Typical Application of the TEXAS Model

TLDR: A simple application of the TEXAS computer model by a traffic engineer in a small city is described and the effects of a change in one of the three parameters (traffic flow, intersection geometry, and intersection control) with only two runs are ascertained.

Abstract: This paper describes a simple application of the TEXAS computer model by a traffic engineer in a small city. (TEXAS is a microscopic model for simulation of traffic at a single intersection. it is currently available from the Texas State Department of Highways and Public Transportation.) TEXAS allows traffic engineers to evaluate changes in intersection parameters (traffic flow, intersection geometry, and intersection control) and to see what effect those changes have on the vehicles' and intersection's performance. TEXAS is comprised of three separate computer programs: GEOPRO, DVPRO, and SIMPRO (see Figure 1). GEOPRO takes geometric information about the intersection system (approach lengths, number of lanes per approach, lane geometry and type, and location of any sight distance restrictions) in a cartesian coordinate manner; it produces a list of possible paths down which vehicles will travel. This path information is used as input to SIMPRO. DVPRO also produces input for SIMPRO. This drivervehicle processor takes volume and headway distribution information and creates a time-ordered list of vehicles. Three types of drivers and 16 classes of vehicles are used. SIMPRO takes these two inputs and a third, which contains the description of intersection control (from unsigned to signed to signalized) and the duration of simulation. Vehicles are \"stepped through\" the system, and speed and delay statistics are gathered for each time increment for each vehicle. At the end of the simulation run, the statistics are summarized for the total intersection, for each approach, and for each turn movement in each approach. During a typical time increment, each car examines the vehicle in front, the adjacent lane(s), and the traffic control at the intersection. Then it makes 'a deterministic decision whether to speed up, slow down, start, stop, or change lanes. Because of the deterministic nature of the model, the traffic engineer is able to ascertain, the effects of a change in one of the three parameters (traffic flow, intersection geometry, and intersection control) with only two runs: \"before\" and \"after\". The following is a description of how I used the model in just this way and was able to make comparisons between two runs. Richardson is a Dallas suburb with a population of 80 000. Its 53 traffic signals are located at arterial intersections on a suburban grid and are, for the most part, noninterconnected and fully actuated. When these signals were installed, multiphase, fully actuated operation was the state of the practice. At many of the locations left-turn phasing was provided, even though during the peak period only three to five vehicles made the left turns each cycle. It had been observed that those three leftturning vehicles were causing unnecessary delays to the opposing through movement. With the increased emphasis today on reducing overall delay and fuel consumption, about 10 locations were targeted for protected left-turn removal in, one or both directions. On January 10, 1981, left-turn green arrows