Research On Data Fusion Algorithm Of Geological Hazard Monitoring Based On Resistivity Imaging Technology

Landslide disaster is a common natural phenomenon of sudden change in rock and soil mass in China.It is a kind of natural disaster with great harm.Because of its sudden characteristics,it is difficult to prevent in real life,causing serious losses of life and property to our country.Therefore,landslide geological disaster monitoring is a very important research.Effective prediction and early warning of disasters can avoid the loss of life and property caused by serious natural disasters.Based on resistivity imaging technology,a deep detection method applied to landslide geological hazard monitoring is proposed in this paper.For landslide geological hazard monitoring and early warning,not only the multi-source sensor system composed of multiple monitoring items such as GNSS,soil temperature and humidity sensor and inclination sensor is used for monitoring,but also the geophysical exploration method resistivity imaging technology is applied to landslide geological hazard monitoring,The resistivity imaging system can deeply explore the internal medium composition and evolution process of landslide geological body,provide a new method for the study of landslide disaster mechanism,and improve the reliability of Landslide Exploration and monitoring.For the multivariate heterogeneous data system composed of rainfall,surface displacement,acceleration,inclination and resistivity,this paper first uses principal component analysis to reduce the dimension of data,and selects soil humidity,temperature,surface displacement,acceleration and inclination as the main factors of geological disaster research.Secondly,taking soil humidity,temperature,surface displacement,acceleration and inclination as the model input and resistivity as the prediction output of the model,the resistivity data is predicted.Finally,the landslide imaging model is formed through RES2 d inversion software.Based on this,the geological disaster prediction model based on multiple regression algorithm is proposed,and the multi-source and heterogeneous real-time monitoring data fusion method is constructed,A prediction model suitable for geological hazard monitoring is established.The above five prediction methods of soil resistivity and slope angle are used as the input data of multiple regression analysis,and the prediction results of soil resistivity and slope angle are good.The effectiveness of the prediction results is also proved by resistivity inversion imaging.The application of resistivity imaging technology in geological hazard monitoring successfully presents the internal characteristics of geological body,which has an obvious effect on the study of Landslide Evolution Mechanism and change process.