Dynamic properties of grid urban networks

Abstract

Urban grid networks might present different ways to structure streets and consequently, traffic, might be accommodated differently. In the present paper, we analyze three grid configurations: two-way streets, two-way streets with prohibited left turns, and one-way streets. We simulate traffic in them with the micro simulation software VISSIM and we distribute trips employing the dynamic traffic assignment integrated in it. In this way, we are able to analyze the dynamic behavior of traffic. Typically, in static simulations, oversaturated links or intersections have really large delays but these delays, contrary to what happens in reality, do not affect other links. We are interested in understanding which configuration is more resilient to the propagation of congestion, and this cannot be seen when employing static models. In addition, the static simulations provide macroscopic indicators such as the total travel time or the total travel distance. However, the dynamic perspective provides a new set of indicators to evaluate our networks, and we are especially interested in using the Macroscopic Fundamental Diagram (MFD) as one of them. The MFD is a traffic monitoring scheme that determines the upper bound of network traffic states relating trip production rate and density of trips. Our goal is to generate MFDs in the different networks and to employ it as indicators of the network performance. Depending on the MFD shape, networks might have different properties when dealing with traffic. In this paper, we discuss this issue by comparing the three different urban grid configurations.