Transportation Planning, Transportation Demand Analysis

1
Transportation Planning, Transportation Demand
Analysis

• Land Use-Transportation Interaction
• Transportation Planning Framework
• Transportation Demand Analysis

2
Land Use-Transportation Interaction
Change in Land use
Transportation serves land uses
Change in Trip generation
Land Values
Accessibility
Change in travel needs
Change in transportation supply (added
services facilities)
Transportation shapes land uses
3
Land Use-Transportation-Environment Interaction

• Land use
• Transport
• Environment

Urban Area
.
Zones
.
Change in land use over time (i.e. change in
residential units, commercial land use,
industrial land use, retail land use, etc.
4
Land Use Patterns, Bid Rent

• Pressure for growth
• Demand for land Bid
  rent
• Land use pattern
  Location of
• 
  activities

Bid rent /sq.km
CBD
Distance from CBD
Jobs
Population
Distance
CBD
5
Purpose of Land Use Models

• To explain/predict
• Change in land use as a function of
• - accessibility to employment
• - land value
• - percent of urban level available vacant land in
  a zone
• - public transit accessibility
• - quality of water sewer services
• - etc..

6
Modelling Travel Decisions

• User Decisions
• 1. To travel (for a given trip purpose at a
  given time)? (Trip generation)
• 2. Destination? (Trip distribution)
• 3. Mode? (Modal Choice)
• 4. Route? (Assignment of trip to network)
• Modelling Approaches
• Four-stage urban transportation modelling system
  (UTMS)
• Unified approaches

7
Urban Transportation Demand Modelling Four-
Stage Modeling System
Population Employment Forecasts
Trip Generation
Trip Distribution
Transportation Network Service Attributes
Modal Split
Trip Assignment
Link O-D Flows, Times, Costs, Etc.
8
Four Stages of Urban Travel Demand Modelling
I
J
Trip Generation
Dj
Oi
Trip Distribution Tij
J
I
J
Tij,auto
I
J
Mode Split
Tij, transit
Traffic assignment
I
J
Path of flow Tij,auto through the auto network
9
Multiple Trip Purposes
Population Employment
HW HS
NWS Generation Generation
Generation Distribution Distribution
Distribution Modal Split Modal Split Modal
Split Road Assignment Transit Assignment

Trip Rates, etc.
Transport Network
Link O-D volumes, times, costs, v/c
ratios, etc.
10
The Traffic Prediction Process
Trip
generation P A Transit network
Road network
Trip
distribution
Modal split Transit person trips
Auto person
trips Occupancy
Occupancy
Transit vehicle trips
Auto vehicle trips
Freight other vehicles
Transit traffic assignment Road
traffic assignment
11
Trip Generation

• Modelling Methods
• Linear regression method
• Cross-classification (category analysis)
  method/trip rate method
• __________________________________________________
  _____
• Trip generation
• Productions Attractions
• Home-based non-home based
• trips

J
I
Zones
12
Trip Productions Attractions
Pi Trip productions of zone i f(land use,
socio-economic characteristics of zone i) Aj
Attractions of zone j f(land use,
socio-economic characteristics of zone
j) Regression Model Examples (P.M. Peak Period
Work Trips) Pi 0.4572 emp - 138 (R2
0.87) Aj 0.1848 pop 9 (R2 0.90) Where
emp is total employment
pop is total population
13
Trip Productions Attractions (Continued)
Regression Model Examples (P.M. Peak Period
Non-work trips) Pi 0.1346pop0.2897emp0.0043GLA
(R2 0.76) Aj 0.0888emp
0.6204DWEL0.0045GLA221 (R2 0.80) Where
emp is total employment pop is
total population GLA shopping
centre gross leasable area (ft2)
DWEL Dwelling units
14
Trip Productions Attractions (Continued)
RegressionModel Development Data Required Zone
Pi Aj pop emp GLA
DWEL 1 . . . .
. .. 2 . . .
. . .. . _____________________
________________________ from O-D survey Data
on other variables obtained from city data base

15
Trip Productions Attractions (Continued)
RegressionModel Development (Continued) Check on
- Partial correlation coefficient (r) .
Should be high between P (the dependent variable)
other variables (the independent variables)
Should be high between A (the dependent variable)
other variables (the independent variables)
. Should be low between pop, emp, GLA, DWEL (I.e.
between independent variables) - Other
statistical measures (t statistic for each
independent variable)
16
Trip Productions Attractions (Continued)
RegressionModel Development (Continued) Check on
- R Multiple correlation coefficient (max.
value of 1.0) - R2 Coefficient of multiple
determination (max. value of 1.0) - Standard
Error of Estimate (for the dependent variable -
e.g. for Pi) Its value can be checked against the
estimated values of the dependent
variable. Example A range of Pi values
1,000-5,000 St. Error of 100 (very low!)
17
Trip Productions Attractions (Continued)
Trip Generation Rates (Cross Classification
Approach) Trip Production Step 1
Family Size Auto
Ownership
0 1 2 or more 1
Trips/household/day 2 3 4 or
more
18
Trip Productions Attractions (Continued)
Trip Generation Rates (Cross Classification
Approach) Trip Production Step 2 Trip
productions for Zone i (Trips/household/day) x
(No. of households of that classification). Trips/
household/day is based on O-D survey No of
households of a given classification to be
forecasted.
19
Trip Distribution Models

• Many models most common is gravity model
• Zone i
• Pi

Zone j Aj
Tij
Zone j Aj
Zone j Aj
20
Trip Distribution Models

• Origin-Constrained Gravity Model

Aj Fij Kij
Tij Pi

S for j(Aj Fij Kij)
Where Tij Trips produced in zone I and
attracted to zone j Pi Trips produced by zone
i Aj Trips attracted to zone j Fij Impedance
of travel from zone I to zone j (a travel time
factor -- expressing an area-wide effect of
distance) Kij A zone-to -zone adjustment factor
21
Trip Distribution Models

• Destination-Constrained Gravity Model

Pi Fij Kij
Tij Aj

S for i(Pi Fij Kij)
Where Tij Trips produced in zone I and
attracted to zone j Pi Trips produced by zone
i Aj Trips attracted to zone j Fij Impedance
of travel from zone I to zone j (a travel time
factor -- expressing an area-wide effect of
distance) Kij A zone-to -zone adjustment factor
22
Gravity Model

• The Fij is usually a some function of the travel
  time or generalized cost of travel between zones
• Fij C-a ij or Fij t-a ij
  Fij

Where a is the calibration constant Fij
Travel time factor C ij Generalized cost
function t ij Travel time Kij A
zone-to-zone adjustment factor (takes into
account special characteristics of ij
combinations
tij or Cij
Example
River
Zone 1
Zone 2
23
Gravity Model

• Note
• Pi S for j Tij
• Aj S for i Tij

Pi
Aj
24
Gravity Model

• Example
• Using a gravity model with an impedance term of
  the form C-a , estimate the number of of trips
  from zone 1 to all other zones. a 1.80. Other
  inputs are shown below.

Zone Travel time to zone 1 (min)
Productions Attractions 1
-- 5000
1000 2 10
2000
4000 3 20
4000 5000 4
15
3000 4000 _______________________
___________________________
25
Gravity Model

• Here, Pi for i zone 1 are to be distributed to
  other zones by using the gravity model. Assume
  all K 1
• For a 1.80 and given travel times Cij,, and
  Aj, we find
• ______________________________________________
• Zone Aj Cij C-a
  AjC- a ij Tij
• 1 1000 -- --
  -- --
• 2 4000 10 1/63.1
  63.40 2716
• 3 5000 20 1/219.7
  22.76 975
• 4 4000 15 1/130.91
  30.56 1309
• 
  Sum 116.72 5000
• T from 1 to 2 5000(63.40/116.72) 2716

26
Gravity Model

• Following iteration 1 of finding Tij from every
  zone to all zones, check to see if Ajs match
  the known values
• If yes, the trip distribution problem is solved.
• If not, the Ajs have to be adjusted.
• The adjustment process is an iterative one (not
  covered here)