Traffic state estimation based on data fusion techniques

The capability to detect and/or forecast traffic conditions is of utmost importance in road management applications. Recent advances in technology have made available numerous new monitoring systems exploiting larger fleet of probe vehicles. Together with traditional volume and time mean speed measurements relative to a local section monitored continuously in time, probe vehicles provide additional type of data, such as space mean speed and travel time, relative to road segments monitored in specific time intervals. Therefore, the aim of this paper is to study how to exploit available information detected by new monitoring devices in the estimation of traffic flow conditions. Different types of data fusion techniques have been analyzed, namely measurement data fusion and state vector fusion, in several simulations carried out on a simple test network, traveled by probe vehicles and composed of 9 cells with an on ramp and an off ramp and with two fixed traffic sensors located in two different cells. Test results are promising and indicate higher accuracy of estimates obtained with new methods, particularly in the case of measurement data fusion.