Advanced traffic data for dynamic OD demand estimation: the state of the art and benchmark study

Abstract

In this paper, the use of advanced traffic data is discussed to contribute to the ongoing debate about their applications in dynamic OD estimation. This is done by discussing the advantages and disadvantages of traffic data with support of the findings of a benchmark study. The benchmark framework is designed to assess the performance of the dynamic OD estimation methods using different traffic data. Results show that despite the use of traffic condition data to identify traffic regime, the use of unreliable prior OD demand has a strong influence on estimation ability. The greatest estimation occurs when the prior OD demand information is aligned with the real traffic state or omitted and using information from AVI measurements to establish accurate and meaningful values of OD demand. A common feature observed by methods in this paper indicates that advanced traffic data require more research attention and new techniques to turn them into usable information.