Application Of Kalman Filtering In GB-InSAR Slope Deformation Monitoring

The slope is constantly deformed under the action of internal and external geological forces.The slope could be stable under a certain deformation range,or it will cause landslides and other geological disasters.Therefore,the measurement of slope deformation has always been a hot topic for scholars.Due to the high precision,high time and spatial resolution,Ground-based Synthetic Aperture Radar Interferometry(GB-InSAR)has been widely used in slope deformation measurement.The traditional Synthetic Aperture Radar Interferometry(InSAR)time-series processing method does not consider the spatial-temporal correlation of atmospheric disturbance in GB-InSAR data,and cannot process the data and predict slope deformation in real time.Besides,the data processing takes up a lot of computer memory.In response to the above problems,this article makes the following researches:(1)This paper considers the characteristics of GB-InSAR atmospheric delay time-space correlation,and proposes an improved multi-temporal GB-InSAR data timing processing method.This method introduces a method based on the winding phase atmosphere removal method to remove the static model atmosphere on the interferogram,using filtering and The interpolation method separates the residual turbulent atmosphere and the deformation,avoids the interference of the time-dependent atmospheric delay phase in the linear construction of the network,and effectively improves the accuracy of the GB-InSAR time series analysis.The GB-InSAR data of Malanzhuang open-pit derived instrument and Zhenan hydropower station are utilized to verify the feasibility and effectiveness of the proposed method.(2)Current existing slope deformation prediction methods can only predict the deformation of a small number of slope monitoring points,and cannot be applied to large-area slope deformation predictions.This article proposes a GB-InSAR slope deformation prediction method based on Kalman filtering.Combined with an improved multi-temporal GB-InSAR data processing method,a Kalman filter based on slope deformation sequence is designed.The deformation sequence is calculated by the improved multi-temporal GB-InSAR method,the partial deformation sequence is used to determine the Kalman filter state transition matrix,and then the Kalman filter is used to filter the deformation sequence and predict the deformation at the next moment.The prediction method can predict and analyze the overall deformation of the slope within the monitoring range of GB-InSAR.The proposed method is verified by multi-temporal GB-InSAR data,and the results show that the proposed method can make full use of the deformation information obtained by multi-temporal GB-InSAR and effectively predict the deformation trend of the GB-InSAR monitoring area.(3)The traditional real-time processing method is based on the entire GB-InSAR data set,or grouped and processed in real time.This type of method is slow,cannot meet the timeliness of slope monitoring,and consumes a lot of computer memory.Therefore,this paper proposes a realtime processing method for GB-InSAR slope deformation monitoring based on Kalman filtering.According to the optimal estimation value of the deformation at the previous time,predict the deformation value at the current time,and interfere the SLC acquired at the current time with the main image to obtain a new interferogram,and use the modified phase of the new interference graph as the Kalman filter observation value Filter estimation.This method does not take up too much computer memory and is highly efficient.It can calculate the best estimate of the current deformation within a few minutes.Malanzhuang monitoring data experiments verify the effectiveness of the method.