KY MUG Trip Distribution and Assignment Issues

1
KY MUGTrip Distribution and Assignment Issues

• The Corradino Group
• October 29, 2004

2
Topics

• Trip Distribution Issues
• Highway Assignment Issues
• Calibration Issues
• Friction factors, gamma, exponential
• K-Factors and reasons
• Calibration/Data Sources
• Distribution of trips for zero-auto households

3
Trip Distribution Issues

• Type of Model (GM, IOM, Destination Choice)
• Speeds (FF vs. Congested, equilibration)
• Trip length frequencies and ATL
• Friction factors, gamma, exponential
• K-Factors and reasons
• Distribution of trips for zero-auto households
• Data Sources

4
Trip Distribution Model Type

• Gravity Model
• Singly constrained (SP STrips)
• Doubly constrained (SP SA STrips)
• Fratar (growth factors)
• Intervening Opportunities (order destinations and
  estimate probabilities)
• Destination Choice Promising, but somewhat
  experimental. Consider more than times.
• More purposes e.g., work trips by income)

5
Trip Distribution Gravity Model

• TijPi(AjFijKij)/(Sj AjFijKij)
• where
• Tij trips
• Pi productions
• Aj attractions
• Fij friction factors
• Kij K- factors

6
Trip Distribution Speeds

• Free-flow
• Congested
• Time-of-day
• Combinations
• Equilibration of speeds

7
Trip Distribution Trip length frequencies and ATL

• Goals
• Match the TLFD shape
• Match average trip length
• Right number of intrazonals
• Iterative calibration
• Gamma function
• Manual methods

8
Trip Distribution Friction factors, gamma,
exponential

• Empirical curves
• Gamma (usual method, curve fitting)
• F(I) a(Ib)e(cI)
• Exponential
• F(I) ae(cI)

9
Trip Distribution K-Factors and reasons

• Different distribution pattern than cannot be
  explained by Ps, As or impedance.
• Stable over time?
• Gives rise to destination choice models.

10
Trip Distribution Distribution of trips for
zero-auto households

• Only for transit models.
• Doesnt make sense to distribute using highways
  skims if auto trips is unlikely.
• Use transit skims.

11
Trip Distribution Data Sources

• Best Local surveys (weighted).
• CTPP for HB- Work.
• Previous studies.
• Similar areas.
• NCHRP 365 Values for gamma

12
Highway Assignment Issues

• Type of Model
• All-or-nothing
• Stochastic
• Averaging methods
• Iterative equilibrium (Dial)
• Iterations
• Capacities
• Volume and delay

13
Highway Assignment Type of Model

• All-or-nothing Suitable if trips dont divert
  in response to congestion.
• Very small areas.
• Long distance trips.
• External-external trips.
• Stochastic Spread out trips
• Averaging methods old methods.
• Iterative equilibrium (Dial) So that in the end
  all paths used are equal. Very sensitive to the
  number of iterations.

14
Highway Assignment Capacities

• Same units for volumes and capacities
  (daily/hourly).
• BPR formula was set up for Practical Capacity
  about LOS D
• Some new models consider intersection delay. Some
  are dynamic.

15
Highway Assignment Volume and delay

• BPR fomula
• TcTi 1 a(v/c)b
• Tc Congested time
• Ti Free-flow time
• v Assigned volume
• c Capacity
• a alpha (traditionally 0.15)
• b beta (traditionally 4.00)

16
Highway Assignment Volume and delay
Source NCHRP 365
17
Highway Assignment Volume and delay
18
Calibration Issues

• NCHRP 365
• Calibration and Adjustment of System Planning
  Models
• Calibration by step
• Common adjustments
• Automatic Model Calibration

19
Calibration Generation

• Trip Rates from Local Surveys
• Special Generators
• NCHRP 365
• Amount of traffic
• About 9 15 person trips per HH per day

20
Calibration Mode Auto Occupancy

• From Local Surveys
• CTPP (work)
• NCHRP 365
• Amount of traffic
• Work seldom greater than 1.1 persons/auto
• Overall seldom greater than 1.5 persons/auto

21
Calibration Mode Distribution

• Screenlines
• Major screenlines with about 10
• Cutlines with about 15
• Match target ATL (local surveys)

22
Calibration Mode Assignment

• Percent Root-mean-square error
• By volume group
• By roadway type
• VMT ratio by AT, FT
• HPMS observed VMT.

23
Calibration Mode Assignment
24
Calibration Mode Assignment

• Minor speed/capacity adjustments (by tables)
  generally frowned upon.
• Local calibration of free-flow speeds, using
  posted speeds, from observed data and signals.
• Adjustments to the volume delay functions.
• Centroid connectors and their locations.

25
QUESTIONS?

• ?