A Cellular Automaton Model for the Traffic Flow in Bogota

TL;DR

This paper introduces a cellular automaton model tailored for Bogota's traffic, accurately reproducing experimental flow data and shock wave behavior, emphasizing the importance of city-specific parameters.

Contribution

The model incorporates hysteresis, delay, and instant braking, with parameters derived from direct measurements, improving traffic flow simulation accuracy for Bogota.

Findings

Model matches experimental fundamental diagrams.

Reproduces shock wave phenomena.

Identifies boundary conditions as sources of artifacts.

Abstract

In this work we propose a car cellular automaton model that reproduces the experimental behavior of traffic flows in Bogot\'a. Our model includes three elements: hysteresis between the acceleration and brake gaps, a delay time in the acceleration, and an instantaneous brake. The parameters of our model were obtained from direct measurements inside a car on motorways in Bogot\'a. Next, we simulated with this model the flux-density fundamental diagram for a single-lane traffic road and compared it with experimental data. Our simulations are in very good agreement with the experimental measurements, not just in the shape of the fundamental diagram, but also in the numerical values for both the road capacity and the density of maximal flux. Our model reproduces, too, the qualitative behavior of shock waves. In addition, our work identifies the periodic boundary conditions as the source of false peaks in the fundamental diagram, when short roads are simulated, that have been also found in previous works. The phase transition between free and congested traffic is also investigated by computing both the relaxation time and the order parameter. Our work shows how different the traffic behavior from one city to another can be, and how important is to determine the model parameters for each city.