Debris flow risk assessment based on machine learning theory
Jiayang Xu, Qizhi Wang, Yujing Fan, Yingying Ye, Fuji Gu, Ruitao Han · 2023
Taking the debris flow in Chongli County of Hebei Province as the research object, based on the field investigation, this paper analyzed the formation conditions, material source characteristics, trigger mechanism, development stage, and susceptibility of debris flow in the Chongli district. According to the characteristics of debris flow gullies in the study area, the risk and vulnerability evaluation indexes were selected. On this basis, forty-nine debris flow gullies were trained by machine learning, and a debris flow early warning model was constructed. The model was based on six main control factors and trained by a support vector machine to obtain a linear regression equation, which can quickly output its risk level according to the current state of debris flow gullies.