Forecasting travel: An agent-based approach
Abstract:
Since the 1950’s transport engineers and planers have been asked to assess the impacts of investments and systems changes on the patterns of travel. They were also asked to do so for very long time horizons to support cost-benefit analyses of major investments. The original models were based on physical analogies, which were later reinvented through the integration of utility maximizing choice models. The advances in computing allowed the shift of higher spatial, temporal and behavioural resolutions. The talk will present MATSim, a popular large-scale agent based simulation framework, one typical example of this last generation of transport forecasting models. MATSim is an open-source project mainly based at TU Berlin and ETH Zürich. Typical use cases are large scale metropolitan areas, e.g. Singapore, LA, Schenzen, but national models, say for Germany as a whole, are feasible. A brief detour will introduce an aggregate alternative, the macroscopic fundamental diagram, as a contrast.
Bio:
Kay W. Axhausen is Professor and chair of the Transport Planning at the ETH Zurich. Before working at ETH, he worked at the University of Innsbruck, the Imperial College London and the University of Oxford. He holds Master of Science degree from the University of Wisconsin–Madison as well as Ph.D. in civil engineering from Karlsruhe Institute of Technology.