Model-based traffic monitoring by means of neural networks

This article is about traffic monitoring by the use of neural networks. Magnetic sensors are used in order to extract online the flow and density variables. Such variables are processed by the monitoring network in order to detect and isolate incidents that disturb the traffic. Our approach is based on a macroscopic model, and more precisely on the fundamental flow-density diagram that characterizes the traffic. An admissible region is defined in the flow-density space from this diagram. Incidents are detected when the measured data are out of the admissible region. The classification properties of neural networks are used to design the monitoring network. The proposed method is applied to the monitoring of a complex road junction in the city of Nancy in France.