Behavioral effect on pedestrian evacuation simulation using cellular automata

Abstract

A pedestrian evacuation simulation model based on the extended cellular automata is proposed with the consideration of heterogeneous behavioral tendencies in humans. We investigated two behavioral tendencies, namely, familiarity and aggressiveness. Our simulation experiments tested the values of the behavioral parameters and the pedestrian flow composition. The model is calibrated and validated by using hand calculation, fundamental diagrams of traffic flow and evacuation experiments. The results show that the evacuation time decreases exponentially with an increase in the evacuees’ site familiarity and increases when pedestrians are either too conservative or too aggressive. The optimum evacuation was associated with higher familiarity and no aggressive preference. Altering the conservative compositions of pedestrian groups showed that there is a turning point for each density, depending on the proportions of conservative pedestrians. Prior to this point, the evacuation time is a relatively constant value. After this point, the evacuation time increases and fluctuates. This finding demonstrates that at high densities, a few leaders in pedestrians can significantly reduce the evacuation time. The model is able to simulate evacuation effectively given the diverse behavior strategies of evacuees, such as herding and communication between evacuees. In addition, the model can explain some basic principles of fundamental diagrams of traffic flow. Furthermore, the model is applied to optimize buildings to reduce evacuation times. Case studies show that the evacuation time can be dramatically decreased if a building is optimized to reduce interactions between different groups of pedestrians by separating flows or expanding the exit width. However, such strategies are effective only under high pedestrian densities. Placing signs and guides is also helpful for improving the evacuation performance. We conclude that behavioral tendencies have a strong influence on evacuation time.