Dynamic Estimation of Transport Demand: Solutions — Requirements — Problems

The following contribution presents requirements, difficulties and first attempts at modelling a temporal short interval estimation of transport demand. For the simulation of motorway traffic (for an area within the state of North-Rhine-Westfalia, NRW) methods of temporal disaggregation of existing trip-matrices have been worked out within the framework of the Northrhine-Westfalian research cooperation for traffic simulation and environmental impacts “NRW-FVU” (Nordrhein-Westfälischer Forschungsverbund Verkehrssimulation and Umweltwirkungen). To simulate urban traffic (example: Wuppertal) methods of estimating temporal short interval trip-matrices were conceived and tested. The matrices were supplied for microsimulation with Cellular Automat (CA) and for the dynamic route choice and traffic assignmemt (DRUM, Dynamische Routensuche and Umlegung). The comparison of both methods, based on the estimated link loads (ADT and hourly loads), supplies deviations ranging within the mean variation of counted values. It therefore can be inferred that these methods, which are different with regard to computing intensity and data requirements, should be used depending on the tasks and the intended precision of the results. The specific pros and cons are important operational criteria. It also becomes obvious that in future methodical advancements should be examined on the basis of activity(-chain)-based approaches.