The Automatic Generation of an Efficient Floor Field for CA Simulations in Crowd Management
Mohcine Chraibi, Bernhard Steffen and Jun Zhang
Realistic simulations of pedestrian dynamics are an important tool to assess critical evacuation scenarios during the process of an event’s management. However, for practicability reasons, the software has to run on available computers faster than real time. Cellular automata models are compared to space-continuous models very fast and hence are a good candidate for this use case. However, due to its limited geometry resolution, they have well-known problems with treating corners and turns. This present paper shows how a standard CA model can be modified to produce a realistic movement of people around bends and obstacles by changing the standard floor field, using simple predictions for the immediate future. By adaptively allowing for the temporary situation using simple predictions for the immediate future. The approach has two tuning parameters that have a distinct meaning and can, therefore, be set correctly. A further application of the proposed method is the automatic generation of guiding lines for any space-continuous model.
Keywords: Cellular automata, modeling, pedestrian dynamics, lanes at corners