urban transport modeling (based on these two sources)

56
Urban Transport Modeling (based on these two sources) A Transportation Modeling Primer May, 1995 Edward A. Beimborn Center for Urban Transportation Studies University of Wisconsin-Milwaukee http://www.uwm.edu/ Dept/CUTS/primer.htm Modelling Transport: Theory and Practice Frank McCabe, Dublin Transportation Office http:// www.icetact.tcd.ie/ icetact/news/ transport/mccabe.html

Upload: danyl

Post on 05-Jan-2016

27 views

Category:

Documents


0 download

DESCRIPTION

A Transportation Modeling Primer May, 1995 Edward A. Beimborn Center for Urban Transportation Studies University of Wisconsin-Milwaukee http://www.uwm.edu/Dept/CUTS/primer.htm. Modelling Transport: Theory and Practice Frank McCabe, Dublin Transportation Office - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Urban Transport Modeling (based on these two sources)

Urban Transport Modeling(based on these two sources)

A Transportation Modeling Primer

May, 1995

Edward A. BeimbornCenter for Urban Transportation StudiesUniversity of Wisconsin-Milwaukee

http://www.uwm.edu/Dept/CUTS/primer.htm

Modelling Transport: Theory and PracticeFrank McCabe, Dublin Transportation Office

http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html

Page 2: Urban Transport Modeling (based on these two sources)

What’s in the Black Box?

Beimborn, Edward A. 1995. A Transportation Modeling Primer.Center for UrbanTransportation Studies, University of Wisconsin-Milwaukee: http://www.uwm.edu/Dept/CUTS/primer.htm (last accessed 11/14/04).

Page 3: Urban Transport Modeling (based on these two sources)

Travel Demand Modeling

First developed in the late 1950s

Helps make decision on future development of (urban) transport systems

Forecasts travel patterns 15-25 years ahead

Model is predictive, but used prescriptively

A clear understanding of the modeling process and assumptions is essential to understanding transportation plans.

Page 4: Urban Transport Modeling (based on these two sources)

Beimborn, Edward A. 1995. A Transportation Modeling Primer.Center for UrbanTransportation Studies, University of Wisconsin-Milwaukee: http://www.uwm.edu/Dept/CUTS/primer.htm (last accessed 11/14/04).

Page 5: Urban Transport Modeling (based on these two sources)

Limitations of Urban Transport Modeling

Only considers factors and alternatives explicitly included in the equations.If models are not sensitive to certain factors, they will not show any effect of them.This could lead to a conclusion that the factors are ineffective.E.g., bicycle or pedestrianIt is therefore critical to consider the assumptions before decisions are made.

Page 6: Urban Transport Modeling (based on these two sources)

Define problem

Define goals and criteria

Collect data

Forecasting (modeling)

Develop alternatives

Evaluate

Finalize an implementation plan

How do models fit in the transport planning process?

Page 7: Urban Transport Modeling (based on these two sources)

How is Travel Modeled?1. What will our community look like in

the future? How many people? (population forecasts) What will they do? (economic forecasts) Where will they do it? (land use pattern)

2. What are the travel patterns in the future? How many trips? (trip generation) Where will the trips go? (trip distribution) What modes will they use? (mode split) What routes will they take? (traffic assignment) What will be the effects of this travel? (impact analysis)

Beimborn, Edward A. 1995. A Transportation Modeling Primer.Center for UrbanTransportation Studies, University of Wisconsin-Milwaukee: http://www.uwm.edu/Dept/CUTS/primer.htm (last accessed 11/14/04).

Page 8: Urban Transport Modeling (based on these two sources)

Population Forecasts

Birth rates

Death rates

Migration rates

Ages

Often use forecasts from other agencies

Page 9: Urban Transport Modeling (based on these two sources)

Economic Forecasts

Employment levelsForecasted in conjunction with population Basic employmentEconomic multipliers used to estimate nonbasic employmentOften use forecasts from other agencies

Page 10: Urban Transport Modeling (based on these two sources)

Land Use Forecasts

Allocate population and economic growth

1. Establish land use goals and land use rates

2. Allocate to specific locations—models can be used to predict nonbasic and residential from basic

Page 11: Urban Transport Modeling (based on these two sources)

Land Use Forecasting:Limitations

No feedback with transportation plans

Current development is fixed – considers only vacant land

Mixed-use benefits not considered

Page 12: Urban Transport Modeling (based on these two sources)

Existing Land Use of Metro Manila (1996)

Page 13: Urban Transport Modeling (based on these two sources)

Study Area

The area within which trip patterns will be significantly affected by the implementation of transport proposals.

Page 14: Urban Transport Modeling (based on these two sources)

Aggregate Modeling

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 15: Urban Transport Modeling (based on these two sources)

Zoning System of Greater Metro Manila

Page 16: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 17: Urban Transport Modeling (based on these two sources)

Using EMME for the UTP of Iloilo City: An Example

Node

Link

CPU

SM City

from/to Guimaras

from/to South

from/to Airport

Page 18: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 19: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 20: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 21: Urban Transport Modeling (based on these two sources)

Trip Generation Example: Iloilo City

Page 22: Urban Transport Modeling (based on these two sources)

Trip Generation

Done separately for each trip purpose

Two main methods:Multiple regressionCross-classification (a.k.a. category

analysis)

Page 23: Urban Transport Modeling (based on these two sources)

Trip Generation byMultiple Regression, e.g.:

Home-to-work trips (R = 0.99)Oi = -43.6 + 0.097 Population + .773 Employed

Residents - 351 Number of Households + .504 Num Cars Trips to shopping (R = 0.95)Oi = -17.9 + 1.19 Area Res. Land + .266 Number of

Cars Light Commercial Vehicle Trips:Oi = 75.9 + .367Number HHs + .267 Total Empl -.339

Office Empl - .0188 Industrial Empl

Page 24: Urban Transport Modeling (based on these two sources)

Trip Generation byCross-Classification

Page 25: Urban Transport Modeling (based on these two sources)

  Advantages Disadvantages

Multiple Regression

1. Familiar methodology2. Statistical significance3. Estimates effect holding other variables constant

1. Assumes linearity2. Aggregate data (ecological fallacy)3. Coefficients not stable over time, or after improvements4. Multicollinearity

Category Analysis

1. Individual data good for predicting individual behavior. 2. Interaction effects 3. Doesn’t assume linearity 

1. Needs individual data2. Hard to resurvey individuals for more variables3. Best with naturally discrete variables.

Page 26: Urban Transport Modeling (based on these two sources)

Common Limitation of Trip Generation Models

Independent decisions

Limited trip purposes

Limited independent variables

Trip-chaining is ignored

Lacks feedback with trip distribution, modal split, traffic assignment

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 27: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 28: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 29: Urban Transport Modeling (based on these two sources)

O-D Matrix (12 zones, Iloilo City)

Page 30: Urban Transport Modeling (based on these two sources)

Trips Attracted (from other zones ) to (Zone 12 – CPU and adjacent areas)

Page 31: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 32: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 33: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 34: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 35: Urban Transport Modeling (based on these two sources)

Common Limitations of Trip Distribution Models

Constant trip lengths (in minutes) into the futureUse of car travel times in gravity model denominatorIgnores socio-cultural-economic factors for O-D pairsLacking feedback with trip generation, modal split, network congestion

Page 36: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 37: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 38: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 39: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 40: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 41: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 42: Urban Transport Modeling (based on these two sources)

Common Limitations of Modal Split Models

Mode choice affected only by time and cost characteristics

Omitted factors (amenities, crime, safety, security) have no effect

No consideration of ease of walking and comfort of waiting for transit

Page 43: Urban Transport Modeling (based on these two sources)

Modal Split Ignores Pedestrian Friendliness

Beimborn, Edward A. 1995. A Transportation Modeling Primer.Center for UrbanTransportation Studies, University of Wisconsin-Milwaukee: http://www.uwm.edu/Dept/CUTS/primer.htm (last accessed 11/14/04).

Page 44: Urban Transport Modeling (based on these two sources)

Traffic Assignmentalso known as

Network assignment

Route assignment

Network loading

Trip assignment

Page 45: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 46: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 47: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 48: Urban Transport Modeling (based on these two sources)

Wardrop Equilibrium"Under equilibrium conditions, traffic arranges itself in congested networks in such a way that no individual trip maker can reduce his path costs by switching routes".

All used routes between any Origin / Destination pair have equal and minimum costs, while all unused routes have greater or equal costs."

Page 49: Urban Transport Modeling (based on these two sources)

Algorithms

A mechanistic iterative procedure with a:

Starting rule

Iteration rule

Stopping rule

Page 50: Urban Transport Modeling (based on these two sources)

Wardrop Equilibrium Algorithm

Starting rule: set all arc flows to 0, compute all arc costs from cost-flow curve.

Iteration (for each O-D pair):Use shortest path algorithm to find the cheapest route for

each O-D pairShift a fraction of the flow from the old route(s) to the

new cheapest route.Recalculate the new arc flows and arc costs.

Check for convergence:

high-cost route – low cost route < (i.e., tolerance)

Page 51: Urban Transport Modeling (based on these two sources)

Common Limitations of Traffic Assignment Models

Intersection delay is ignoredTravel only on simplified networkIntrazonal travel ignored (affects pollution estimates)Capacities are simplified as a function of number of lanes and type of roadModel ignores time of day. Peak-hour adjustment factor is critical.Peak spreading is not considered.

Page 52: Urban Transport Modeling (based on these two sources)

Final Outputs from Traffic Assignment

Amount of travel, congestion, speed per link

Total transport cost, time, VMT

By applying other coefficients:AccidentsAir pollution

Page 53: Urban Transport Modeling (based on these two sources)

Improving UTMSBetter data

Biking and walking

Auto occupancy

Time of day factors

More trip purposes

Transit-friendly and bike/walk-friendly design

Land use feedback

Intersection delays

Page 54: Urban Transport Modeling (based on these two sources)

Other Critiques to UTMS

Ignores non-transport impacts (energy, neighborhoods, etc.)

Sequential 4-step model doesn’t always match transport decision-making process.

Ecological fallacy (of aggregate data)

Page 55: Urban Transport Modeling (based on these two sources)

McCabe, Frank. Modelling Transport: Theory and Practice. Dublin Transportation Office: http://www.icetact.tcd.ie/icetact/news/transport/mccabe.html (last accessed 11/14/04).

Page 56: Urban Transport Modeling (based on these two sources)

Disaggregate Methods

Parcel-based microsimulation

Household Activity Travel Simulation (HATS)