Etiological study on forest fire accidents using Bow-tie model and Bayesian network

Forest fires will do great harm to the ecological environment. At present, the most important research field is the study of the causes of forest fires, and the failure to determine the causes will greatly hinder the prevention, control and safety management of forest fires. In order to solve this problem, this paper constructs a Bayesian network to calculate the state probability and sensitivity after determining the factors related to forest fire. Secondly, on the basis of constructing Bayesian network, the Bow-tie model is constructed according to the consequences of forest fire. The constructed Bow-tie model contains 7 influencing factors and 5 consequences, and 26 corresponding control measures. Finally, through the analysis of Bayesian network and bow tie model, the forest fire prevention system is constructed. According to the calculation results of the Bayesian network, the probability of forest fires being at high risk of occurrence is higher for three factors: agriculture, forestry and water expenditure, rural population density and precipitation. Changes in rural population density and humidity have the greatest impact on the occurrence of forest fires. In combination with the Bow-tie model, human factors are the most important factors leading to the occurrence of forest fires, and the material, environmental, and management factors are the secondary factors. Based on the forest control system, the prevention and management measures are put forward. This research can provide reference for the control of forest fire and reduce the occurrence of similar forest fires.