FHWA Traffic Analysis Tools Measures of Effectiveness

1
FHWA Traffic Analysis Tools Measures of
Effectiveness

• Richard Dowling
• July 28, 2006

2
The Problem

• The HCM and traffic simulation tools rarely give
  same performance results.
• No guidelines on how to interpret performance
  results.

3
Project Objectives

• To develop information and guidance on
• Which MOEs should be produced,
• How they should be interpreted, and
• How they should be defined and calculated in
  traffic analysis tools

4
Approach

1. Review current practice.
2. Review how MOEs computed from field data.
3. Review how MOEs are defined and computed within
   traffic simulation models and the HCM.
4. Develop innovative approach to interpret MOEs.
5. Demonstrate approach in a case study.

5
Status

• Draft report being reviewed by expert panel.

6
Microsim MOEs

• Currently
• Travel time, speed, delay, queues
• Critique
• Computed differently in different software
• Breakdown at high congestion levels
• Hard to interpret

7
Examples of Computation Differences

• Mean speed
• Definition of Queue
• Queue length

8
1. Mean Speed on Arterial Link
Time
Signal
Analysis Period
Vehicle Trajectories
Distance
Link Length
9
Mean Speed - SimTraffic
Time
SpeedVMT/VHT
Distance
10
Mean Speed - CORSIM
SpeedVMT/VHT
Time
11
All traces used for VHT. Only traces 1-11 for VMT.
Distance
1
11
Mean Speed - AimSun
Time
SpeedVMT/(VHT)
0
ExitTime
EntryTime
Distance
12
Recommendation

• Adopt analysis domain that is limited to only
  some of the vehicles in the simulation period.
• Analysis period is specified by user.
• Simulation period is amount of time required to
  clear all vehicles entering (or attempting to
  enter) link during analysis period.

13
(No Transcript)
14
2. Definition of Queue

• HCM vehicle waiting to be served.
• SimTraf speed lt 11 km/h
• Corsim speed lt 3-9 km/h
• Vissim user defined, (def. lt 5 km/h)
• Paramics user defined, (def. lt 7 km/h)
• Aimsun user defined.

15
Signal Example
16
Queue Overflows Link
HCM, Synchro, Vissim(2)
Aimsun, Corsim, Vissim Paramics, SimTraf
Distance
Link Length
17
Queue Does Not Clear
HCM/Synchro Max.Back.Queue
Simulation Max.Back.Queue
Simulation Max Queue
Analysis Period
Distance
Link Length
18
Recommendation

• Adopt standard maximum speed and minimum spacing
  for queues in US.
• Adopt queue performance measures that identify
  overflow potential when queue exceeds 90 of
  storage capacity.
• Do not rely upon simulation reported queue
  lengths when queue exceeds 90 of storage
  capacity.

19
Breakdown of Some MOEs

• Under parking lot conditions
• Speed goes to zero.
• Density goes to jam density.
• Stops go to zero.
• Variance goes to zero.

20
Recommendation

• Place greater reliance on travel time and delay
  performance measures that do not flatten out
  under extreme congestion (parking lot) conditions.

21
MOEs for Decision Makers

• Indices that have obvious good and bad values.
• Reference each index to a desired state
• Continue to be meaningful under extreme LOS F
  congested conditions.

22
Recommended MOEs

• Travel Time Index
• Percent Delay
• Buffer Index
• Intersection Blockages (, Hours)
• LOS F Segments (, Hours)
• Fuel Consumption Ratio
• Air Pollution (NOx) Emissions Ratio

23
Travel Time Index

• Ratio of actual to free-flow travel time.
• For freeways TTI gt 1.5 means over capacity.
• For signalized arterials, TTI gt 2.5 means over
  capacity and/or poor signal coordination.

24
TTIs at Capacity Uninterrupted Flow Facilities
25
TTI at Capacity Uncoordinated Arterials
26
TTI at Capacity Coordinated Arterials
27
Percent Delay

• Ratio of mean delay to mean trip time at
  free-flow speeds.

28
Buffer Index

• Excess travel time that must be budgeted to
  ensure 95 probability of arriving on-time
  divided by the mean trip time.
• Values of 25 to 33 typical of urban commutes.

29
Fuel Consumption Index

• Ratio of actual fuel consumed to hypothetical
  fuel consumed at fuel efficient travel speeds, no
  stops.
• Values near 1.00 indicate near optimal operations.

30
Air Pollution Index

• Ratio of nitrous oxides (NOx) emissions to
  emissions at minimal polluting travel speed (no
  stops).
• Values of near 1.00 indicate near optimal
  operation.