Study on displacement prediction of landslide based on neural network

In recent years, the economic losses caused by landslides were as high as several billion dollars. Therefore, the study of landslide risk has become a hot topic today. Landslide displacement prediction is a highly nonlinear and extremely complex issue. In most cases, it is difficult to use mathematical models to describe the process clearly. Data mining technology, which uncovers the hidden data patterns and models effectively, can be used for landslide displacement prediction. Series of system models are trained, validated, and applied to a landslide study along the Three Gorges and cases from the literature. Based on these monitoring data from deformation displacement of the Bazimen landslide in Zigui County, natural rainfall in the area of Shazhen and Guizhou, water level changes of the Three Gorges Dam, we propose a theoretical method suitable for analyzing the impact of landslide deformation. By drawing duration curve about each monitoring point displacement and predisposing factors on the valid data, we use regression analysis method to analyze correlation and hysteresis of landslide displacement variation and predisposing factors through the actual situation of monitoring monthly reports.