Landslide disaster prediction method based on depth belief network

Abstract

The invention discloses a landslide disaster prediction method based on a depth belief network. Firstly, a monitoring and early warning system of a landslide body is established, a large number of disaster inducing factors are collected, and the main disaster inducing factors are screened by using a MIV algorithm. The selected disaster inducing factors were standardized and divided into test samples and training samples according to specific proportion. Then, a landslide hazard prediction model based on depth belief network is constructed, which is composed of two-layer RBM and three-layer BPnetwork. CD algorithm is used to pre-train the RBM and update the network parameters. Genetic algorithm is used to supervise the training and learning to ensure the overall optimization of DBN network. Finally, the optimized landslide hazard prediction model is reconstructed to classify the landslide grade and predict the possibility of landslide occurrence. The method disclosed by the invention accelerates the convergence speed by extracting characteristic disaster inducing factors, prevents falling into local optimization, and improves the accuracy of landslide disaster prediction.

Description

technical field [0001] The invention belongs to the technical field of geological disaster forecasting methods, and relates to a landslide disaster forecasting method based on a deep belief network. Background technique [0002] Landslide is one of the important types of geological disasters, which threatens the safety of human life and property, and has a great destructive effect on the infrastructure and ecological environment of the disaster area. Therefore, how to use corresponding technical means to monitor and forecast landslide disasters in real time and minimize losses has become the main content of our attention. [0003] There are many existing landslide disaster prediction methods, and the research stage is divided into several periods. The first stage was in the 1960s and 1970s, mainly based on phenomenon forecasting and empirical forecasting, namely the famous "Saito method". Experts infer landslide instability based on landslide failure phenomena, but this me...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a landslide disaster prediction method based on a deep belief network, which solves the problems of slow convergence speed and low prediction accuracy of the algorithms used in the existing disaster prediction, and accelerates the convergence speed by extracting characteristic disaster-inducing factors. Prevent falling into local optimum and improve the prediction accuracy of landslide disaster

Images