| dc.description.abstract | Some approaches have been proposed by literature to describe the traffic state for a network, such as kinematic wave theory (using concepts from Physics), cell transmission models or macroscopic traffic simulation models. However, many of them have severe limitations regarding traffic state change or require a lot of computation time. For this reason, researchers have been examining for last years the existence of a simple and fast way that can sufficiently describe the dynamics of a road network. As a result, the concept of the Macroscopic Fundamental Diagram (MFD) - an object (empirical relation, theoretical model or both) that relates the average flow to the average density of a network, capturing so the essential network situation - was developed. Once the MFD of the network is known, all that is needed to have a traffic state estimation is to locate where the system is on the MFD at any desired moment, so it serves as a fundamental object for macroscopic traffic flow models. These family of models allow describing the spatio-temporal evolution of traffic density, for instance, and lead to clever solutions that optimize the existing traffic system. Thus, the objective of this project is to present a method for obtaining a network MFD using bus GPS data and a data structure developed by Uber (Uber’s H3 Hexagonal Hierarchical Spatial Index). We use a raw data collection of latitude and longitude data points of buses in Rio de Janeiro, Brazil, from January 2018 to December 2018. It is worth mentioning that the resulting MFD of the proposed method serves as a basis to support the development of public transportation management systems, which is able to make accurate traffic state predictions. The findings confirm the usefulness of bus GPS data and Uber H3 structure in finding a Macroscopic Fundamental Diagram, especially the Density-speed one, and future research directions are addressed. | |
