REAL-TIME ESTIMATION AND PREDICTION OF ORIGIN-DESTINATION MATRICES PER TIME SLICE

As part of the DRIVE-II project DYNA (Dynamic Network Assignment), an approach for the real-time estimation and prediction of origin-destina tion (OD) matrices per time slice has been developed. This method is u seful to support real-time traffic management of a freeway network. Th e proposed indirect approach combines historical information with traf fic counts at on-ramps. It has been tested on the Italian freeway A4, where the existing pay toll system enables us to measure the real orig in-destination matrices. The results show that these prediction proced ures can perform almost as well as direct prediction methods based on knowledge of the demand.