Research on public security risk assessment of emergencies based on scene coupling driven

The security risk in the process of emergency occurrence and development is composed of the interaction of various risk elements, which have many distinct characteristics different from the normal. This paper constructs a public security risk prediction model adapted to scene coupling drive by combining with the risk interaction coupling characteristics of HHM-RFRM theory. The qualitative, quantitative filtering, rating and risk assessment of public security risk scenarios are carried out by using Bayesian theorem and model. Combined with the actual data of multidimensional risk scenario, the coupling relationship is effectively analyzed to realize the transition from "single risk" to "coupling risk" early warning. It is found that the method has strong consistency with the actual data, the evaluation accuracy is further improved, and it has stronger adaptability to the security risk of emergencies evolution.