成果匯報最終版

1、SPACE SYNTAX FINAL REPORTTake Zhong Shan Road Subway Station and surrounding land for example 王懿寧姚嘉倫尤智玉張嬙常銘瑋杜孟鴿李金王藝霖1王櫟盛強指導教師：研究生領隊：成員：The Influence Factors of Traffic flowCatalogueTopological ParameterRoad WidthComfort DegreeFurther Research about Pedestrian Flow Separate the influence of pedestria

2、n from subway station FunctionsComparison with Different Subway StationsRevamping Options Visibility graphAnalysis of SiteThe Average Walking DistanceCoverage areaSiteAnalysis of SiteChina301Main RoadSecondary RoadPathRoads GradeN500m4Zhongshan RoadNorth StationSubway StationBus StationPublic Transp

3、ortationSubway LineN500m5FunctionsMall&MarketHotelSchoolGreenlandHospitalBankN500m6Public facilities Greenland500m1000mN500m7Public facilities Hotel500m1000mN500m8Public facilities Mall&Market500m1000mN500m9Public facilities School500m1000mN500m10Public facilities Hospital500m1000mN500m11Public faci

4、lities Bank500m1000mN500m12The Influence Factors of Traffic flowCatalogueTopological ParameterRoad WidthComfort DegreeFurther Research about Pedestrian Flow Separate the influence of pedestrian from subway station FunctionsComparison with Different Subway StationsRevamping Options Visibility graphAn

5、alysis of SiteThe Average Walking DistanceCoverage area01Peak flow on Weekday Average flow on Weekday Peak flow on Weekend Average flow on Weekend PedestrainBicycleAutomobile14Statistics of traffic flowRegression analysis and thinking based on flowRegression analysis and thinking based on flowR=1600

6、R=2400R=3200R=5000R=8000R=20000R=50000R=0.244R=0.676R=0.687R=0.616R=0.645R=0.405R=0.522The relevance of NACH under different R and weekday average automobile flow weekday peak value of automobile flow weekend average automobile flow01R=nR=0.644R=0.287R=0.287R=0.707R=0.717R=0.685R=0.569R=0.653R=0.685

7、R=0.258R=0.559R=0.608R=0.559R=0.487R=0.568R=0.559R=0.40015R=0.289R=0.673R=0.643R=0.535R=0.524R=0.544R=0.526The relevance of Integration under different R and weekday average automobile flow weekday peak value of automobile flow weekend average automobile flowR=0.526Regression analysis and thinking b

8、ased on flow01R=0.298R=0.461R=0.450R=0.452R=0.470R=0.452R=0.326R=0.457R=0.452R=0.586R=0.449R=0.456R=0.669R=0.636R=0.585R=1600R=2400R=3200R=5000R=8000R=20000R=50000R=nR=0.45616Regression analysis and thinking based on flow01R=0.510Regression on NACH R8000 and Integration R2400We found that NACH R8000

9、 is correlate to Integration R2400, so we cant use them to do multiple regression analysis. Thus we only use the NACH R8000 to do the prediction, which is more steady than multiple regression analysis.Zhongshan Road space DNA of automobile =7.02689*value(“NACH R8000”)-3.62783NACH R8000 Integration R

10、2400R=0.67617Nach R400Weekday average bike flow Regression on NACH under different R and bike flowNach R1600Nach R800Nach R1200Nach R2400Nach R3200R=1600mNach R400R=0.156Nach R800R=0.251Nach R1200R=0.346Nach R1600R=0.522Nach R2400R=0.482Nach R3200R=0.370Nach R nR=0.11818Integration R400average bike

11、flow Integration R1600Integration R800Integration R1200Integration R2400Integration R3200R=3200mIntegration R400R=0.160Integration R800R=0.247Integration R1200R=0.341Integration R1600R=0.431Integration R2400R=0.461Integration R3200R=0.448Integration R nR=0.446Regression on Integration under differen

12、t R and bike flow19R=0.156 R=0.247 R=0.346 R=0.522 R=0.482 R=0.370 R=0.446R=0.160 R=0.247 R=0.341 R=0.431 R=0.461 R=0.448 R=0.446IntegrationNach01Regression analysis and thinking based on flow20R=0.48901Regression on NACH R1600 and Integration R3200R=0.491Zhongshan Road space DNA of bike=0.006913*va

13、lue(“Integration R3200”)+0.456013Regression analysis and thinking based on flow21500mR=1000mNPath TrackingZhongshan RoadSubway Station01Normalized Angular ChoiceLowHighNACH Zhongshan RoadKunwei RoadLvwei RoadJinwei RoadHuangwei Road（expressed by colour）Pedestrian flow & NACHR=0.082The correlations b

14、etween pedestrian flow and total Nach is low, so that we will continue to analysis relevance between pedestrian flow and Nach in different radius22Regression analysis and thinking based on flowNACH-R400 NACH-R3200 R=0.116NACH-R800NACH-R1600NACH-R1200NACH-R2400NACH-R400 R=0.201NACH-R3200NACH-R1600 R=

15、0.206NACH-R1200R=0.205 NACH-R2400 R=0.163NACH-R800R=0.201R=1600m The Relevance of NACH (different radius) and Pedestrian Flow23500mNPath TrackingR=1000mZhongshan RoadSubway StationT1024 IntegrationLowHighIntegration （expressed by colour）Pedestrian flow & IntegrationR=4.82147e-005The correlations bet

16、ween pedestrian flow and total integration is low, so that we will continue to analysis relevance between pedestrian flow and integration in different radius2401Regression analysis and thinking based on flowintegration-R400Integration-R3200 R=0.101integration-R800integration-R1600integration-R1200in

17、tegration-R2400integration-R400 R=0.222integration-R3200integration-R1600 R=0.362Integration-R1200R=0.309integration-R2400 R=0.193Integration-R800R=0.284R=1600m The Relevance of Integration (different radius) and Pedestrian Flow25SUMMARY OUTPUT回歸統(tǒng)計Multiple R0.742982R Square0.552022Adjusted R Square0

18、.548319標準誤差0.096584觀測值123方差分析dfSSMSFSignificance F回歸分析11.390911.39091149.10237.71E-23殘差1211.1287560.009329總計1222.519665Coefficients標準誤差t StatP-valueLower 95%Upper 95%下限 95.0%上限 95.0%Intercept0.7084120.0382718.510954.24E-370.6326470.7841770.6326470.784177X Variable 10.0020710.0001712.210757.71E-230.0

19、017350.0024070.0017350.002407R=0.552VSConclusion: The Relevance of Integration-R1600 and NACH-R1600Integration-R1600 metric is highly correlated with NACH-R1600, so we cannot introduce both factors into regression analysis.X=T1024 Integration-R1600 metricY=NACH-R160026RestaurantR=0.115EducationR=0.0

20、02HotelR=0.004Relevancy between functions and pedestrian flowShopping mallsR=0.08BanksR=0.08HospitalTransferenceSmall shopsOfficeResidenceR=0.005R=0.045R=0.24R=0.025R=0.031Simple Linear Regression Model27Relevancy between functions and pedestrian flowMultiply Linear Regression ModelWe have to discus

21、s the autocorrelation between functions and Nach or IntegrationSUMMARY OUTPUT回歸統(tǒng)計Multiple R0.641625R Square0.411683Adjusted R Square0.357709標準誤差0.807338觀測值120方差分析dfSSMSFSignificance F回歸分析1049.715144.9715147.6274283.9E-09殘差10971.045580.651794總計119120.7607We are sure that functions has an effect on pe

22、destrian flow,But when we predict pedestrian flow, do we have to introduce the variable of function?28Relevancy between functions and pedestrian flowAutocorrelation verificationSUMMARY OUTPUT回歸統(tǒng)計Multiple R0.493071R Square0.243119Adjusted R Square0.173681標準誤差45.9149觀測值120方差分析dfSSMSFSignificance F回歸分析

23、1073811.787381.1783.5012120.000496殘差109229791.42108.178總計119303603.1We found that the functions ie autocorrelated with intrgration(R=1600)So we dont need to consider function when we predict the pedestrian flowWe can consider integration(r=1600) only in the pedestrian flow DNA 29SUMMARY OUTPUT回歸統(tǒng)計Mu

24、ltiple R0.601787R Square0.362148Adjusted R Square0.356877標準誤差0.723261觀測值123方差分析dfSSMSFSignificance F回歸分析135.9369535.9369568.6991931.8186E-13殘差12163.295820.523106總計12299.23277Coefficients標準誤差t StatP-valueLower 95%Upper 95%下限 95.0%上限 95.0%Intercept2.1650150.2865797.5546828.848E-121.59765562.7323741.59

25、76562.73237388X Variable 10.0105280.001278.2884981.819E-130.008013150.0130420.0080130.01304241R=0.362人流預測模型: The Relevance of Integration-R1600 and NACH-R1600After regression analysis, we finally determine to use Integration R1600 to predict the pedestrian flow in this district中山路地鐵周邊人流DNA=0.010258*

26、value（“Integration R1600”）+2.16501530The Influence Factors of Traffic flowCatalogueTopological ParameterRoad WidthComfort DegreeFurther Research about Pedestrian Flow Separate the influence of pedestrian from subway station FunctionsComparison with Different Subway StationsRevamping Options Visibili

27、ty graphAnalysis of SiteThe Average Walking DistanceCoverage area02Flow-widthWeek day peak valueWeek day averageWeekend averageWeekend peak value Regression analysis and thinking based on road widthRoad width（expressed by width）automobileflow（expressed by colour）02Regression analysis and thinking ba

28、sed on road width1200400800160024003200R=0.015 R=0.086 R=0.140 R=0.193 R=0.277 R=0.357 R=0.517n02Regression analysis and thinking based on road width1200400800160024003200R=0.059 R=0.113 R=0.124 R=0.201 R=0.549 R=0.612 R=0.549n02Regression analysis and thinking based on road widthIntegrationNach1200

29、400800160024003200nR=0.015 R=0.086 R=0.140 R=0.193 R=0.277 R=0.357 R=0.517R=0.059 R=0.113 R=0.124 R=0.201 R=0.549 R=0.612 R=0.54902Regression analysis and thinking based on road width&Nach-WidthR=nR=0.517Integration-WidthR=nR=0.517Test on autocorrelationAutocorrelationR=0.289Nach-IntegrationR=nNo su

30、perposition！02Regression analysis and thinking based on road widthR=0.481R=0.405Nach R-road widthIntegration R-road width (road width)0.25=1570.0167*Integration .86 (road width)0.25=0.4347*Integration+0.169944602Regression analysis and thinking based on road widthWidth of Road Flow Function of Archi

31、tectureinfuluenceinfuluenceDesitinationdecideMore shade Less vehicles Quieter Preference of narrow roadsdecide02Regression analysis and thinking based on road widthWidth of Road Flow Function of ArchitectureinfuluenceinfuluenceDesitinationdecideMore shade Less vehicles Quieter Preference of narrow r

32、oadsdecideComfort Evaluation02Regression analysis and thinking based on road widthThe Influence Factors of Traffic flowCatalogueTopological ParameterRoad WidthComfort DegreeFurther Research about Pedestrian Flow Separate the influence of pedestrian from subway station FunctionsComparison with Differ

33、ent Subway StationsRevamping Options Visibility graphAnalysis of SiteThe Average Walking DistanceCoverage area(1)Road Flatness(2)Pavement Capacity(3)Pedestrian Crossing Facilities Comfort(4)Pedestrian Flatness(5)Road Greening Rate(6)Leisure Space Humanization Design(7)Shade of Architecture(8)automob

34、ile Flow(9)Clean Degree(10)Safe Degree +03Regression analysis and thinking based on comfort degree1.Pavement Capacity 3.Shade of Architecture 1 Little Shade2 A Little Shade3 Much Shade2.Road Greening Rate1 Low Greening Rate2 Relatively High Greening Rate 3 High Greening Rate4.Clean Degree1 Dirty2 Re

35、latively Clean3 Clean2 Relatively High Capacity 3 High Capacity1 Low Capacity03Regression analysis and thinking based on comfort degreeAverage automobile flow-Peak automobile flowR=0.958weekday automobile flow-weekend automobile flowR=0.915Average automobile flow-Peak automobile flowR=0.958Average p

36、edestrian flow-Peak pedestrian flowR=0.935weekday pedestrian flow-weekend pedestrian flowR=0.851Average bike flow-Peak bike flowR=0.968weekday bike flow-weekend bike flowR=0.84403Regression analysis and thinking based on comfort degree Comfort evaluation-average automobile flowR=0.539X1：Pavement cap

37、acity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree03Regression analysis and thinking based on comfort degree Comfort evaluation-average automobile flowR=0.539X1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degreeThere is a negative correlation b

38、etween pavement capacity and automobile flow.Clean degree is the most relevant factor to automobile flow.Possible explanation：The wider the road, the cleaner the environment .03Regression analysis and thinking based on comfort degreeR=0.526X1：Pavement capacity X2：Road greening rate X3 : Shade of arc

39、hitecture X4 : Clean degree Comfort evaluation-average bike flow03Regression analysis and thinking based on comfort degreeR=0.526X1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree Comfort evaluation-average bike flowPavement capacity has a strong relation to bike

40、 flow, much stronger than other factors.03Regression analysis and thinking based on comfort degreeR=0.584X1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree Comfort evaluation-average pedestrian flow03Regression analysis and thinking based on comfort degreeR=0.584

41、X1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree Comfort evaluation-average pedestrian flowPavement capacity has a strong relation to pedestrian flow, much stronger than other factors.There is a negative relation between clean degree and pedestrian flow.Possibl

42、e explanation: This is because of the vegetable market in the site, which is of low clean degree but with large pedestrian flow.03Regression analysis and thinking based on comfort degreeR=0.537R=0.302R=0.321With road widthWith Nach RWith Integration 3200Weight The trend of bike flow and pedestrian a

43、re similar.The two lines of automobile flow and bike flow show opposed status.There is a negative correlation between clean degree and pedestrian flow.Clean degree is only highly related to automobile flow.multiple regression analysis on flow without clean degree03Regression analysis and thinking ba

44、sed on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of architecture R=0.092(was 0.539) Comfort evaluation （without clean degree）-average automobile flow03Regression analysis and thinking based on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of architectu

45、re R=0.092(was 0.539) Comfort evaluation （without clean degree）-average automobile flowautomobile flow is not related to pavement capacity, road greening rate or shade of architecture03Regression analysis and thinking based on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of ar

46、chitecture X4 : Clean degree R=0.458(was 0.526) Comfort evaluation （without clean degree）-average bike flow03Regression analysis and thinking based on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree R=0.458(was 0.526) Comfort evaluation （without

47、clean degree）-average bike flowAmong the four factors of comfort evaluation, pavement capacity is a decisive factor to bike flow.03Regression analysis and thinking based on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree R=0.596(was 0.584) Comfor

48、t evaluation （without clean degree）-average pedestrian flow03Regression analysis and thinking based on comfort degreeX1：Pavement capacity X2：Road greening rate X3 : Shade of architecture X4 : Clean degree R=0.596(was 0.584) Comfort evaluation （without clean degree）-average pedestrian flow Among the

49、four factors of comfort evaluation, pavement capacity is a decisive factor to pedestrian flow.03Regression analysis and thinking based on comfort degreeLogging Data Correlation Analysis Conclusion1.List influential factor2.Screening according to importance and maneuverability1.Four factors correlati

50、on analysis2.Three factors (withoutClean Degree) correlation analysis fort degree is almost not correlative to automobile flow, while strongly correlative to pedestrian as well as bike flow.2.Among the four factors that taken into account, clean degree is nearly not related to bike or pedestrian flo

51、w.3.Among the four factors that taken into account, pavement capacity is the most relative factor to bike and pedestrian flow. Summary03Regression analysis and thinking based on comfort degreeThe Influence Factors of Traffic flowCatalogueTopological ParameterRoad WidthComfort DegreeFurther Research

52、about Pedestrian Flow Separate the influence of pedestrian from subway station FunctionsComparison with Different Subway StationsRevamping Options Visibility graphAnalysis of SiteThe Average Walking DistanceCoverage area500mR=1000mNPath TrackingBACRegression analysis and thinking based on path track

53、ing01Path Tracking60Most destinations are transference and service industryMost destinations are residenceMost destinations are residenceRelevancy vs nachR=0.07Relevancy vs Angular Step DepthR=0.25Relevancy vs Metric Step DepthR=0.40Relevancy AnalysisRelevancy vs integrationR=0.09500mNLow passing ra

54、teHigh passing rate61Relevancy AnalysisRelevancy vs Metric Step DepthR=0.49Metric Step DepthVs Angular Step DepthR=0.4Relevancy vs Angular Step DepthR=0.26Relevancy vs Metric Step Depth& Angular Step DepthR=0.5862Relevancy AnalysisRestaurantBIG DATAShopping mallsHotelsR=0.0323R=0. 0996R=0.199463Tran

55、sferenceEducationHospitalsBanksRelevancy AnalysisR=6.32189e-005R=0.122618R=0. 00404484R=0. 1174764Relevancy AnalysisShopsBIG DATAOfficeResidentialentranceR=0.0684017R=0. 00143563R=0.0065371265Relevancy AnalysisRelevancy vsFunctionR=0.39Restaurant EducationHotelsShopping mallsBanksHospitalsTransferen

56、ceShopsOfficeResidentialentranceFunction=0.065318221*Restaurant+0.071495699*Education-0.017822371*Hotels +0.527148111*Shopping malls+0.175788615*Banks+0.052079119*Hospitals +0.344863152*Transference-0.004134893*Shops+0.010209514*Office -0.032750532*Residential entrance66Relevancy Analysis多元分析Functio

57、n vs Angular Step DepthR=0.0342454Metric step depth vs Angular Step DepthR=0. 1139Metric step depth vs FunctionR=0.237048AutocorrelationAngular Step Depth Metric Step Depth地鐵站釋放人流(Subway distribution prediction )DNA=-0.25872*value（Angular Step Depth）-0.00087*value（Metric Step Depth ）+1.528017 67Sepa

58、rate The Influence of station from the parameters which influence pedestrian flowExit-CExit-AExit-B01Time:9:0010:00Time:14:0015:00Time:17:0018:00Exit-CExit-AExit-BExit-CExit-AExit-BSum: 960 / hAverage pedestrian flow: 1196 / hSum: 1220 / hSum: 1407 / h68Separate The Influence of station from the par

59、ameters which influence pedestrian flow01DNA（Per hour）= (1196/400) * DNA(Per 400 persons) =-0.7735728*value（Angular Step Depth）-0.0026013*value（Metric Step Depth ）+4.56877083DNA（Per hour) = (Average pedestrian flow from station /400 )*Subway distribution prediction model (400 people) Transference fr

60、om Subway distribution prediction (per 400 people) to Subway distribution prediction (per hour) 69Separate The Influence of station from the parameters which influence pedestrian flow01Simulated flow From stationSimulated all the flowInfluence =-0.7735728*value（Angular Step Depth）-0.0026013*value（Me