Method and device for constructing a forest fire danger prediction model

Abstract

The invention provides a method and device for constructing a forest fire danger prediction model. The method comprises the steps of acquiring the historical natural forest fire danger information, historical meteorological data information and vegetation type information of a target area; filtering the data records with the missing items in the obtained information to obtain filtered information;utilizing a logistic function to carry out smoothing processing on the filtering information to obtain smooth information; performing normalization processing on the smooth information to obtain normalization information, and enabling the normalization information of the historical natural forest fire danger when the fire danger occurs to correspond to the normalization information of the historical meteorological data when the fire danger occurs; and by taking the historical meteorological data normalization information and the vegetation type information as input and taking historical natural forest fire danger normalization information mapped by the historical meteorological data normalization information as output, training the long-short time memory method model until the trained model meets a preset precision requirement, and obtaining a forest fire danger prediction model. And the prediction precision of forest fire danger can be effectively improved.

Description

technical field [0001] The present application relates to the technical field of fire risk prediction, in particular, to a method and device for constructing a forest fire risk prediction model. Background technique [0002] Forest resources are one of the most important natural resources on the earth. On the one hand, forests are the main components of the earth's ecosystems, and have the functions of regulating climate, producing oxygen, preventing wind and sand, eliminating noise, and purifying the environment. On the other hand, forests also It provides a large number of resources for people's production and life, for example, it provides people with various forestry products and plant processing by-products such as raw wood, pulp wood, and artificial boards. Therefore, forest protection has irreplaceable economic, ecological and social significance. Among them, how to prevent natural forest fire hazards is a key research field of forest protection. [0003] At present...

AI Technical Summary

Problems solved by technology

[0009] However, in this forest fire risk prediction method, since each fire risk factor generally has multi-dimensional complexity and each fire risk factor is related to each other, the forest fire risk prediction using the logistic model of linear regression has certain limitations, and the accuracy of the prediction results needs to be improved. ; Furthermore, for big data in the Internet era, the logistic model of linear regression cannot meet the analysis and processing requirements of big data

Images