Simulation of Mixed Traffic Flow

Simulation of Mixed Traffic Flow on Two-Lane Roads by Dey, P., Chandra, S. And Gangopadhyay, S.

This article describes a study of two-lane roads conducted using a computer simulation modeled on various traffic situations. This simulation was based on observed behaviors of certain types of vehicles with various directional delineations for each lane on the road., which were classified according to assigned Passenger Car Units (PCUs). The purpose of the study was to determine the capacity and efficiency of two-lane roads in several parts of India. After gathering initial data from observing traffic at various two-lane roads throughout the areas of study, models were built and the computer simulation constructed using the Visual Basic programming language.

Once the computer model could accurately predict traffic patterns, road capacity, and efficiency for observed situations, the authors manipulated various factors of the model to create hypothetical situations, enabling them to determine the major agents of change in traffic patterns and road capacity. Specifically, the authors adjusted the model to include more cars of certain classes of vehicles. They found, in part, that all-tractor simulations showed an unsurprising decrease in road capacity and efficiency, while simulations that included only two-wheeled single- or double-passenger vehicles achieved the highest capacity.

One of the more surprising results that the authors came to in this study was the fact that capacity decreased when the road became one way; that is, when the model was adjusted such that traffic only flowed in one direction along both lanes. This is at least partially due to the fact that passing is actually achieved more efficiently when the lanes are moving in opposite directions; this ensures that slower drivers will not be holding up both lanes. Any move away from a fifty-fifty distribution in the direction of travel resulted in a linear decrease in road capacity and efficiency.

Personal Opinion:

found this article to be incredibly detailed in its organization and in the controls the authors used in building their model. The ingenuity of their variable assignments, especially in regards to the PCU classification of vehicles rather than a simple size classification added the refinement to their set-up that was necessary for achieving reliable and accurate results. The fact that, as the authors put it, "lane discipline is not strictly observed in India," was well-accounted for in a project that attempted to know the efficiencies and capacities of the road during periods and lengths of exclusive and rigidly-defined directional travel for each or both lane(s).

That being said, very little of what the researchers found was of great import, nor was it incredibly earth shattering. Especially when using time as a measure for capacity (how many vehicles are able to pass a given point in an hour was the measure used by the researchers), it is not surprising to find that there is a lower capacity when larger, slower vehicles are using the road, and a much higher capacity when the smallest and most maneuverable crafts are the only vehicles on the road. The research was sound, however, and though the facts themselves might not be useful on the surface, they do underline the human component of all engineering pursuits.

This study shows that two-lane roads would be most efficiently used by assigning strict directional rules, with one lane going one way and one lane going the other. This is not how people in India drive, however -- it only worked that way in the author's computer model. The authors failed to suggest ways in which roads or signage might be designed and/or implemented to encourage drivers to make this a regular practice.

Speed and Lateral Vehicle Position Models from Controlled Nighttime Driving Experiment

By Stodart, B., and Donnell, E.

This article's main goal was to study an aspect of the way that engineers study drivers. That is, the authors here conducted an observational study to determine if there was a correlation between various aspects of driver behavior that are generally measured and scrutinized separately in engineering models. Specifically, the author's observed the relationship, if any, between speed and lateral position of vehicles on a stretch of two-lane highway in Pennsylvania during nighttime hours. Observed behaviors were then modeled using a variety of mathematical regressions, some of which proved more accurate in describing behaviors than others.

The use of multiple regressions was used as a way to determine whether or not a relationship indeed existed between the speed at which a driver traveled and their average lateral position to the vehicle in front of them. If the same regressions remained fairly accurate in describing both speed and lateral position, than a correlation between the two driver behaviors could be said to exist. The results of this study found, however, that at least within the confines of the observed experiment speed and lateral position were entirely unrelated. Speed was best modeled by a single-equation random effects regression, whereas the least-squares regression model was more accurate in describing the lateral position of vehicles.

The goal of this research had been to determine if engineers were correct in analyzing these behaviors separately and implementing engineering solutions and geometric designs in order to influence driver behaviors such as speed and following distance. The hope was that the discovery of a correlation would lead to increased effectiveness in engineering-based behavior modification, but this research up held the current thinking that these behaviors are indeed uncorrelated. The authors did find a correlation between the roadways itself and speed, however.

Personal Opinion:

The research conducted in this study had a focus that was far too narrow for the results to be widely applicable. Though the nighttime two-lane highways was used as a way of controlling the complexity and sheer number of variables in a real-world, observed experiment, it also severely limited the study's findings and the scope of its impact. The authors acknowledge this fact, and they did not let it deter them from extracting some useful information from their data, but a broader study might have provided more accurate and interpretable data.

The author's also seem to at least partially make the misstep of assigning too much influence to engineering elements of the roadway. In their conclusion, they acknowledge that individual drivers are the primary determiners of both speed and lateral difference; there is much more likely a psychological correlative between various aspects of driver behavior than an engineering one.

Asphalt Material Characterization in Support of Mechanistic-Empirical Pavement Design Guide Implementation in Virginia

By Flintsch, G., Loulizi, A., Diefenderfer, S., Diefenderfer, B., and Galal, K.

The constituent components of asphalt and other paving materials are of huge import to the engineering aspects of road design and maintenance, and it is for that reason that the authors of this paper subjected several mixtures paving materials to a series of tests as laid out in the new procedural guidelines as laid out in the Guide for Mechanic-Empirical Design of New and Rehabilitated Pavement Structures. Called the MEPDG for brevity's sake, the methods of paving characterization outlined here was used by the authors to determine the appropriateness of certain mixtures of hot-mix asphalt as a paving material in Virginia. The results were also compared to certain predictive equations to determine the accuracy of current models.

Samples of eleven different asphalt mixtures were obtained from various plants across Virginia, and each was analyzed to determine its components and the percentages of their appearance in the mixture, as well as their response to certain stress tests as described in the MEPDG. Dynamic modulus, the strength of the various mixtures under vibratory stress, is considered by the authors and the MEPDG to be the most important diagnostic test used to characterize the different asphalt mixture as it most closely approximates the stress experienced by paving materials in real-world situations. After conducting tests to the determine specific gravity and asphalt content of each mixture, samples of the loose mixtures were compacted and subjected to test for dynamic modulus, tensile strength, and creep compliance.