Research On Early Warning System Of Slope And Landslide Based On ANN-LSTM Hybrid Prediction Method

China is one of the countries with frequent landslide geological disasters,and a large number of economic and life and property losses are caused by landslide disasters every year.The results show that real-time monitoring and early warning technology is one of the important means to effectively reduce the losses caused by landslide disasters.Although a large number of studies have been carried out at home and abroad on slope monitoring and early warning technology,there are still problems such as false alarms of disaster hazards and single functions of early warning systems,resulting in huge losses of life and property caused by landslide geological disasters in China every year.Therefore,in order to reduce the losses caused by landslide geological disasters,the study of efficient monitoring and accurate prediction and forecasting of unstable slopes is one of the key scientific problems that urgently need to be solved.Based on the slope of Wenxi Village in Fujian Province as the research background,this paper systematically studies the real-time monitoring and early warning technology of the slope.Firstly,the intelligent real-time monitoring system of the slope of Wenxi Village is constructed by using communication and automation technology;due to the absence of monitoring data and random noise,a set of monitoring data preprocessing methods are constructed based on the methods of information science and statistics;in order to realize the advanced perception of the development trend of slope deformation,a mixed prediction model of landslide displacement is established based on the method of machine learning;based on the idea of process early warning,a multi-layer multi-level early warning model of landslide in Wenxi Village is proposed;and finally through the means of secondary development,Successfully developed the visualization platform of the intelligent real-time online monitoring and early warning system for the slope of Wenxi Village and applied it in the project.The specific work content and conclusions of this article are as follows:(1)Firstly,the basic geological conditions such as topography and stratigraphic lithology of the slope of Wenxi Village are expounded in combination with the geological survey data.Geo Studio numerical software was used to calculate and analyze the stability of the slope of Wenxi Village under extreme rainfall conditions.It is obtained that the slope is in an unstable state under natural conditions,and with the increase of rainfall intensity and rainfall duration,the slope stability decreases significantly until instability occurs.(2)Combined with the calculation results of slope stability,the potential landslide area and monitoring scheme of Wenxi Village were determined by using satellite remote sensing and onsite geological survey technology.In order to solve the problem of poor communication due to the poor geographical environment of the slope area of Wenxi Village,the intelligent real-time monitoring system of the slope based on long-distance wireless communication technology is studied by using communication and automation technology.In order to solve the problem of missing and random noise in the obtained monitoring data,the monitoring missing data is supplemented by Newtonian interpolation,and the noise reduction method using the combination of unbiased risk estimation threshold principle and hard threshold function is the best by studying the comparative numerical experiments of different wavelet threshold principles and threshold functions.(3)In order to achieve the purpose of advanced perception and prediction of landslide disasters,a mixed prediction model of slope displacement ANN-LSTM was constructed by using machine learning methods,and the accuracy of the model was verified by using evaluation indicators.Based on the idea of process early warning,combined with the actual situation of slope deformation in Wenxi Village,a three-layer four-level early warning model of The slope of Wenxi Village is constructed based on the comprehensive criteria of displacement deformation rate,rate increment and improved tangent angle.(4)Finally,based on the secondary development methods,the Wenxi Village Slope Intelligent Real-time Monitoring and Early Warning System Visualization Platform was successfully developed that integrates monitoring data management,data preprocessing model,prediction and early warning model,etc.