Anda di halaman 1dari 32

The SANDAG Activity

Activity--Based Travel Model

Institute of Traffic Engineers Luncheon Presentation


April 1, 2010

Trip-Based Model Overview

Trip Generation

Trip Distribution

Mode Choice

Trip Assignment

Trip Productions (Pi)


Zone

Pi

34

66

Trip Attractions (Aj)


Zone

Aj

82

18

From
Zone

To Zone
1

27

55

Total
Aj

82

Mode

Trips
(Tijm)

Route

Trips
(Tijmr)

34

Auto

30

Route A

18

11

66

Transit

25

Route B

18

100

Total
Pi

Institute of Traffic Engineers Luncheon


April 1, 2010

Criticisms of Trip-Based Models


Lack of Behavioral Fidelity
The only model based on actual decision-making theory is mode
choice (in most model systems)

Aggregation Bias
No information on non-home-based trips
Each additional market segment (socio-economic category, trip
purpose, time period) significantly increases runtime

Lack of Policy Responsiveness


Time-of-day shifts
Socio-economic changes
Induced travel

Institute of Traffic Engineers Luncheon


April 1, 2010

What is an Activity-Based Travel Model?


Travel is a derived demand it results for the need for
people to engage in activities outside the home.
Activity-based travel models attempt to replicate how people
decide:

whether to travel
where to travel to
when to travel
how to travel

Activity-based travel models are based on behavioral


decision-making theory
This makes them more suited to address policies that affect
how people make travel decisions than trip-based models
Institute of Traffic Engineers Luncheon
April 1, 2010

Activity-Based Travel Models


Model travel by individuals
All important socio-economic characteristics are tracked
explicitly
Micro-simulation used

Model trips as part of tours


A series of trips beginning and ending at home or work (anchor
locations)
Primary destination, intermediate stops
No more non-home-based trips!!

Schedule tours into available time-windows


Consistent daily activity patterns that replicate survey data

Institute of Traffic Engineers Luncheon


April 1, 2010

Complex Planning Issues Addressed with


AB Models

Affects of transport policies on time-use


Demographic changes & equity analysis
Demand management policies (HOV, pricing)
Changes in accessibilities
Better interface with traffic simulation models
Telecommuting
Global transportation policies & taxation
Evacuation planning
Institute of Traffic Engineers Luncheon April 1,
2010

ABM: Tours and Trips


Work-Based Tour
Zone 1

Zone 3

Home-Based Work Trip

Non-Home-Based Trip

Non-Home-Based Trip

Work Tour
Origin
Home-Based
Other Trip

Primary
Origin
Non-Home-Based
Destination
Trip

Zone 2

Zone 4

Primary
Destination

Intermediate
Stop

Data View:
HH #

Per #

Tour #

Purp

Origin
TAZ

Destin.
TAZ

Outbound
Stop1 TAZ

Return
Stop1 TAZ

Mode

Subtour

Sub-Tour
Destin.

1023

Work

Transit

Yes

Institute of Traffic Engineers Luncheon


April 1, 2010

Activity-Based Models: Mode Consistency


Work Tour
Zone 1

Work-Based Tour
Zone 3

Zone 2

Bus to Work = Drive alone not available for lunch

Institute of Traffic Engineers Luncheon


April 1, 2010

Zone 4

Activity-Based Model: Micro-simulation


A synthetic population is created that represents the
actual population
Travel is explicitly modeled for each person/household
Monte Carlo simulation is used instead of fractional
probability aggregation: Discrete choices made for each
traveler
Results are aggregated and:
Assigned to transport networks
Compiled into reports

Institute of Traffic Engineers Luncheon


April 1, 2010

Activity-Based Model For San Diego


Based on the CT-RAMP (Coordinated Travel Regional
Activity-based Modeling Platform) Family of Activity-Based
Travel Demand Models

Main features:
Explicit intra-household interactions
Continuous temporal dimension (Half-hourly time
periods)
Logit formulations for choice models
Sensitive to a wide range of socio-economic variables,
transportation costs/accessibilities, and land-use
changes
Java-based package for model implementation
Institute of Traffic Engineers Luncheon
April 1, 2010

Activity-Based Models In the United States


Seattle
Oregon
Ohio
San
Francisco

Sacramento
Lake Tahoe

NY
Columbus

Denver

Bay Area

Los Angeles
San Diego

Phoenix
Atlanta

Developed by PB
Developed by others
CT-RAMP Family

(Atlanta model co-developed with MTC)


Institute of Traffic Engineers Luncheon
April 1, 2010

1. Population Synthesis
2. Long-term

3. Mobility

2.1. Usual workplace / school

3.1. Free Parking Eligibility

4. Daily

3.2. Car ownership

3.3. Transponder Ownership

4.1. Person pattern type & Joint Tour Indicator


Nonmandatory

Mandatory

Home

Joint
(household level)

Available
time budget

Residual time

Individual
Mandatory Tours

Joint NonMandatory Tours

Allocated Tours

4.2.1. Frequency
4.2.2. TOD

At-work sub-tours

4.3.1. Frequency

4.4.1. Frequency

4.3.2. Party

4.4.2. Allocation

Individual
Discretionary
Tours
4.5.1. Frequency

4.3.3. Participation

4.6.1. Frequency
4.6.2. Destination
4.6.3. TOD

4.3.4. Destination

4.4.3. Destination

4.5.2. Destination

4.3.5. TOD

4.4.4. TOD

4.5.3. TOD

5. Tour level
5.1. Tour mode

6. Trip level

5.2. Stop frequency

5.3. Stop location

5.4. Stop Departure

6.1. Trip mode


6.2. Auto parking
6.3. Assignment

Institute of Traffic Engineers Luncheon


April 1, 2010

12

12

A relevant cartoon

Institute of Traffic Engineers Luncheon


April 1, 2010

Activity Types
TYPE

PURPOSE

DESCRIPTION

CLASSIFICATION

ELIGIBILITY

Work[1]

Working at regular workplace


or work-related activities
outside the home.

Mandatory

Workers and students

University

College +

Mandatory

Age 18+

High School

Grades 9-12

Mandatory

Age 14-17

Grade School

Grades K-8

Mandatory

Age 5-13

Escorting

Pick-up/drop-off passengers
(auto trips only).

Maintenance

Age 16+

Shopping

Shopping away from home.

Maintenance

5+ (if joint travel, all


persons)

Other Maintenance

Personal business/services,
and medical appointments.

Maintenance

5+ (if joint travel, all


persons)

Social/Recreational

Recreation, visiting
friends/family.

Discretionary

5+ (if joint travel, all


persons)

Eat Out

Eating outside of home.

Discretionary

5+ (if joint travel, all


persons)

10

Other Discretionary

Volunteer work, religious


activities.

Discretionary

5+ (if joint travel, all


persons)

Institute of Traffic Engineers Luncheon


April 1, 2010

Modes
Choice

NonMotorized

Auto

Drive-Alone

DriveAlone Free

DriveAlone Pay

Shared-Ride 2

Shared 2
Free

Shared 2
Pay

Shared-Ride 3+

Shared 3+
Free

Walk

Shared 3+
Pay

Explicit toll versus non-toll choice


Explicit treatment of line-haul transit modes

Transit

Bike

Drive-Transit

Walk-Transit

Walk-Local

Walk-BRT

Drive-Local

Drive-BRT

WalkExpress

Walk-LRT

DriveExpress

Drive-LRT

WalkCommuter
Rail

Institute of Traffic Engineers Luncheon April 1,


2010

DriveCommuter
Rail

Treatment of Space: TAZs and MGRAs

About 10
MGRAs to 1 TAZ
32k MGRAs Tot.
All origins and
destinations
located at MGRA
level
Highway
assignments still
use TAZs

Institute of Traffic Engineers Luncheon


April 1, 2010

Transit Network, Stops and Access Points

About 2,500
transit access
points (stops)
Stop-to-stop
skims
(TransCAD)
All transit
boardings/alights
located at TAPs

Institute of Traffic Engineers Luncheon


April 1, 2010

Transit Paths

On-the-fly pathbuilding from


origin MGRA, to
boarding TAP, to
alighting TAP, to
destination
MGRA

Institute of Traffic Engineers Luncheon


April 1, 2010

Tour Destination, Time-of-Day, Mode, Stop Location


1.

Select Primary Destination

2. Select Departure/Arrival Period

3. Select Primary Mode

4. Select Stop Location


Institute of Traffic Engineers Luncheon
April 1, 2010

Time-Use Concept

Recalculate residual time windows

5-6

1-Work

18-19

7-17

2-Discret
20-23

23

Institute of Traffic Engineers Luncheon


April 1, 2010

Tour-Based Model Output


Household Data, Person Data, Tour/Trip List
HID
1
1
1
1

PID
1
1
2
3

TID
1
2
1
1

PUR
2
1
4
2

MOD
1
2
1
4

SB
0
1
0
1

SA
1
0
0
1

OTAZ
943
943
943
943

DTAZ
987
731
952
565

S1TAZ
0
856
0
698

S2TAZ
964
0
0
982

TLOR
1
3
1
1

TLDS
3
3
2
2

Work Trip Frequency Distribution:


Auto Ownership 1, Income Group 1-2
Estimated vs. Observed
6.0%

5.0%

Trip Tables

Frequency

4.0%
Observed
Estimated
3.0%

2.0%

1.0%

Peak Highway Travel Time (minutes)

Maps, Graphics

Assignment
Institute of Traffic Engineers Luncheon
April 1, 2010

Other Summaries

59

56

53

50

47

44

41

38

35

32

29

26

23

20

17

14

11

0.0%

What Sort of Measures/Visuals are Now Possible?


ABM results in a complete activity diary for all SANDAG
residents
A wealth of activity/travel results
Just about any custom report/query/visual is now possible

Scenario Testing

Capacity improvements
HOV, HOT lane scenarios
Cordon Pricing
Land-use policies
New Starts
Equity Analysis

Example Outputs
Institute of Traffic Engineers Luncheon
April 1, 2010

Tracing of Activities/Tours

Institute of Traffic Engineers Luncheon


April 1, 2010

Mode Share by Person Type

Institute of Traffic Engineers Luncheon


April 1, 2010

90

75

Travelers By Age

60

45

30

15

Institute of Traffic Engineers Luncheon


April 1, 2010

Persons Not At Home By TAZ and Hour

Institute of Traffic Engineers Luncheon April 1,


2010

Persons By TAZ and Hour (Daytime Population)

Institute of Traffic Engineers Luncheon April 1,


2010

Mean Delay Peak Period Travel

Institute of Traffic Engineers Luncheon April 1,


2010

Time Spent Traveling by Income & Person Type

Institute of Traffic Engineers Luncheon


April 1, 2010

1. Population Synthesis
2. Long-term

3. Mobility

4. Daily
Mandatory

2.1. Usual workplace / school

3.2a. Car ownership

4.1. Person pattern type

Nonmandatory

Home

Year 1 (2009):
Individual
Mandatory Tours

Individual NonMandatory Tours

4.2.1. Frequency

4.5.1a. Frequency
& Purpose

At-work sub-tours
4.6.1a. Frequency

5. Tour level

5.2a. Stop frequency & Purpose

Daily Trip Productions By Purpose

6. Trip level (4-Step Models)

Trip Distribution
Mode Choice
Trip Assignment

Institute of Traffic Engineers Luncheon


April 1, 2010

Simplified activity-based
travel generation models
estimated, implemented,
and calibrated

Model Development Schedule


Year 2 (2010)
On-board survey data available
Tour mode choice, time-of-day choice, destination choice

Year 3 (2011)
Trip-level models estimated, implemented
Toll transponder ownership
Employer-provided parking and parking lot choice

Year 4 (2012)
Special market models (visitors, air passengers, special events)
PECAS (land-use model) integration
Model validation
Institute of Traffic Engineers Luncheon
April 1, 2010

Questions and Discussion

Joel Freedman
freedman@pbworld.com
503.478.2344
Institute of Traffic Engineers Luncheon
April 1, 2010