A continuous floor field cellular automata model with interaction area for crowd evacuation

Abstract Discrete and continuous models are generally used to study the evacuation dynamics. Considering the low accuracy of the discrete model such as floor field cellular automata model (FFCA) and the high time complexity of the continuous model such as social force model (SFM), a continuous FFCA (CFFCA) model is proposed. Pedestrians no longer occupy cells but are treated as particles in continuous space. To reduce the time complexity caused by continuity, the social force is replaced by interactions by drawing on SFM from different perspectives. The obstacle and pedestrian interaction area are proposed to simulate the repulsion, friction and attraction between pedestrians and obstacles, and between pedestrians. By simulating different evacuation scenarios, it is found that the simulation results are close to the experiments data. CFFCA evacuation path is more reasonable than that of FFCA and the execution time is shorter than that of SFM. Because of the superiority of interaction area used in CFFCA, pedestrians can move freely without being restricted by the spatial scale of cells, and the complex psychology of pedestrians can be simulated under lower time complexity. Our study may be useful to find the trade-off between model realism and tractability.