Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 115


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 114
DYNAMIC TRAFFIC ASSIGNMENT MODEL BREAKDOWN 103 input table for the demand that is to be simulated for the follow simple rules of the road with regard to traffic sig- network. nals; a vehicle encountering a red light at a traffic signal To develop the Vista network data from the Atlanta (or the end of a queue waiting at a traffic signal) waits regional model data, the link and node tables were trans- and moves forward after the signal changes to green (or formed directly into the format required by Vista by the queue moves forward). In Vista, traffic signals are reordering the data fields and converting some of the defined as simple preset signals. It is possible to include units. For the demand data, the regional model demand permitted, protected, and permittedprotected combina- matrices were first exported to test files using the Citi- tion phasing. In the implementation described here, per- labs Cube (see citilabs.com) software with which they mitted left-turn phases were defined if left-turn flows were created, and those text files were then manipulated warranted a separate phase, and permitted phases were to produce the required Vista demand table. The trip defined for through and right-turn movements. matrices from the regional model were developed by Traffic control settings for signalized intersections using typical aggregate model methodologies. The result- were determined by applying a straightforward green ing matrices defined flow rates of vehicles per 4-h time allocation methodology to the approaches at sig- demand period by vehicle class (sov, hov, truck). The nalized intersections. A custom-written program was flow rate from an origin zone to a destination zone could used to read movement flows from an initial Vista model in general be a noninteger number and often was some run and to calculate cycle length, number of phases, fraction of a trip. The Vista demand table required a phase lengths, and order of phasing. From this informa- record for every discrete vehicle, so the conversion tion, the tables required by Vista to represent traffic con- process included converting flow rates to discrete vehi- trol were written. cles while conserving the total number of vehicles and One further note about traffic control might be rele- included assigning departure times, in seconds from the vant. One might be tempted to believe that preset signal start of the simulation period. The departure time profile timing parameters are a limitation given that in reality from the regional model, which specified the proportion many intersections have traffic-actuated traffic control. of vehicles departing during each hour of the 4-h period, This would clearly be the case in a microscopic context was used to control the departure time assignments. where traffic patterns are more or less fixed and the DTA model requirements for network data are some- microscopic model is used to evaluate the impact of times purer than is typical for aggregate travel demand details such as geometry, capacity reductions, traffic con- model networks. Specifically, where an aggregate model trol, and traffic merging and weaving. In the mesoscopic might define the speed field to be free-flow speed as DTA model, however, traffic is simulated to produce observed or as an expected value, including the influence dynamic user-equilibrium route assignment for the vehi- of traffic signals, a DTA model will require posted speed cles in the demand table. Simulating traffic control that for links and should be allowed to simulate the relation- varies as traffic varies is problematic for the dynamic ship among traffic, traffic control, and speed. Similarly, network equilibrium methodology, and the use of preset where an aggregate model might have network capaci- traffic control makes the problem much more tenable. ties defined to ensure that congestion effects are ade- quately represented in assigned network travel times and flows, a DTA model requires that network saturation SOLVING FOR THE DYNAMIC flow rates on surface streets and service flow rates on NETWORK EQUILIBRIUM uninterrupted flow facilities be defined so that the cor- rect aggregate fundamental traffic flow relationships can The dynamic user-equilibrium solution procedure be represented. involves a sequence of steps that include simulating the Because the movement of vehicles through the net- movement of vehicles along predetermined routes, calcu- work is represented by a traffic simulation model to lating those routes between origindestination zone pairs determine network characteristics for the dynamic user- by time interval, and allocating vehicles to one of a set of equilibrium route choice procedure, it is necessary to competing routes. When the routes used by vehicles represent the traffic control system. The traffic simula- between origin and destination by departing assignment tion will simply move vehicles along network links on interval are equal for all origins, destinations, and assign- predetermined paths over time at their free speed as long ment intervals, and no additional lower travel-time as there is room for them to move. When more vehicles routes exist, the dynamic user-equilibrium solution has want to move on a link than the link has room for, con- been determined. gestion will develop, and vehicles will need to progress The vehicle simulation is based on the propagation of according to some rules, usually some kind of first-in, vehicles according to the cell transmission model (CTM) first-out rule. In line with this general way of moving (1). Network links are divided into cells, and vehicles are vehicles along their paths, the simulation model must moved from cell to cell along links and between links as