A hierarchical approach for dynamic origin-destination matrix estimation on large-scale congested networks

In this paper we propose a hierarchical approach for decomposing and simplifying the dynamic OD estimation procedure for large-scale congested networks. The main idea is to subdivide the network into multiple hierarchical levels. Next an OD estimation is performed on each level separately, starting with the highest level, and the output of this estimation is used as input for the OD estimation on a lower level. The main advantage of this approach is that different levels of complexity (of the DNL/DTA model and of the estimation method) can be used for different parts of the network as required. In addition, this hierarchical approach solves many practical and theoretical limitations of traditional OD estimation methods, which have been identified in previous research.