Roadway Traffic Assignment (Last Edit August 26, 2005)

Similar documents
Some network flow problems in urban road networks. Michael Zhang Civil and Environmental Engineering University of California Davis

Origin-Destination Trips and Skims Matrices

VISUM State-of-the-Art Travel Demand Modeling VISUM

The Secrets to HCM Consistency Using Simulation Models

Travel Time in Macroscopic Traffic Models for Origin-Destination Estimation

APPENDIX H: TRAVEL DEMAND MODEL VALIDATION AND ANALYSIS

US 75 Integrated Corridor Management System Using Technology and Partnership to Maximize Transportation System Capacity

SUTRA : Sustainable Urban Transportation for the City of Tomorrow

The new Uber for Business. How to make the most of the new features

AN INTEGRATED TRANSPORTATION PLANNING AND OPERATIONS MODEL FOR MEMPHIS, TENNESSEE

Maximize Transportation System Capacity

The Policies section will also provide guidance and short range policies in order to accomplish the goals and objectives.

TPB Scenario Study Task Force: Proposal for Development and Analysis of Two New Scenarios

I-66 Corridor Improvements Outside the Capital Beltway in Northern Virginia, USA

AMPO Annual Conference Session: Performance (Part 1) October 18, 2017 Savannah, GA

Equity Issues Associated with Roadway Pricing

Quantifying the performance of a traffic data collection system: Scout Connect match rate evaluation

AIR QUALITY AND CLIMATE CHANGE EVALUATION GUIDANCE

Management. VA SITE Annual Meeting June 27, 2013 Jay Styles Performance and Strategic t Planning Manager, Business Transformation Office

Are We Ready for Autonomous and Connected Vehicles?

Truck GPS Data for Freight Planning

Niagara Region Transportation Master Plan Niagara-Hamilton Trade Corridor Technical Paper

PRICING STRATEGIES PRESENTED BY JEFFREY D. ENSOR MALAYSIA TRANSPORT RESEARCH GROUP TO THE NOVEMBER 25, 2003

I-30 Express Lanes Survey Report

Appendix B5 PT Model Validation

Regional Initiatives on Land Use and Transportation

Dallas ICM & 511 System. 87 th Annual Transportation Short Course

Transform 66 Multimodal Project: Prioritization Process and Evaluation Criteria Approved March 3, 2016

Construction Alternative Screening with Regional Travel Demand Model

MONITORING IMPLEMENTATION AND PERFORMANCE

2040 Transportation Policy Plan Conversion to Performance Based Planning Basis. 25 th Annual CTS Transportation Research Conference May 21, 2014

MEMORANDUM EXAMPLES FOR ILLUSTRATIVE PURPOSES NEXT STEPS. Item 3 Long-Range Plan Task Force May 17, 2017

APPENDIX B. Public Works and Development Engineering Services Division Guidelines for Traffic Impact Studies

BOSTON REGION METROPOLITAN PLANNING ORGANIZATION

Greenhouse Gas Emission Reduction Strategies

TRANSMIT: Travel-Time System for Border Crossings in Ontario s Niagara Region. Stephen Erwin, P.Eng., Ministry of Transportation of Ontaro

Integrated Corridor Management --An Overview --

IMPLEMENTING TOLL PLAZA ANALYSIS INTO FREEPLAN

Congestion Management in a Rapidly Growing Economy: Optimizing Transportation Network Performance

Central Phoenix Transportation Framework Study

A Network Demand Model for Rural Bypass Planning. Paper Number

Transit Service Guidelines

Curbfront Case Studies

INTERSTATE CORRIDOR PLANNING

Risk assessment and incident impact traffic flow estimation techniques for a monitor integrated road safety system.

Evaluation Summary for the San Diego Testbed. Final Report August 2017 FHWA-JPO

Appendix D: Functional Classification Criteria and Characteristics, and MnDOT Access Guidance

Technical Report Documentation Page 2. Government Accession No. 3. Recipient's Catalog No. 1. Report No. FHWA/TX-14/

TRAFFIC & TRANSPORTATION COMMISSION AGENDA REPORT

DART s Role in Transit - Oriented Development. Jack Wierzenski,, Director Economic Development & Planning Dallas Area Rapid Transit

NATMEC June 30, 2014 Anita Vandervalk, PE, PMP

Buffalo Niagara Integrated Corridor Management Project. ENTERPRISE Webinar Keir Opie, Cambridge Systematics September 26, 2016

FUTURE BUS RAPID TRANSIT PLANS

Operations in the 21st Century DOT Meeting Customers Needs and Expectations

Travel Time Reliability in the SLOCOG Region. October 27, 2014 San Luis Obispo Council of Governments Transportation Education Series

Study of Transit Alternatives in the Dulles Airport Access Road Corridor

Woodburn Interchange Project Transportation Technical Report

Day/Night Noise Level Assessment Tool Users Guide

Chapter 7 Entity Transfer and Steady-State Statistical Analysis

IH 820 Corridor Improvement Study Randol Mill Road To North SH 121 Interchange. Public Meeting. Thursday April 4, 2013

Model development. Peter Mijjer 4 oktober 2010

I-710 Project Committee Meeting

VALUING GHG SAVINGS FROM ITS IMPLEMENTATION IN URBAN AREAS

Mobility on Demand for Improving Business Profits and User Satisfaction

OPTIMIZING RAMP METERING STRATEGIES

Washtenaw Area Transportation Study Request for Proposal for Model Update and Improvements

New Jersey Pilot Study

Volume to Capacity Estimation of Signalized Road Networks for Metropolitan Transportation Planning. Hiron Fernando, BSCE. A Thesis CIVIL ENGINEERING

Transit Pre-Planning Planning &

Analytical Tools and Decision Support Systems SANDAG I-15 ICM. ITS PA January 30 th, 2014

FULLY AUTONOMOUS VEHICLES: ANALYZING TRANSPORTATION NETWORK PERFORMANCE AND OPERATING SCENARIOS IN THE GREATER TORONTO AREA, CANADA

REGIONAL SOLICITATION ROADWAY APPLICATIONS: POTENTIAL CHANGES. TAB September 20, 2017

METRA UP-W LINE. Locally Preferred Alternative Report

Multi-modality assessments in promoting sustainable urban traffic

TRAFFIC MODELLING PLATFORMS IN TEMA-TT

GO RAIL NIAGARA SERVICE EXTENSION

National Capital Region Congestion Report

Trans-Texas Corridor-35 Update

Induced Demand and Rebound Effects in Road Transport

LATTS II - Freight Investment Decision Principles

Environmental Traffic Assignment: Developing Emission-based Models

Impact of Traffic Conditions and Carpool Lane Availability on Peer-to-Peer Ridesharing Demand

TRANSPORTATION FACTS. OUR CUSTOMERS: Travel Patterns

IOWA TRANSPORTATION BY THE NUMBERS:

Dynamic Trip Assignment-Simulation Model for Intermodal Transportation Networks

Passenger Transport Modelling [T1]

Evaluation of traffic gases emissions: the case of the city of Tunis

Emergency Incident Management, Benefits and Operational Issues

CALIFORNIA AIR RESOURCES BOARD REPORT ON REGIONAL GREENHOUSE GAS EMISSION REDUCTION TARGETS PURSUANT TO SB375

Strategic Transportation Plan. Presented to: ECO-Rapid Transit Board of Directors Presented by: Gill V. Hicks, Cambridge Systematics, Inc.

REQUIREMENTS FOR THE CALIBRATION OF TRAFFIC SIMULATION MODELS

CITY OF JACKSONVILLE 2030 MOBILITY PLAN STUDY UPDATE GOALS, OBJECTIVES, AND PERFORMANCE MEASURES PREPARED FOR: CITY OF JACKSONVILLE

HB2 Implementation Policy Guide

Energy Savings by replacing old facility o Energy savings o Emissions

Automated Permitting and Routing for OS/OW. GIS-T, 2010 Session Jay Adams - Oklahoma DOT

EVALUATION OF INCIDENT MANAGEMENT BENEFITS USING TRAFFIC SIMULATION

TRANSPORTATION IMPACT ANALYSIS GUIDELINES

Highways England s Regional Traffic Models Challenges over the past year Saturn User Group 2 nd November Alison Cox

HCM 6th Edition Planning and Preliminary Engineering Applications Guide (PPEAG) to the HCM Bringing Back-of-the-Envelope Back.

This page intentionally left blank.

Transcription:

Roadway Traffic Assignment (Last Edit August 26, 2005) This chapter focuses on the Roadway Traffic Assignment module of the Dallas-Fort Worth Regional Travel Model (DFWRTM). Based on the input vehicle trip tables, roadway link attributes, and link performance functions, the traffic assignment output contains the roadway link flows. Commonly referred to as link traffic volumes, the link flows are estimates of the number of vehicles on each link. Each vehicle trip table introduces a different vehicle class in the assignment process. The traffic assignment procedure estimates the link volume (by vehicle class) for non-fixed-route vehicle types that choose their paths. In other words, fixedroute vehicles, such as transit buses, are not directly considered in this procedure. The interaction between roadway and transit travel times is explained in the transit assignment chapter. The following is the description of the overall design of the traffic assignment process, including the inputs created in previous stages of the DFWRTM. After model time periods and vehicle classes are introduced, a brief discussion about selection of the user equilibrium (UE) method for traffic assignment is provided. The link volume-delay performance functions, value of time assumptions, and generalized cost procedures are also explained. The chapter ends with documentation of the traffic assignment implementation within the TransCAD environment. Overall Design Traffic Assignment consists of a set of procedures after the Time of Day module and before either the Update Time or Transit Assignment modules (depending on whether feedback to trip distribution is performed). Figure 1 shows the position of traffic assignment in the full model run flow chart. The objective of a Base Year traffic assignment validation is to load the origin-destination (OD) vehicle trip demand tables onto the roadway network in such a way that the resulting traffic volumes on the roadways provide a realistic estimate of actual traffic counts. Since drivers travel behavior and/or roadway system conditions may change depending on the time of day of travel and the type of user, there needs to be a breakdown of trips into time of day and vehicle classes. Traffic Assignment 1 NCTCOG Regional Model

Figure 1:Roadway traffic assignment module location in the full model run flow chart Time Periods A time-of-day (TOD) breakdown of weekday vehicle trips is important because roadway flows vary significantly for different hours of the day. The application of TOD traffic assignments results in more realistic estimates of both volumes and speeds by time period. Such detailed information is used for many transportation studies, including an air quality conformity analysis that requires TOD link speed and vehicle classification volumes. Three time periods are used for preparation of three TOD vehicle trip tables that are then used for three separate traffic assignments: AM: Morning peak period, 6:30 AM to 8:59 AM (2.5 hours) PM: Evening peak period, 3:00 PM to 6:29 PM (3.5 hours) OP: Off-peak times, 9:00 AM 2:59 PM and 6:30 PM 6:29 AM (18 hours) Traffic Assignment 2 NCTCOG Regional Model

Vehicle Classes The roadway network consists of some facilities that are not used by all drivers, e.g., a typical HOV lane is open only to a specific group of vehicles. In some cases (such as an Air Quality analysis), the model user needs statistics for specific groups of vehicles such as trucks. To address roadway network restrictions for different vehicle groups and output requirements of the model, four classes of vehicles in the traffic assignment are used. The vehicle classes and their corresponding roadway network are: DA: the drive alone vehicles with only a driver in them, can use all links except rail and regular HOV links SRHOV: the shared-ride vehicles with more than one person in them, can use all links except rail links SRNOHOV: the shared-ride vehicles with more than one person in them, can use all links except rail and HOV links [Note: the factors used to distribute SR vehicles into HOV and NOHOV groups was established from a model calibration exercise] TRUCK: commercial vehicles with six or more tires, can use all links except rail and HOV links (or other links specifically identified by the model users) Assignment Method Different algorithms can be used to estimate link traffic volumes. The User Equilibrium (UE) generalized cost method is adopted for the DFWRTM for three reasons: The UE method represents standard good modeling practice in the United States Paths are based on roadway operating costs, toll costs, and travel times The congested travel times are sensitive to the capacity and volume of a roadway The behavioral assumption of the UE method is that each vehicle travels on the zone-to-zone path that minimizes that vehicle s generalized cost of travel. This choice rule implies that, at equilibrium, the link flow pattern is such that the travel times on all used paths connecting any given Origin-Destination pair will be equal; the travel time on all of these used paths will also be less than or equal to travel time on any of the unused paths. At this point, the network is in user equilibrium. This method is formulated as an optimization program, which converges toward the equilibrium condition through iterative adjustment of link volumes for each vehicle class. UE provides a stable and consistent method for traffic assignment. Traffic Assignment 3 NCTCOG Regional Model

The Incremental Assignment method was used in the previous Dallas Fort-Worth regional travel model, in which a fixed portion of the trip table is loaded onto the network at each iteration. The link travel times are then updated and an additional portion of the trip table is then loaded onto the network. The result of an incremental assignment may resemble an equilibrium assignment when many increments are used, but an incremental assignment can never guarantee a true UE solution. The selected method of assignment, user equilibrium with generalized cost, requires performance function for each link to estimate link travel time, cost value for each link, and value of time for each vehicle class to convert time to cost and calculate generalized cost. Volume Delay Functions Since roadway facilities have limited capacities for carrying traffic, the purpose of the traffic assignment volume-delay function is to calculate a link travel time for the next UE assignment iteration that takes into account the link s capacity and previous iteration volume. As the number of vehicles using a facility increases, the interaction of vehicles decreases the performance (speed) of that facility. The volume-delay or performance function calculates a congested travel time that reflects the uncongested free speed travel time and the delay time in terms of volume over capacity (vol/cap) ratio of the roadway link. In theory, the assigned link volume should never exceed the available link capacity, so the link travel time calculated for a case where the volume exceeds the capacity should be infinite. Nonetheless, traffic assignment models are used for general planning rather than detailed vehicle operations studies, so the typical volume-delay function used in most regional models allows for the assigned volume to exceed the capacity. The DFWRTM volume-delay function is similar in form to the BPR-type functions used in other regional models, in that link speed decreases as the vol/cap ratio increases. However, NCTCOG s equation includes a maximum minutes of delay per mile assumption that effectively limits the drop in link speed that can occur, as the vol/cap ratio increases. The general form of NCTCOG s volume-delay function is as follows: TravelTime = FreeFlowTime + ( Length) Min{ A. e B.( vol ) cap, C} Traffic Assignment 4 NCTCOG Regional Model

Values for A, B, and C for Freeway and non-freeway links are shown in Table 1. Figure 2 shows the relation between speed and vol/cap ratio for Freeway and Arterial links assumed in the traffic assignment. Table 1: Delay function parameters Parameter Freeway Arterial A 0.015 0.05 B 6.0 3.90 C 4.0 5.0 Delay Functions 70 60 Speed (mph) 0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 50 40 30 20 10 0 Freeway Arterial Volume / Capacity Figure 2: Speed versus volume / capacity ratio for roadway links Values for capacities and times are stored in the network file under the following fields: AMCAP_*: Capacity for AM period PMCAP_*: Capacity for PM period OPCAP_*: Capacity for OP period PKFRTIME_*: Free flow time for AM and PM periods OPFRTIME_*: Free flow time for OP period Traffic Assignment 5 NCTCOG Regional Model

Roadway Link Toll and Cost Two types of costs are assigned for each roadway link: Vehicle operation-related costs and roadway tolls. While vehicle operating costs can include both long-term costs (insurance, car payments, repairs, etc.) and short-term out of pocket costs (primarily the fuel cost), the NCTCOG strategy is to consider only short-term costs as an influence on choosing a path in the assignment module. The auto operating cost is assumed to be 7.3 cents per mile, expressed in constant 1999 dollars. Tolls can be considered within the model by two methods: A fixed dollar value for a link, or a dollar per mile value. When the location of toll plazas and the amount of the toll are known, the fixed value is used. In other cases, the toll per mile can be entered for a roadway link and subsequently calculated internally as a link toll. In both cases, the link tolls are adjusted to constant 1999 dollars. Toll adjustment factors are shown in Figure 3. Managed lane facilities (HOV lanes that can be used by the Drive Alone class by paying a toll but free for Share Ride classes), have a higher cost for DA class, which is reflected in the roadway network file in a different field. Based on these two cost values, the direct dollar cost of traveling through each link is calculated as: [ FixedToll + ( TollMile)( Length) ] AdjustedToll = ( AdjustmentFactor OperatingCost = ( Length)( CostMile) + AdjustedToll yr ) AdjustmentFactor yr : Adjustment factor for tolls to convert the toll values for the year yr to 1999 Dollar CostMile : Operating cost per mile for all links, constant 0.073 Dollar/Mile Cost values for links are stored in the following fields in the network file: TOLL_*: Value of fixed toll at Toll booth in Cents TOLLMILE: Value of toll per mile in cents OPERCOST_*: Adjusted dollar value for passing through a link for non-drive Alone classes in Dollar OPERCOSTDA_*: Adjusted dollar value for passing through a link for Drive Alone class in Dollar TOLLROAD: Code fields used for managed lane facilities, value 3 means manages lane Traffic Assignment 6 NCTCOG Regional Model

MODEL_LENGTH: Length of the link in mile Toll Adjustment Factor 1.6000 1.4000 Factor 1.2000 1.0000 0.8000 0.6000 0.4000 0.2000 1.0000 0.0000 1980 1990 2000 2010 2020 2030 2040 Year Figure 3: Toll Adjustment Factor based-on 1999 Dollar Generalized Cost and Value of Time For UE optimization, each link s generalized cost is composed of travel time, an auto operating cost (7.3 cents per mile in 1999 constant dollars), and the toll cost (expressed in 1999 constant dollars). Since the generalized cost for a specific link must be expressed as a single value, the travel time on each link needs to be converted to a 1999 constant dollar amount by means of a value of time (VOT) assumption. Calculation of the correct or most appropriate VOT for use with traffic assignments is often a difficult undertaking because there are many factors that may influence an individual person s VOT (e.g., trip purpose and the socio-economic characteristics of the traveler). The VOTs used in the DFWRTM traffic assignment are based on a 1999 Base Year model calibration process that resulted in a good match between the modeled and observed toll road volumes. The resulting VOTs were as follows: Traffic Assignment 7 NCTCOG Regional Model

$10 / hour ($0.167 / minute) for each of the three auto-based vehicle classes (DA, SRHOV, and SRNOHOV) $12/ hour ($0.2 / minute) for the truck class The total cost of traveling though a roadway link is calculated as: Generalize dcost = OperatingCost + ( VOT )( TravelTime) Generalized cost is used in UE assignment module. Implementation of Roadway Traffic Assignment in TransCAD As noted previously, the DFWRTM consists of AM Peak, PM Peak, and Off-peak traffic assignment runs. For each assignment there are two key steps: first a TransCAD network file is created, and then, the TransCAD UE multi-modal multi-class assignment is run. For the AM peak time period, the following steps are taken to create the AM network: 1. Load ACTRDWY.DBD file (the roadway link geographic file) 2. Create a selection of links with AMHRCAP_AB > 0, call this selection AMLinks 3. Create a selection of nodes with CENTROID > 0, call this selection Centroids 4. Create a network file with these settings: a. Create links from: AMLinks b. Read length from: MODEL_LENGTH c. Other Link fields: A_PK, B_PK, C_PK, MODEL_LENGTH, AMCAP_*, OPERCOSTDA_*, OPERCOST_*, PKFRTIME_* d. Check drop duplicate link option e. Save this network as AM.NET 5. Open the network setting dialog box, under Centroid section, choose Create from selection set radio button and select Centroids selection set. 6. Click OK. The network file is created and Centroids are identified in it. Screen shots for the creation of AM network file and its setting are shown below: Traffic Assignment 8 NCTCOG Regional Model

Figure 4: Screen shot for creating AM network file Figure 5: Screen shot for changing the setting of AM network file for Centroids For the AM peak time period, the following steps are taken to run the AM peak assignment: Traffic Assignment 9 NCTCOG Regional Model

1. Load ACTRDWY.DBD file 2. Load AM.NET file 3. Load AM.MTX file (contains AM vehicle trip tables for four classes) 4. Create a selection of links with AMHRCAP_AB > 0, call this selection AMLinks 5. Create a selection of links with AMHRCAP_AB > 0 and FUNCL = 8 and TOLLROAD <> 3, call this selection ExclusionSetDA 6. Create a selection of links with AMHRCAP_AB > 0 and FUNCL = 8, call this selection ExclusionSetSRNOHOV 7. If exc_truck field exists in the link layer, create a selection of links with AMHRCAP_AB > 0 and (FUNCL = 8 or exc_truck = 1). If exc_truck field does not exist, create a selection of links with AMHRCAP_AB > 0 and FUNCL = 8. Call this selection ExclusionSetTruck 8. Bring up Multi-modal Multi-class Assignment dialog box and use the following setting: a. Press Network button, under Toll tab, choose Toll Links in selection set radio button and select AMLinks b. Delay Function: NCTCOG c. Method: User Equilibrium d. O-D Matrix: AM e. Class information, use following table Matrices PCE VOT Fixed Toll Road Toll Exclusion Set Drive Alone 1.0 0.167 OPERCOSTDA_* - ExclusionSetDA SRIDE NOHOV 1.0 0.167 OPERCOST_* - ExclusionSetSRNOHOV SRIDE HOV 1.0 0.167 OPERCOST_* - - Truck 1.0 0.2 OPERCOST_* - ExclusionSetTruck f. Delay Function Parameters, use following table Name Field Default Value Time PKFRTIME_* N/A Capacity AMCAP_* N/A Length MODEL_LENGTH N/A A A_PK 0.015 B B_PK 6.2 C C_PK 60 Preload None N/A g. Press Option button, under other options, check class flows h. Use 30 for number of iterations and 0.00 for convergence value Traffic Assignment 10 NCTCOG Regional Model

i. Press OK, and specify the output file locations Traffic assignment may take 3 hours to run for AM period. Screen shots of TransCAD dialog boxes for traffic assignment are shown in Figure 5 and 6. Similar steps, with adjustment related to time periods, are taken for running OP and PM traffic assignments. After the traffic assignment is done for each time period, estimated volumes for each class are stored in the output files. Total traffic volume for each time period is calculated and stored in the roadway network file. Also, the volumes are used in a post-processed procedure for link travel time estimation, which are also stored in the roadway network file. If a feedback run is needed for a model run, new link estimated times are used for skim matrices. Total volumes and times are stored in the following fields: AMVOL_*: Total link volume for AM period PMVOL_*: Total link volume for PM period OPVOL_*: Total link volume for OP period PKTIME_*: Loaded link travel time in AM period PMTIME_*: Loaded link travel time in PM period OPTIME_*: Loaded link travel time in OP period Traffic Assignment 11 NCTCOG Regional Model

Figure 6: Screen shot for changing the setting of AM network file for Toll links Figure 7: Screen shot for reporting assigned vehicle classes in the output file Traffic Assignment 12 NCTCOG Regional Model

Figure 8: Screen shot for MMA setting for AM assignment Traffic Assignment 13 NCTCOG Regional Model