Authors: Nick Malleson*, University of Leeds, Tomas Crols, University of Leeds, Jonathan Ward, University of Leeds, Andrew Evans, University of Leeds
Topics: Geographic Information Science and Systems, Quantitative Methods
Keywords: agent-based modelling, data assimilation, data-driven simulation, urban mobility, ensemble Kalman filter.
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Bayside A, Sheraton, 4th Floor
Presentation File: No File Uploaded
The 'data deluge', coupled with related 'smart cities' initiatives, have led to a proliferation of models that capture the current state of urban systems to a high degree of accuracy. However, they are typically 'black box' in nature, which makes it difficult to use them to better understand the fundamental dynamics that drive the systems under study. Furthermore, many example models of 'smart cities' appear limited in their ability to *forecast* future system states. Agent-based modelling (ABM) is well suited to modelling urban systems, but is limited in its ability to incorporate up-to-date data as they arise in order to reduce uncertainty in model forecasts. This limits the usefulness of ABM a forecasting tool. This paper presents ongoing work that adapts data assimilation techniques from fields such as meteorology (specifically the 'ensmble Kalman Filter') in order to allow agent-based models to be optimised using streaming data in real time. Here, a simple example of an agent-based model used to simulate the movement of people as they travel along a street is illustrated. Importantly, the model is optimised dynamically in response to hypothetical data streams. The paper will also demonstrate some important secondary uses of the technique as a means of capturing model uncertainty. Ultimately, we work towards a combination of ABM and data assimilation methods that will be able to assimilate streaming ‘smart cities’ data into models in real time.