基于传染病机制的突发事件下群体情绪感染模型

Frequent emergencies have the characteristics of suddenness, high uncertainty, persistent and destructiveness, and always bring massive casualties, while practice has showed that the generation and spread of crowd panic emotion is an important cause of those casualties. Therefore, the modeling and simulation of crowd emotional contagion under emergencies have extensive demands for social applications. Nevertheless, its accuracy and practicability also face great challenges, including complicated emotion theories, emotional contagion mechanisms, inadequately considered contagion models, and non-intuitive simulation for crowd emotional contagion models. Based on the epidemic mechanism of SIRS (Susceptible-infected-recovered-susceptible) model, this paper proposes an improved crowd emotional contagion model E-SIRS, which takes the consideration of the influences of crowd environment, individuals and social factors on individual emotional states in emergency under self-organization conditions. This proposed model classifies the crowd into three states according to their panic level: susceptible state, infected state and recovery state. Besides, this paper also establishes the state transition relation during the spread of panic and its quantitative expression. And then, from the perspective of complex dynamical systems, the Lyapunov method is utilized to analyze the stability of this model. There is a non-zero threshold during the spread of crowd panic emotion in homogeneous network. When parameter is less than the threshold, panic emotion will finally disappear. When parameter is greater than the threshold, the proportion of panic people will eventually converge a non-zero value. Finally, the NetLogo tool is used to build a dynamics simulation model to observe the number of trends for crowd emotional contagion. The results showed that E-SIRS can describe the dynamic changes of emotion of panic crowd comprehensively, and provide the scientific basis for real-time monitoring and early warning of mass incidents, especially, unexpected group incidents in the future.