Geological disaster multi-disaster comprehensive risk evaluation method based on random forest

Abstract

The invention provides a geological disaster multi-disaster comprehensive risk evaluation method based on a random forest. The geological disaster multi-disaster comprehensive risk evaluation method comprises the following steps: (1) constructing a geological disaster risk evaluation system; (2) performing sample selection and risk grading based on geological disaster detailed survey points; (3) performing random forest model construction and sample training; (4) carrying out model precision evaluation and index contribution degree calculation; and (5) obtaining a geological disaster risk evaluation result. According to the method, from the perspective of multiple disaster types, common influence factors of the multiple disaster types of geological disasters are selected, an index system is constructed from the dangerous and vulnerable dimensions, subjective weighting and subjective threshold delimitation are avoided, and the method has good adaptability to comprehensive geological disaster risk evaluation of the multiple disaster types. And the evaluation result can better reflect the actual disaster risk level of the region. According to the point-to-surface disaster risk evaluation and distribution method provided by the invention, main control factors of geological disaster risks can be effectively identified, and regional geological disaster risk grading and zoning are scientifically realized.

Description

technical field [0001] The invention relates to the technical field of geological disasters, in particular to a random forest-based multi-hazard comprehensive risk assessment method for geological disasters. Background technique [0002] Geological hazards are one of the important types of natural disasters, including landslides, debris flows, collapses, ground subsidence, ground cracks, etc. Their occurrence directly leads to the destruction of infrastructure, property loss, and even life-threatening. In recent years, with the increase in the frequency of extreme climate events, the impact of strong regional earthquakes, and the intensification of engineering construction disturbances, the frequency and harm of natural disasters in my country, especially geological disasters, have increased significantly. For example, in 2010, the "8.17" catastrophic mudslide disaster in Zhouqu, Gansu, and the "8.20" heavy rainfall in Wenchuan, Sichuan, the catastrophic torrential debris f...

AI Technical Summary

Problems solved by technology

For example, a lot of data preprocessing is required before using the decision tree, and it is easy to fall into local optimum; the support vector machine is inconvenient to apply because of its complex mathematical functions, and it is not enough to solve multi-classification problems; the artificial neural network model has over-learning , local minimum and slow convergence, etc.

Images