Research And Construction Of Forest Fire Risk Prediction System In Yunnan Province Based On Spring Boot

Yunnan Province is a large province of forest resources in China,and it is also one of the areas where forest fires occur frequently.Forest fires are very sudden and destructive,and timely and effective forest fire risk prediction is an important task for forest fire prevention in this area.In order to provide scientific basis for forest fire prevention and control,it is necessary to research and build the forest fire risk prediction system in Yunnan Province.This paper introduces the development technology,prediction model and the process of research and construction of forest fire risk prediction system in Yunnan Province.The prediction methods include Poisson distribution model,negative binomial distribution model,zero expansion negative binomial distribution model and water content model of small combustible materials.Based on the consideration of reference factor categories in long-term prediction of forest fire risk,the prediction of forest fire risk is divided into time dimension and administrative division dimension.In the annual scale and monthly scale of time dimension,meteorological factors are selected as independent variables,and Poisson distribution model and zero inflation negative binomial distribution model are used to predict forest fire risk data.In the annual scale and monthly scale of the administrative division dimension,meteorological factors,road density,agricultural population density and other factors are selected as independent variables,and the negative binomial distribution model and zero expansion negative binomial distribution model are respectively used to predict the forest fire risk data.In the two dimensions of daily scale and real-time scale,meteorological factors,combustible and other factors are selected as the independent variables of the moisture content model of small combustible to achieve the prediction of forest fire risk data.Based on the understanding of the way of receiving forest fire risk prediction information from Yunnan Forestry Bureau,this paper introduces the related technologies and implementation methods used in the research and construction of forest fire risk prediction system in Yunnan Province,including Spring boot integrates the development framework of Paoding Rose Jade,API call of real-time meteorological data,construction of prediction model database,access mechanism and call of model database,embedding and data display of ECharts graphic database,etc.Through to the Yunnan province of forest fire danger forecasting system requirement analysis,system design,coding,and testing of the entire process,it has realized the function of displaying the historical data of forest fire in the past 20 years and the future data of forest fire prediction in the form of charts or maps in Yunnan Province.The time scale can choose annual scale,monthly scale and daily scale to fully ensure the timeliness of forecast data.The regional level can choose provinces,cities and counties.Through the combination of time dimension and administrative division dimension,the fire protection business needs of users’ multi angle statistics,analysis and prediction are fully met.It can be used for reference for research and development in this field.