Transcript Regional Travel Modeling

Regional Travel Modeling
Unit 6: Aggregate Modeling
Puget Sound Regional Council’s travel
model
• Forecast future travel patterns and conditions within
the four counties (King, Kitsap, Pierce, and Snohomish)
of the Puget Sound region
• Represents the state of the art for regional travel
modeling
• Analyze the likely impacts of travel forecasts on the
region’s transportation infrastructure and environment
• Interacts directly with the land use model (UrbanSim),
and it is the primary source of input data for the air
quality (EPA Moves) and benefit-cost analysis (BCA)
tools
* Puget Sound Regional Council (PSRC) is the Seattle region’s MPO
Modeling Framework
Seven basic components
•
•
•
•
•
•
•
Household vehicle availability
Person trip generation
Trip distribution
Mode choice (passenger travel)
Time of day
Trucks
Vehicle trip assignment
Commercial and Passenger Handled
Separately
• Truck trips are generated, distributed, and then
assigned to specific routes.
• There is no truck mode choice model as all travel is
assumed to occur via the truck mode.
• Uses TAZ-level employment data provided by
UrbanSim to estimate truck trip generation.
• The truck model converts trucks to passenger car
equivalents (PCE) prior to route assignment.
– accounts for the fact that large trucks occupy more space
on roadways than do passenger cars
– important for representing the contribution of truck travel
to congestion and overall travel speeds.
– light trucks represent 1.0 PCE; medium trucks represent
1.5 PCEs and; heavy trucks represent 2.0 PCEs.
Basic unit: TAZ
• Transportation Analysis Zone
• 938 in 4 county region
• Estimate truck trips at the TAZ level based on
employment in that zone
5 time periods
1) AM Peak Period (6:00 am to 9:00 am);
2) Mid-Day (9:00 am to 3:00 pm);
3) PM Peak Period (3:00 pm to 6:00 pm);
4) Evening (6:00 pm to 10:00 pm);
5) Night (10:00 pm to 6:00 am).
Three truck types
• Light Trucks
– Four or more tires, two axles and weighing less than
16,000 pounds
– commercial vehicles such as taxis, rental cars, school
buses, ambulances, etc
• Medium Trucks
– Six or more tires, two to four axles, and weighing
between 16,000 and 52,000 pounds
• Heavy Trucks
– Double- or triple-unit, having 5 or more axles, and
weighing more than 52,000 pounds
Trips estimated from employment
• Confidentiality issues result in generally underestimated total
employment
• Certain kinds of manufacturing that are not included in the PSRC
land use model (primarily construction and resources employment)
–
–
–
–
–
–
–
–
–
–
–
Ag/forestry/fishing
Mining
Construction
Manufacturing – products
Manufacturing – equipment
Communication, transportation, utility
Wholesale
Retail
Financial, real estate, service (fires)
Government
Education
Trip rates
• Attraction and production both estimated
from employment using different factors
• Only change due to changes in employment
– Not transportation system performance
– No elasticity of demand with respect to travel cost
• Special generators
– Ports and warehousing districts
– Survey based
External trips
• Generated and/or destined for outside the
region
• 1997 TRANSEARCH data
• 2001-2003 SFTA data
Validation
• The FASTruck model was validated system
wide based on a limited set of truck counts for
medium and heavy trucks.
• Light-truck travel was validated on a system
wide basis to ensure that the assignments of
these trucks did not significantly affect the
assignments of passenger cars.
• The existing FASTruck model does include all
non-personal-use vehicles.
Trip destination
• Gravity model approach
– Closer TAZes attract more trips
– Different for each truck class
• Estimated from consultant report
• SFTA survey
• Distributed over the day
– From PSRC screenline counts from trucks
Network
• Traffic assignment with equilibrium
assumption
• Congestion effects captured with link
performance functions
• Only the highway network is used
– Recall zone to zone travel
• Trucks assigned along with all other traffic
– All vehicles on same link travel same speed
– Reduce truck speeds by 25%