Simulation of Crowd Evacuation under Toxic Gas Incident Considering Emotion Contagion and Information Transmission

Abstract

Evacuation can be the most necessary and effective method to save human lives in emergency scenarios due to natural or human-caused disasters. The emergency egress behavior of humans is not only affected by damage caused by danger sources directly, but also depends on time-varying internal states of evacuees, such as their familiarity with the environment, the information perceived, and the emotional responses triggered by the incident and casualties. This study presents a method of crowd evacuation simulation under a toxic gas incident considering the effects of emotion contagion and information diffusion, which consists of four main parts: the gas dispersion model, the information diffusion model, the emotion contagion model, and the modified social force model. First, the gas dispersion model characterizing the concentration of the toxic gas is used to determine the impairment of the pedestrian. Second, the spread of the information about the incident and the exits is modeled by the information transmission model. The emotion contagion model was adopted to quantify the emotion and panic. Then the social force model is modified to integrate these effects to generate the movements in the evacuation process. Numerical simulations show that the incorporation of emotion contagion and information transmission can cause significant differences, as compared to existing models, in terms of evacuation time, number of casualties, and number of collisions. The influences of the number of evacuees, the perception radius, and the number of authority figures on the evacuation site are examined in the parametric study.