Determining dynamic origin-destination matrices in urban networks

Abstract: The article presents methods for determining origin-destination matrices, with a particular focus on the dynamic approach to the problem. Due to the potential for applying methods of determining origin-destination matrices based on traffic volumes on transport network links in dynamic traffic control and management, an attempt was made to generalize the computational process in a schematic way. The generalized scheme can serve as a basis for developing a modular concept containing a series of partial models. To streamline calculations, an original IT tool was developed.1. INTRODUCTION
Analyses related to traffic flow planning, traffic forecasting, and designing organizational changes require knowledge of origin-destination matrices.
They reflect transport demand in terms of movement relations.
In effective traffic management, when knowledge of the travel destinations of individual road users is necessary in critical situations (accidents, road works, breakdowns, or other unusual situations), so-called dynamic origin-destination matrices, which are updated at specific time intervals, are often used. Only on this basis can optimal detour routes be determined in urban networks.
Individual cells of a two-dimensional origin-destination matrix represent the volume of traffic flow expressed as the number of trips made between a pair of transport zones. The centroids determined for each zone represent the accumulation point of the zone’s departure and arrival potential. In practical solutions, these centers are usually transferred via so-called connectors to the nearest nodes of the technical network (e.g., road, rail, etc.). Therefore, determining the origin-destination matrix between two zones can be reduced to determining the traffic volume between two nodes of the technical transport network.
Origin-destination matrices can be built for individual modes of transport, for specific time intervals, for different travel motivations, and for groups of road users. These matrices can be determined for the existing state or for a future state (traffic forecasting). Individual elements of the origin-destination matrix can be expressed as values of the number of trips or as percentage shares of the flow moving from a given zone in various directions. (…)

Determining dynamic origin-destination matrices in urban networks – PDF file for download