An extended model for crowd evacuation considering pedestrian panic in artificial attack

An extended floor field cellular automata model considering pedestrian panic was proposed to investigate pedestrian dynamics under the threat of artificial attack. In the model, a new parameter D was introduced to reflect quantitatively the extent of the impact of panic on pedestrian evacuation. Besides, pedestrian movement is governed by the static, attack threat and herding floor field. And some modifications were made to introduce the influence of panic on the decision-making of the individual. Further, the motion and assault of the artificial attacker and the death of pedestrians were taken into account simultaneously. The simulation results show that the proposed model can reproduce some real characteristic features of escape panics, such as (1) Pedestrians become less rational. (2) At exits, arching and clogging are observed. (3) People show a tendency towards mass behavior, i.e., pedestrian herding. (4) Alternative exits are overlooked or not efficiently used. As the extent of the panic increases, pedestrians spend more time to escape from the room and are more sensitive to the movement of the crowd and the action of the attacker. The phenomenon of pedestrian herding and jamming can be clearly observed as D increases. And the results also indicate that the tension caused by moderate panic can help reduce casualties, whereas excessive panic will increase casualties. Besides, enhancing pedestrian initiative, i.e., the ability that pedestrian finds the exit independently in panic, is beneficial to the evacuation. In addition, more pedestrians would lose their lives due to attack and the evacuation efficiency would be reduced if the attacker starts the attack from the positions of the exits, so the security forces near the exits should be strengthened.