Converting a Static Macroscopic Model Into a Dynamic Activity-Based Model to Analyze Public Transport Demand in Berlin
Transport models demanded by public transport companies today should not only deliver the basis for future planning of the regional transport system, but also provide detailed information about passenger flows of different user groups. This paper presents the successful transformation of a static macroscopic model (built using PTV VISUM) into an integrated activity based demand and dynamic assignment model (MATSim) performed for a real application in the Berlin/Brandenburg metropolitan region. While the two models clearly differ in their methodology, overall key values can be reproduced showing similar results. It is shown that by the use of the activity chain distributions and their timing activity based demand can be reproduced with respect to the trip distribution of the origin-destination matrices from the macroscopic model. The process flow defined in this paper allows to use both models for planning purpose, case studies and effect analysis, enabling public transport companies to analyze effects on the macroscopic level of detail as well as on the agent based level to capture specific customer groups and/or time ranges during the day. The microscopic model is then used for further analyses, of which a selection is presented in this paper. Notably, the model allows for researching effects generated by the interaction of public transport vehicles and regular private car traffic, or for researching user-group specific behavior.
Published in: Proceedings of the 13th Conference of the International Association for Travel Behaviour Research (IATBR), International Association for Travel Behaviour Research