Incorporating Traffic Operations into Demand Forecasting Model Daniel Ghile, Stephen Gardner 22nd international EMME Users

1
Incorporating Traffic Operations into Demand
Forecasting ModelDaniel Ghile, Stephen
Gardner 22nd international EMME Users
Conference, PortlandSeptember 15 -16, 2011
2
Presentation Outline

• Background
• Conventional (regional) Model
• Sub-area Model
• Operational Elements
• Model Result Comparison
• Model of Off-Street Parking Lots
• Volume Calibration and validation
• Applications

3
Background

• Location Kelowna, West Canada
• Downtown Plan

4
Background

• Objective Assess implications of transportation
  and land use changes on traffic operations in the
  downtown core
• Approaches applied to assess the detailed traffic
  operational analysis
• Sub-area demand model using EMME
• A micro simulations model focused on network
  impacts
• Intersection Capacity analysis using synchro
• The focus is on the Sub-area travel demand model

5
Regional Model

• Study area covers City of Kelowna, West Kelowna,
  Lake Country and Peachland
• Calibration 2008 fall/spring traffic conditions
  
• Link-based capacity approach
• Drawbacks/limitations
• Coarse Zone system and network in the downtown
  area
• Calibration limited to link volumes at screen
  lines
• Off-street parking lots not treated as special
  zones

6
Sub-area Model

• Uses the regional model as a base
• Detailed zone system and network in the downtown
  area
• Transition area provides a transition between
  the detailed zone system in downtown and the
  coarse regional zone system
• Turn capacity based on Highway Capacity Manual
  (HCM) operational methodology
• Traffic assigned to off-street parking lots
• Detail Intersection coding

7
Operational Elements Model attributes

• Model Procedures
• Regional model approach applied to trip
  generation, distribution and mode split
  calibration
• Traffic assignment uses fixed demand generalized
  cost multi-class assignment with class specific
  volumes
• Travel time is key input to trip distribution
  and assignment and consists of
• Link travel time
• Delay

8
Operational Elements-Travel time

• Link Travel time
• Modified Bureau of Public Roads (BPR) equation
  applied in both Regional and Sub-area models
• Adjustment factor applied to capture difference
  in link capacity for various road classes (0.8 to
  1.1)
• Intersection Delay
• Key component of travel time in congested network
• Volume and capacity are key inputs to delay
• Proper representation of turn capacity is
  required to realistically capture delay at
  intersections

9
Operational Elements- Capacity

• Regional Model Capacity
• Link based capacity
• Drawbacks
• Approach capacity is insensitive to conflicting
  volumes
• All stop controlled approaches have fixed
  capacity (400 veh/h)
• Exclusive left turns have fixed capacity (200
  veh/h)
• Right and left turns are assumed to have the same
  capacity
• Signal phasing and timing are not taken into
  account
• Sub-area Model Capacity
• HCM operational methodology approach applied to
  estimate capacities at signalized and
  unsignalized intersections
• Dynamic adjustment of capacities based on
  projected volumes

10
Operational Elements- Signalized inter.

• Signalized intersection Capacity
• Approaches disaggregated to lane groups
• Saturation flow rate estimated as per HCM
  guideline
• Assumptions made on lane width, grade, percentage
  of heavy vehicles, bus, parking and pedestrian
  activities
• Right turn and left turn adjustment factors
  estimated based on HCM equations
• Left turn adjustment factor is very complicated
• Five cases considered to estimate left turn
  adjustment factor

11
Operational Elements- left turn factor

• Left turn adjustment factor for permitted phase
• HCM approach applied
• Adjustment facto for the lane from which
  permitted left turn are made is estimated by
• --Eq.1
• Adjustment facto for the lane group from which
  permitted left turn are made is estimated by
• --Eq2
• PL proportion of left turn traffic in shared
  lane, EL1 through car- equivalent for permitted
  left turn, N number of lanes
• Parameters used in Eq. 1 and 2

12
Operational Elements- Sample Input and outputs,
Capacity comparison

• Turn Capacity Equations
• EMME turn attributes
• Operational input and output attributes for a
  sample signalized intersection
• Capacity Comparison Signalized intersection

13
Operational Elements-Capacity Comparison
14
Operational Elements- Unsignalized int.

• Unsignalized intersection Capacity
• Opposing Volumes, critical gap, follow-up time
  ,base capacity estimated as per HCM approach
• Capacity adjusted by impedance factor
• Operational input and output attributes for a
  sample unsignalized intersection
• Capacity Comparison Unsignalized intersection

15
Operational Elements- Unsignalized int.
16
Off-street Parking lot Modeling

• Parking lots defined as special zones
• Base year volume established based on parking
  spaces, observed parking occupancy, and parking
  duration
• A procedure was developed that assigns auto trips
  to parking lots
• Volume deducted from regular zones to account for
  the volume allocated to parking lots
• Volume deduction is proportional to the trip
  generation of the zones and inversely
  proportional to the travel impedance between the
  parking lot and the adjacent zone

17
Link Volume Deviation
18
Link Volume Deviation

• Up to 20 screen lines defined within downtown
  core
• Up to 92 of the link volumes fall within the
  acceptable deviations
• Key link volumes all within the acceptable limits

19
Goodness of fit test

• Coefficient of determination (R2)- 0.91
• Slope between model and observed volumes almost
  1.0
• The coefficients show good relationship between
  the model and observed volumes

20
Intersection Turn Volumes

• Turn volumes compared at key intersections along
  Highway 97
• Most major turn movements replicate actual
  observed volumes

21
Model Application

• Produce O-D input to micro simulation model
• Test alternative packages of road/transit
  network improvements
• Evaluate the transportation impact of various
  land use scenarios
• Evaluate alternative transportation demand
  management (TDM) strategies

22
Summary of Findings

• Demonstrated capability to replicate HCM turn
  capacities
• Potential applications include
• Sub-area model,
• Local intersection improvements
• Projection of turn capacities at major
  intersections
• Limitations include
• Intensive coding and error prone
• Application is limited to short term operations
• Detail input required for new or changed traffic
  control
• All vehicles are served irrespective of capacity