Researchers in Granada Spain have used real-time wireless tracking nodes to capture, analyze and predict future traffic flow, establishing predictive technology to be used by Granada’s Mobility Office of the City Council.
Several nodes of the so-called MOBYWIT system were deployed at important urban points to collect real-time vehicle movement. The data was processed into Origin/Destination matrices and accurate displacement times and estimation of real traffic flow via Time Series Forecast is computed.
Four different forecasting techniques have been applied to test the quality or value of the gathered data for performing useful predictions. These include Linear Regression, MLP (Multi-Layer Perceptron), NARX (Nonlinear AutoRegressive with Exogenous (external) Input) and NARX-CLN (Closed Loop Network).
The results validate the system as a tool for urban traffic flow monitoring, analysis and prediction, which could be used as a part of an intelligent transportation system, the research team says.
The research presents “a complete study of a real urban scenario” and the datasets allow for analysis and prediction of traffic flows for both direction and traffic volume. The researchers believe their forecasting system is able to anticipate traffic congestion and the data can be used to manage traffic signals.
READ the Research Paper HERE …
(Photo José Sáez, Flickr)