Design And Implementation Of Landslide Monitoring System Based On Random Forest

70% of China's land area is covered by a complex mountainous environment,and it is one of the countries most affected by landslide disasters in the world.At present,scholars at home and abroad have carried out a lot of research on landslide monitoring and achieved certain results.However,due to the large number of hazard factors causing landslides,the safety factor of the internal state of the slope is calculated to be uncertain and singular.Forecasting and warning of landslide status is still a world problem.This paper designs a monitoring and early warning system based on the Random Forest model to analyze the landslide state in real time,and realizes the effective monitoring of landslide state risk classification and time prediction in high-risk areas,which has certain practical significance.In this paper,a mountain condition monitoring system based on Stochastic Forest model is designed by using socket communication technology,database technology and machine learning technology.The system mainly includes three parts:landslide monitoring algorithm end,Socket communication program and human-computer interaction interface.The landslide monitoring algorithm mainly analyzes,processes and monitors the data collected by the lower computer.The data of the pore water pressure,earth pressure,rainfall,landslide displacement,vibration,temperature and humidity collected by the lower computer are transmitted to the upper computer through the GPRS network.The landslide monitoring algorithm first performs data cleansing and missing value filling,selects the landslide displacement as the target variable,performs correlation analysis by calculating the Spearman correlation coefficient between the two variables,and eliminates the linear correlation higher variable to select the feature.Regression prediction is carried out on the slope state space.The system compares the correct rate of the logistic regression model and the random forest model by cross-validation results,selects the random forest model with higher score,and uses the least squares third-order orthogonal polynomial of the moving smoothing idea.The smoothing algorithm smoothes the curve,reduces its accidental error,converts the spatial deformation-time curve into a tangent angle curve to classify the slope state,and uses the Verhulst biological model to predict the landslide time of the high-risk slope.The Socket communication program is mainly used.Responsible for coordinating databases,clients,and calculations Normal work between the various parts of the method model,using the completion port in the C/S communication mode,using the system multi-core processor,using multi-line threads to handle multiple tasks,building threads in advance for port message query,reducing thread switching time At the same time,the asynchronous communication of the system is realized,and the problem of system balance is solved at the same time;the human-computer interaction interface is written in C#form,in order to ensure the security of the system,the user division is divided by the minimum privilege level,and the database is used for record review,so that each The user operation has traces and the responsibility is pursued to the individual;the system uses the standard CBC working mode 3DES algorithm,using the initial vector(IV)+key+plaintext for data storage,network communication and password access control system for data block encryption and Decrypt,After the laboratory simulation of the mountain simulation,when the lower machine sends the mountain information data,the system can stably classify the landslide state,monitor the early warning and forecast the rough landslide time.The system adopts the mountain condition monitoring model based on Stochastic Forest model,which improves the adaptability to meteorological conditions and topography,reduces the consumption of human resources.The accuracy of spatial prediction and classification of mountain condition can reach 82%.It has good real-time,reliability and stability.It has certain reference significance for mountain disaster monitoring in the future.