniquesITE(InstituteofTransportationEngineers)TripRates.Typicallyusedfortrafficimpactanalysis(bytrafficengineers)CrossClassification(orCategoryAnalysis)UsealotfortripproductionMultipleRegressionWorkbetterasatripattractionmodel.26ITETripRatesITETripGenerationHandbookTripGentransportationsoftware(Version9)Ratesbasedupondemographics(averagehouseholdsize,businesstype,numberofemployees,etc…)AcompilationofdatafromalloverNorthAmericaonmanydifferenttypesoflandusesProductionsandattractionsforeachtypeoflandusearerelatedtosomemeasurablevariablesUsealotinsiteimpactanalysis27TripGenerationRateExampleSingle-familydetachedhousingAveragetriprateperdwellingunit調(diào)查資料整理分析北京市交通出行率指標(biāo)（北京交通發(fā)展研究中心）28北京市交通出行率指標(biāo)2930Cross-Classification(orCategoryAnalysis)HouseholdsinTAZsaggregatedintogroupsRatesforeachgroupusedtodeterminethenumberoftripsTripratesbasedonhouseholdcharacteristics(incomelevel,vehicleownership,householdsize,…)31CrossClassificationModelsAssumestripratesarestableovertimeDerivedfromtravelsurveysGrouphouseholdsindifferentstrataAdvantagesGroupingsareindependentoftheTAZsystemNopriorassumptionabouttherelationship(i.e.,linear)RelationshipscandifferfromcelltocellSimple,easytounderstandDisadvantagesDoesnotpermitextrapolationNogoodnessoffitmeasuresRequireslargesamplesizeNoeffectivewaytochooseamongvariablesforclassificationortochoosebestgroupsforagivenvariable32Cross-ClassificationExample(1)Thefollowingdatahasbeengatheredforathree-zonecity.DistributionofAutosperHHbyIncome

Householdswithcars(%)Zone

Household

MeanHHInc. 0123+1 4323 $40,000 101060 202 7072 $30,000 153045103 3016$20,000 20403010

33Cross-ClassificationExample(2)

PersonTripsperHHbyCarsandTripPurpose

PersonTripsbyCars

(%)bypurposeZone

MeanHHInc.0123+

HBWHBONHB1 $40,00081115 171236242 $30,00061014161834223 $20,000571012303020

Question:WhatarethetriptotalsforHBW,HBO,andNHBtripcategories?34Cross-ClassificationExample(3)TotalPersonTripsbyTripPurpose

Generated

MeanTrips

TripsbyPurposeZone

TotalTrips

perHH

HBWHBONHB1 61819

14.3

74182725512364 2 83450 11.8 1502128373200283 24128 8.0723872385791Total 169397 11.7 296785786638182 8(.1)+11(.1)+15(.6)+17(.2)=14.314.3(4323)=61819 (61819)(.12)=7418MeantripsperHHHH%bypurpose35Cross-ClassificationExample(4)Households(%),PersonTripsperHH,andTripsasafunctionofHHIncome36RegressionModelsSimpleregression:onlyoneindependentvariableMultipleregression:morethanoneindependentvariablesNumberoftrips=f(population,autos,numberofdwellingunits,…)Trippredictors(i.e.,independentvariables)needtobeindependentFactorExamplePersonalcharacteristicsGender,Age,Income,OccupationHouseholdcharacteristicsFamilysize,Carownership,Numberandageofchildren,HouseholdincomeZonalcharacteristicsLanduse,Residentialdensity,Numberofemployment,Accessibility37RegressionModelsforTripGenerationRegressionAnalysisHHTrip/Day13122325210212726310312311411414410512618812816915101638OrdinaryLeastSquaresYyieie1e2e3e4e5X(xi,yi)xi39OrdinaryLeastSquares(Cont.)Question:Intheabsenceofavalidrelationshipb/tyandx,whatisyourbestguessforanyvalueofx?40GeometryoftheRegressionLineYXDeviationoftheobservedvalueYfromtheregressionestimateDeviationoftheregressionestimatefromthemean41GeometryoftheRegressionLine(Cont.)TotalSum-of-the-Squares ObservedvaluesaboutthesamplemeanRegressionSum-of-the-Squares Estimatedvaluesaboutthesamplemean Variation“capturedby”or“explainedby” themodel’sexplanatoryvariablesErrorSum-of-the-Squares Errorterm,minimizedinOLSDefinitionExpression42GoodnessofFitSSTisfixedoncethedatasetisselected.SSEisminimumundertheOLSestimator.SSRismaximizedsinceSSTisfixed(i.e.,maximizetheexplanationpoweroftheregressionline).CoefficientofDetermination(orR2)ThecloserthevalueofR2isto1.0.,thebetterthelinearmodelisintermsofaccountingforthevariationinthedependentvariabley.43GoodnessofFit(Cont.)SimplePearsonCorrelation(standardizedmeasureofcovariance)Correlationb/tdependentvariableyandindependentvariablexMultipleR(orMultipleCorrelationCoefficient)RangeofRisgreaterthanR2sinceCOV(x,y)maybepositiveornegative.44AnalysisofVariance(ANOVA)RegressionErrorsTotalSource

Sum-of-Squares

df

MeanSS

F-Ratiokn-k-1n-1MSR=SSR/kMSE=SSE/(n-k-1)F=MSR/MSEEquivalentF-RatioSinceandGeneralrule:F≥445StandardErroroftheEstimateAnestimateofthevarianceofthepopulationyaboutthetrueregressionline.46SignificantoftheRegressionCoefficientsStandardErroroftheBivariateRegressionCoefficient(b1)iny=b0+b1xwhereGeneralrule:(95%certainthatb1exceedsapproximatelytwiceitsstandarderror)StandardErroroftheMultipleRegression(i.e.,y=b0+b1x1+…+bkxk)whereistheithdiagonalelementoftheinvertedcorrelationmatrix.Sameasabove47StepsinDevelopingRegressionModelsIdentifyindependent(causal)variable(s)Basedonexperience,judgmentandopinionsfromexpertsCollecthistoricaldataonbothdependentandindependentvariablesDependentvariables:tripproductionortripattractionIndependentvariables:socio-demographicsforproductionandlanduseandemploymentforattractionPlotandidentifythefunctionalrelationshipLinear,exponentialorpolynomialDeterminetheparametersUseanystatisticalsoftware(e.g.,MicrosoftTool=>DataAnalysis)48FunctionalFormsofRegressionModelsForm

Model

Dependent

Independent

Parametersb0,b1b0,b1,…,bnb0,b1b0,b1yyyyxx1,x2,…,xnxxSimpleMultipleExponentialNonlinear49Example1:SimpleRegressionAlanduseplannerobservedthatin5zonesofthecitythenumberofgasstations(y)inrelationtothepopulation(x)in1000wasasfollows:(x,y):(1,2),(5,7),(3,3),(2,5),(4,8)SetupalinearequationconnectingyintermsofxDetermineR2andapplythet-Test50Example1:SimpleRegression51Example1:SimpleRegression52Example1:ResultsfromExcel53Example254Example2:Results55Example2:ResultsModelComment:ModelissignificantsinceitpassesbothF-Test(F>4)andt-Test(t>2).Also,thesignofthecoefficientforLABFispositive,whichiswhatwewouldexpect.案例——綿陽出行生成綿陽出行生成模型出行產(chǎn)生模型——交叉分類模型；根據(jù)家庭車輛保有情況（3類），職業(yè)（4類），出行目的（6類），交叉得到每種人群的出行產(chǎn)生率。出行吸引模型——基于小區(qū)各類用地性質(zhì)的線性模型。案例——綿陽出行生成車輛保有模型

模型建立了家庭月均收入水平和家庭小汽車保有量之間的關(guān)系。在規(guī)劃年，家庭收入將隨著GDP的增長(zhǎng)而增長(zhǎng)，因此就可以預(yù)測(cè)未來年各個(gè)收入水平下家庭小汽車的保有量。

模型將預(yù)測(cè)不同的收入水平下沒有小汽車、擁有一輛小汽車和擁有兩輛及以上小汽車的百分比。公式如下：P(N)i為第i小區(qū)家庭擁有N輛車的百分比;Ii為第i小區(qū)的收入；

K，b，bh為待標(biāo)定的參數(shù)。案例——綿陽出行生成車輛保有模型

P(1+)=S1h/(1+exp(-LP1)S1h=saturationlevelforHH,assume1

LP=k+(b+bh)ln(y)

-19.241.58170.54961

IncomeP(1+)-surveyskbbhS1hLPexpP(1+)-modelsqerror10000.021167-19.241.58170.54961-4.51891.6551%0.00020000.045208-19.241.58170.54961-3.04120.9215%0.00030000.095776-19.241.58170.54961-2.1778.816110%0.00040000.181044-19.241.58170.54961-1.5634.775317%0.00060000.296645-19.241.58170.54961-0.6992.012333%0.00180000.507921-19.241.58170.54961-0.0861.0948%0.001100000.635964-19.241.58170.549610.38940.677460%0.002120000.64-19.241.58170.549610.7780.459369%0.002rsq0.987413

0.006調(diào)查所得的總的具有1輛及以上車的家庭收入比例模型得到對(duì)應(yīng)不同收入層次擁有車輛數(shù)的比例案例——綿陽出行生成車輛保有模型

模型擬合標(biāo)定擁有一輛及一輛以上車的百分比（P1+）和擁有2輛及2輛以上車在擁有1輛及1輛以上車中所占的百分比的系數(shù)（P2+），然后通過計(jì)算可得：沒有車的百分比P(0),擁有一輛車的百分比P(1)和擁有兩輛及以上車的比例P（2）。得到

zone平均收入ncoco1co2+1017000.0769230.5215263240.4680141850.0104594911024250.156250.6680301780.3308632210.0011066011032900.20.8981265140.101869983.5062E-0610400.9892072410.0107927592.62861E-101054250.250.6680301780.3308632210.0011066011063366.1759780.8268476350.1731125523.98132E-051073363.8636360.8268476350.1731125523.98132E-051083212.650.8268476350.1731125523.98132E-051093502.9986230.8268476350.1731125523.98132E-051102695.7227270.8981265140.101869983.5062E-061112091.2454550.8981265140.101869983.5062E-0611200.9892072410.0107927592.62861E-101132570.4824560.8981265140.101869983.5062E-061142875.3750.8981265140.101869983.5062E-061151861.50.9543809910.0456189011.07967E-0711600.9892072410.0107927592.62861E-101173136.6818180.8268476350.1731125523.98132E-0511836670.8268476350.1731125523.98132E-051192235.1439390.8981265140.101869983.5062E-061203187.750.8268476350.1731125523.98132E-05案例——綿陽出行生成出行產(chǎn)生模型模型具有以下計(jì)算表達(dá)式：其中：為i區(qū)出行的產(chǎn)生量為c類的平均出行產(chǎn)生率為i區(qū)c類的總?cè)丝贜為交叉分類的總分類數(shù)案例——綿陽出行生成出行產(chǎn)生模型

模型將人群按不同的車輛擁有情況（0輛，1輛，2輛及以上），職業(yè)情況（在職，在校大學(xué)生研究生，中小學(xué)生，其他），出行目的（HBW基家工作、HBCOL基家去大學(xué)、HBSCH基家去中小學(xué)、HBSH基家購(gòu)物、HBO基家其他、NHB非基家出行），由調(diào)查數(shù)據(jù)通過excel統(tǒng)計(jì)得到各類的出行數(shù)量除以

各分類的人數(shù)（通過excel統(tǒng)計(jì)可得），得到各類的出行率（去除邏輯上不可能的組合，見下頁表格。）

案例——綿陽出行生成出行目的個(gè)人類型NCOCO1CO2+全家庭HBW在職1.481.521.401.49其它0.360.390.110.36HBCol在校大學(xué)生、研究生0.850.620.410.78