Traffic flow modeling in transportation networks

One of the main urban problems is congestion. To mitigate its effects, it is necessary to know how traffic flows are distributed in the cities’ road networks.

This project addresses the modeling and analysis of traffic flows in transportation networks. The main research problem is the estimation of source-destination flows, also known as OD matrix estimation problem. The objective is to estimate the origins and destinations of the traffic observed on a road network, and thereby better planning interventions in the transportation system and forecasting the temporal evolution of traffic.

For the modeling of traffic, we use **state space models **, which are stochastic dynamic models whose inference is made from a point of view view of Bayesian statistics. Additionally, we have been using tools from data science to analyze traffic data obtained through sensors installed on the city road network from Fortaleza, Brazil.

The project has the collaboration of Prof. Luis Carvalho, from the Department of Mathematics and Statistics at Boston University.

This project has financing from Conselho Nacional de Desenvolvimento Científico e Tecnológico - CNPq , under grant number 422464/2016-3.