Research And Implementation Of Geological Disaster Monitoring And Early Warning Platform Based On BIM+GIS

Geological disasters represent a major threat to life and property in many areas of the world,especially in China.At present,traditional geological disaster monitoring and early warning systems are mostly based on two-dimensional geographic information system,which often have disadvantages such as poor visualization capabilities and unreliable model prediction accuracy.Based on the current severe situation of geological disasters in my country,research on an early warning system that can effectively predict the occurrence of geological disasters has become an urgent task for the prevention and control of geological disasters.This paper takes the the Hupo Village landslide monitoring project in Yushui District,Xinyu City,Jiangxi Province as an example,aims to combine the 3D-BIM geological model,GIS data and Beidou high-precision reference station network with the geological disaster monitoring and early warning system together,designs and implements a monitoring and early warning platform suitable for multiple hidden hazardous locations of geological disasters,a new combination model is proposed to predict geological disasters,which improves the visualization and early warning capabilities of the monitoring and early warning platform,and provides a scientific basis and decision-making reference for disaster prevention and reduction,and loss reduction.The research content and results of the thesis are as follows:1.Introduce the BIM 3D geological model to the earth disaster monitoring and early warning platform.This paper has realized the 3D modeling and visual analysis of geological objects based on borehole data and SuperMap iDesktop software.By superimposing dynamic voxel grid 3D space field data on the 3D BIM geological model,It not only realizes the ascending expression of data,but also solves the shortcomings of the traditional static BIM geological model with weak real-time and cooperative expression.2.The research has realized a combined prediction model based on the displacement and deformation of geological disasters.This paper proposes a new optimal weight combination model based on machine learning and time series analysis,which can predict the occurrence of geological disasters through displacement and deformation data.The basic model of the combined model is composed of ridge regression,SVM,GRU and ARIMA.The weight calculation process draws on the Soft Voting weight voting idea of integrated learning in machine learning,which solves the limitation that the traditional single displacement and deformation prediction model cannot accurately describe the deformation and evolution law.In addition,a relatively complete data processing process was proposed in the data preprocessing stage,including abnormal data processing,missing data interpolation,and Kalman filtering method to reduce noise.Through comparative analysis of the prediction results,the combined model with the best weight has a higher prediction accuracy than any basic model.And it's better than the Voting Regressor method of the scikit-learn machine learning algorithm library.3.Designed and implemented a BIM+GIS-based geological disaster monitoring and early warning platform and a supporting mobile WeChat Mini program.The platform adopts the B/S structure,and the framework of the monitoring and early warning platform adopts the method of separating the front and back ends.The platform realizes a large screen of integrated monitoring of 3D geological disasters based on BIM+GIS,basic data management,real-time monitoring,group monitoring and group defense,early warning rule configuration,early warning prediction,and early warning release modules.And the mobile terminal for supporting group defense inspector to patrol is implemented based on WeChat Mini program development platform.