TRANSIMS

1
TRANSIMS Mobile Emissions

• David Roden
• AECOM Consult
• March 3, 2003

2
Portland Case Study

• Track 1 Network Simulation Strategies
• Integrate TRANSIMS with Standard Models
• Stabilize and Validate Highway Assignment
• Early Deployment of Router/Microsimulator
• Track 2 GEN2 Model Development
• Utilize TRANSIMS Model Structure
• Complete Application with Sensitivity Tests
• Track 3 Strategic Topics
• Think Ahead

3
TRANSIMS - MOBILE6 Study

• EPA Air Quality Analysis
• Issues of Compatibility
• Methods of Reconciling Differences
• EMME/2 Conversion
• Utilize Standard Modeling Networks and Trip
  Tables as input to TRANSIMS
• Facilitate Comparison with Portland METROs
  MOBILE6 Application

4
EMME/2 Network Conversion

• Convert EMME/2 Network to TRANSIMS
• Nodes, links, lanes, speeds, functional class,
  use
• Automatically/Parametrically Add Features
• Activity locations, parking lots, and process
  links
• Thru links, pocket lanes, and lane connectivity.
• Traffic controls ? signs, signals, detectors,
  phasing and timing plans.
• External stations ? activity locations
• Use GIS to map Zones to Activities Locations

5
Trip Table Conversion

• Convert daily Trips by trip purpose and TAZ to
  Activities at activity locations by time of day
• Split trip tables into P?A and A?P directions
• 15 minute Diurnal Distribution curves
• Randomly assign P?A/A?P trips to activity
  locations and trip start/return times
• Add activity durations and travel time estimates
• Combine trips into household activity patterns
  and travel tours

6
Lessons Learned

• METRO EMME/2 coding conventions
• Lanes, speeds, functional class, link length
• lt 2 activity locations per zone
• Issues with the Track 1 approach
• Refine activity location weights (e.g., airport)
• Auto access on Transit Only roadways
• Smooth diurnal distributions
• Software impacts of unlinked trips
• Multiple access points and methods
• Parking lots, intersections, and intrazonal
  networks

7
Assignment Strategies

• Initial Assignment ? Incremental
• Start with Free Flow Speeds
• Split Households into 30-20-20-10-10-5-5
  Increments
• Build Paths for the Next Increment based on the
  loaded speeds of the Previous Increment
• Subsequent Assignments ? Iterative
• Random Re-Routing
• Off Plan Problems

8
Incremental Assignment
9
A Closer Look
10
Stabilization Strategies

• Stabilization vs. Convergence
• Path-Based Adjustments (Router)
• Trip-Based Adjustments (Activity Generator)
• Iterative Assignments
• Re-Route Off Plan Households
• Random 5-10 per Iteration
• Stabilization Exit Rule
• Less Than 5 of the Activities are Off Plan
• Maximum of 50 of Households Re-Routed

11
Iterative Assignment
12
Validation Checks

• Compare Daily and Peak Period simulated volumes
  to traffic counts by
• Functional class, screenline, district, volume
  group
• Compare TRANSIMS volumes to EMME/2 volumes
• Subjective assessments of
• Time-of-day variations in speeds, travel times,
  and queues
• Output Visualizer animations

13
TRANSIMS Emission Estimator

• Microsimulator output 30 meter segments by one
  hour increments and six speed bins
• 0.0, 16.8, 33.6, 50.3, 67.1, 83.9 mph
• Data Disaggregated by Emission Estimator
• Light-duty and heavy-duty fleet composition
• Splits light-duty into 20 speed bins (4 mph), 34
  power levels, and 23 vehicle types
• Splits heavy-duty into 20 speed bins, 4 power
  levels, and 4 truck and 3 bus vehicle types
• Evaporative Estimator is Trip-based

14
TRANSIMS Traffic Aggregation

• MOBILE6 inputs are typically
• Links or functional classes
• Hours or time periods (peak and off-peak)
• Average volume-delay-based speeds
• What impact does the aggregation level have on
  the emission estimates?
• What aggregation level is most compatible with
  current air quality analysis procedures?

15
Aggregation Strategies
16
Portland MOBILE6 Application

• Compare the TRANSIMS generated emissions with the
  EMME/2 generated emissions
• Same input data, different analysis tools
• Quantify impact of analysis year on MOBILE6
• Same input data, three different analysis years
• Quantify impact of traffic operations
• 1994 base ?1994 with operational improvements
• Re-run TRANSIMS Microsimulator and Emission
  Estimator

17
Project Objectives

• Identify the best aggregation strategy
• Develop methods of adjusting the TRANSIMS outputs
  to account for MOBILE6
• Overall Emission Estimates
• Impact of Analysis Year
• Develop methods of adjusting the MOBILE6 outputs
  to account for traffic operations
• Develop conclusions about the impacts of
  microsimulation on air quality analysis