Evaluation Of Landslide Susceptibility Based On BP Artificial Neural Network And GIS In Dingcheng District,Hunan Province

The phenomenon of geological disasters in China is relatively serious,with a large number,wide distribution,frequent occurrence and strong destructive characteristics,which has a great impact on human living environment,national economic development and infrastructure construction.Due to the continuous development of our country in recent years,the construction of infrastructure is increasing day by day,which leads to the increasing frequency and intensity of geological disasters.In all the types of geological disasters,the most important type is landslide.Landslides have the characteristics of centralized distribution and mass occurrence.Therefore,the hazards caused by landslide geological disasters are very serious.Landslide susceptibility evaluation is the basic work of landslide prevention and control,which provides a reliable guarantee for infrastructure planning and human life safety.Dingcheng District is located at the boundary of Changde City,Hunan Province.Due to the large population and abundant engineering activities,the landslide has a strong impact.It is of great significance to evaluate the landslide susceptibility of Dingcheng District for the local development.In this paper,through the geological data collection and data extraction in the study area,the influencing factors of landslide susceptibility are sorted out.Based on BP artificial neural network and GIS technology,the landslide model and susceptibility zoning of Dingcheng District are established.The work includes the following aspects:(1)According to the characteristics and distribution law of landslides in the study area,the evaluation factors were selected from the factors of topography,formation lithology,rock and soil type,meteorology and hydrology,human engineering activities,etc.(2)According to the previous experience and the physical and geographical conditions of the study area,the slope,aspect,plane curvature,section curvature,lithology,rainfall,river and road are selected as evaluation factors.(3)The BP neural network model is used to train a part of the quantitative landslide data in the study area.The BP neural network prediction model is compared with the actual landslide geological disaster situation.When the output value reaches the expected value,the evaluation model is used to evaluate the remaining quantitative landslide data,and the quantitative value of landslide susceptibility is obtained.(4)By using the spatial processing and analysis function of Arc GIS,the quantitative data of landslide susceptibility learned by BP neural network evaluation model are evaluated in different areas,and the landslide susceptibility evaluation map of the study area is obtained.This paper studies the susceptibility of landslide geological hazards in Dingcheng District,which can provide reliable basis and help for the local government's geological hazard zoning work and prevention and control measures of geological disasters,and promote the local social and economic development.