THURSTON REGION MULTIMODAL TRAVEL DEMAND FORECASTING MODEL IMPLEMENTATION IN EMME/2 - Presentation at the 15th International EMME/2 Users Group Conference.
Post on 26-Mar-2015
217 Views
Preview:
Transcript
THURSTON REGION MULTIMODAL THURSTON REGION MULTIMODAL TRAVEL DEMAND FORECASTING MODEL TRAVEL DEMAND FORECASTING MODEL
IMPLEMENTATION IN EMME/2IMPLEMENTATION IN EMME/2
- Presentation at the 15th International - Presentation at the 15th International EMME/2 Users’ Group ConferenceEMME/2 Users’ Group Conference
Oct. 18, 2000Oct. 18, 2000
Jin Ren, PE, Transportation EngineerJin Ren, PE, Transportation Engineer
Thurston Regional Planning CouncilThurston Regional Planning Council
Olympia, WA (www.trpc.org)Olympia, WA (www.trpc.org)
TRPC Technical Modeling Process
T h u rs to n C o u n ty T ra ve l S u rve ys M u lt im o d a l N e tw o rk B u ild ing
V e h ic le T rip C a lib ra tion
M u lt i-C la ss A u to A ss ig n m e n ts(B y T im e P e rio d s)
T ra n s it T rip C a lib ra tio n
T ra n s it P e rso n T rip A s s ig n m e n ts(B y T im e P e rio d s)
T im e o f D a y M o d e ls
D a ily M o d e C h o ic e s by P u rp o ses
D a ily T rip D is trib u tio n b y P u rp o ses
T rip G e n e ra tion
H o u se ho ld S u b-M o d e ls
Travel “Skims” Data Travel “Skims” Data PreparationPreparation
EMME/2 Multimodal Network BuildingEMME/2 Multimodal Network Building Travel Time and Distance by Modes Travel Time and Distance by Modes
(Walking/Biking/Auto/Transit) (Walking/Biking/Auto/Transit) Intrazonal Travel Time and DistanceIntrazonal Travel Time and Distance Distance-Based Housing/Employment Distance-Based Housing/Employment
Density by Traffic Analysis Zones (TAZ)Density by Traffic Analysis Zones (TAZ) Travel Time-Based Transit AccessibilityTravel Time-Based Transit Accessibility Mix Use Index (Area-Based Densities)Mix Use Index (Area-Based Densities) Area-based Local Intersection DensityArea-based Local Intersection Density
Household Sub-ModelsHousehold Sub-Models(Multinomial Logit Choice Modeling)(Multinomial Logit Choice Modeling)
Household Worker (0, 1, 2, 3+)Household Worker (0, 1, 2, 3+) Household K-12 Schoolchild (0, 1, 2, 3+)Household K-12 Schoolchild (0, 1, 2, 3+) Household Auto-Ownership (0, 1, 2, 3+)Household Auto-Ownership (0, 1, 2, 3+)
Trip Generation ModelsTrip Generation Models
Cross-classified Household Trip RatesCross-classified Household Trip Rates: : 1998/1999 Household Travel Survey1998/1999 Household Travel Survey
Truck Freight Trip ModelTruck Freight Trip Model: : 1997 Riebee Freight Survey Data1997 Riebee Freight Survey Data
External Trip Generation ModelExternal Trip Generation Model: : 1997 I-5/SR-101 O-D Surveys and 1997 I-5/SR-101 O-D Surveys and Vehicle Classification CountsVehicle Classification Counts
Household Cross-Classification Household Cross-Classification Schemes for Trip ProductionSchemes for Trip Production
1998 Daily Trip 1998 Daily Trip ProductionProduction Calibration Calibration
Trip PurposeBefore-
CalibrationProductions
After-Calibration
Productions
Model TripProductionDistribution
ExpandedSurvey Trip
Production %HB-Work 127,785 140,564 17.7% 18.0%HB-Other 218,925 240,818 30.4% 29.6%
HB-Shopping 72,238 79,462 10.0% 10.2%HB-School 52,394 57,633 7.3% 7.4%HB-College 7,197 7,917 1.0% 1.1%Work-Other 85,970 94,567 11.9% 12.1%Other-Other 156,510 172,161 21.6% 21.6% Daily Total 721,019 793,122 100.0% 100.0%
AverageTrips/Household
8.99 9.89 8.66
Note: Trip production expansion factor is found to be 1.10.
Daily Destination Choice ModelDaily Destination Choice Model(Multinomial Logit Models with Size Variables)(Multinomial Logit Models with Size Variables)
O-D Travel Time from Auto O-D Travel Time from Auto Assignments Assignments
1998 Households, Employees by Retail, 1998 Households, Employees by Retail, Office, Service, Government and OtherOffice, Service, Government and Other
The standard formula for utilities is: The standard formula for utilities is:
Utilij= exp(*timeij+*timeij2+*timeij
3+ln(1…k*Employmentj1...jk
+ j*Householdsj))
WhereWhere
, , , , , , and and are parameters or estimated coefficients are parameters or estimated coefficients
1…k stand for different employment sectors 1…k stand for different employment sectors
i represents a ‘production’ TAZ i represents a ‘production’ TAZ
j represents an ‘attraction’ TAZj represents an ‘attraction’ TAZ
Daily Mode Choice Modeling Daily Mode Choice Modeling
Drive-Alone Vehicle or Person Drive-Alone Vehicle or Person TripsTrips
Drive-with-Passenger Vehicle or Drive-with-Passenger Vehicle or Person TripsPerson Trips
Passenger-Only Person TripsPassenger-Only Person Trips Transit Person TripsTransit Person Trips Walk Person TripsWalk Person Trips Bike Person TripsBike Person Trips
Variables Impacting Mode Variables Impacting Mode ChoicesChoices
(Multinomial Logit Choice Modeling)(Multinomial Logit Choice Modeling) Land Use Variables (XLand Use Variables (Xii)): Employment Density, : Employment Density,
Transit Accessibility, Mixed-Use, & Parking Cost Transit Accessibility, Mixed-Use, & Parking Cost Household Variables (YHousehold Variables (Yjj)): Household Size, Auto-: Household Size, Auto-
Ownership, Worker Size and Income StatusOwnership, Worker Size and Income Status Network Skims Variables (ZNetwork Skims Variables (Zkk)): Local : Local
Intersection Density and Point-to-Point Travel Intersection Density and Point-to-Point Travel Time Time
The standard logit utility function:The standard logit utility function:
Utilij= exp( +i*Xi+j*Yj+k*Zk)
Where Where , , , , , and , and are parameters or estimated coefficients are parameters or estimated coefficients
Time-of-Day ModelsTime-of-Day Models
Production-Attraction and Attraction-Production-Attraction and Attraction-Production Peaking Factors (Time-of-Day Production Peaking Factors (Time-of-Day Factors)Factors)
1998 AM Peak Hour Trip Tables by Modes1998 AM Peak Hour Trip Tables by Modes 1998 Mid-Day Hour Trip Tables by Modes1998 Mid-Day Hour Trip Tables by Modes 1998 PM Peak Hour Trip Tables by Modes1998 PM Peak Hour Trip Tables by Modes Add 1998 Inbound/Outbound/ Through Add 1998 Inbound/Outbound/ Through
Vehicle Trips for AM, Mid-day and PM HoursVehicle Trips for AM, Mid-day and PM Hours
Trip AssignmentsTrip Assignments
1998 Multi-Class Auto 1998 Multi-Class Auto Assignments by Time Periods Assignments by Time Periods
1998 Transit Person Trip Multi-1998 Transit Person Trip Multi-Path Assignments by Time PeriodsPath Assignments by Time Periods
Feedback and Looping Process to Feedback and Looping Process to Reach Ideal EquilibriumReach Ideal Equilibrium
Model Calibration ProcessModel Calibration Process
Goodness-of-Fit Statistical TestingsGoodness-of-Fit Statistical Testings Control Total or Percentage Checks: Control Total or Percentage Checks:
- Household Numbers - Household Numbers - Trip - Trip Productions - Productions - Mode Splits Mode Splits - Average Vehicle Occupancies - Average Vehicle Occupancies
Screenline Analysis by 18 ScreenlinesScreenline Analysis by 18 Screenlines Transit Ridership Calibration to 1998 Transit Ridership Calibration to 1998
Transit Ridership SurveysTransit Ridership Surveys
In ConclusionIn Conclusion For the first time, our region is developing a For the first time, our region is developing a
multimodal travel demand forecasting modelmultimodal travel demand forecasting model For the first time, we are using local survey For the first time, we are using local survey
data to develop a regional model data to develop a regional model Model estimation and application hand in handModel estimation and application hand in hand Peer review groups and documentationPeer review groups and documentation
Effective integration of software tools for Effective integration of software tools for
data preparation and analysis in housedata preparation and analysis in house
Robust EMME/2 Matrix Manipulation and Robust EMME/2 Matrix Manipulation and
MacrosMacros
top related