Research And Application Of Ground-based SAR Deformation Monitoring Data Processing

The earth’s surface bears human life and is the cornerstone of people’s production activities.The stability of the surface is related to the personal safety of the masses and the orderly progress of people’s production and life.In the past two decades,under the background of global warming,the frequency of extreme weather has increased year by year,and the hazards of geological disasters have also continued to intensify,which also puts forward higher requirements for surface deformation monitoring.Compared with the traditional surface deformation measurement technology,Synthetic Aperture Radar(SAR)has become one of the favorable means of surface deformation monitoring due to its advantages of all-day,all-weather,high resolution and wide range.Spaceborne SAR has inherent defects in the revisit cycle and imaging geometry,which cannot meet the needs of sudden natural disaster monitoring.Compared with spaceborne SAR,Ground-based Synthetic Aperture Radar(GBSAR)has high time resolution(minutes to seconds),high spatial resolution(sub-meter level),and high precision(submillimeter level)and other advantages,and has shown excellent performance in refined real-time deformation monitoring,and has become an effective supplement to spaceborne SAR.Therefore,based on the ground-based synthetic aperture radar interferometry(GB-In SAR)technology,actively carry out research on surface deformation monitoring technology,which has important practical significance for disaster monitoring and early warning,and emergency mapping for disaster relief.At present,the data processing methods of ground-based SAR are mostly borrowed from the spaceborne In SAR technology.However,the unique high time sampling frequency characteristic of ground-based SAR prevents the spaceborne In SAR technology from being well applied.In addition,the current ground-based SAR data processing methods cannot quickly cope with the real-time processing of large amounts of data brought about by high temporal resolution.In view of this,this paper carried out the research and application of ground-based SAR deformation monitoring data processing technology.The main research contents and innovations are as follows:(1)Aiming at the problem of spectrum leakage when traditional FFT is used for range imaging,a ground-based SAR range imaging algorithm based on all-phase FFT is proposed,which achieves the effect of suppressing spectrum leakage and improves the recognition of observation targets,to attenuate the noise.This algorithm replaces the traditional FFT with an all-phase FFT,which combines the excellent spectrum leakage suppression performance of the all-phase FFT with the important characteristics of "invariant initial phase",and realizes that the initial phase of the highest spectral line is strictly equal to the initial phase of the input signal.This ensures that the highest spectral line is not affected by spectral spread.The algorithm is verified by simulation experiments and real experiments.Compared with traditional FFT,in the spectrum diagram of all-phase FFT,the main frequency line corresponding to the observation target is more obvious,and the surrounding side frequency lines are effectively suppressed.Under this imaging algorithm,the deformation measurement accuracy of ground-based SAR indoors can reach the submillimeter level.algorithm replaces the traditional FFT with an all-phase FFT,which combines the excellent spectrum leakage suppression performance of the all-phase FFT with the important characteristics of "invariant initial phase",and realizes that the initial phase of the highest spectral line is strictly equal to the initial phase of the input signal.This ensures that the highest spectral line is not affected by spectral spread.The algorithm is verified by simulation experiments and real experiments.Compared with traditional FFT,in the spectrum diagram of all-phase FFT,the main frequency line corresponding to the observation target is more obvious,and the surrounding side frequency lines are effectively suppressed.Under this imaging algorithm,the deformation measurement accuracy of ground-based SAR indoors can reach the submillimeter level.(2)Aiming at the problem that the existing atmospheric delay phase compensation method is difficult to distinguish the deformation in the ground-based SAR high-frequency observation signal from the atmospheric delay change,a joint estimation model of ground-based SAR surface deformation and atmospheric model parameters is proposed.Features and spatial characteristics of atmospheric delay,the joint estimation of deformation parameters and atmospheric model parameters is realized by using least squares,which improves the accuracy of deformation estimation.The core idea of this method is to model the surface deformation phase and atmospheric delay phase of the selected PS point respectively,and then construct the function model and random model between the interference phase,surface deformation and atmospheric model parameters,and use the least squares The method is used to estimate the parameters,so as to solve the atmospheric delay phase by back calculation.The feasibility analysis of the algorithm was carried out through design experiments,and in order to verify the accuracy superiority of the estimation model,the joint estimation method was compared with the classic external meteorological parameter correction method and polynomial fitting correction method.Accuracy evaluation is carried out by indicators such as error and root mean square error.The experimental results prove that the standard deviation and root mean square error of the joint estimation method proposed in this paper are significantly improved compared with the other two methods,which verifies the reliability and precision superiority of the joint estimation method.In addition,the method was applied to the field bridge vibration monitoring in Changsha City,Hunan Province,and the bridge vibration was successfully separated from the atmospheric delay.(3)In view of the low post-processing efficiency of ground-based In SAR and the poor real-time performance of existing ground-based In SAR real-time processing algorithms,a real-time deformation function model of ground-based In SAR based on variable sequential adjustment is proposed,and the deformation parameters are scientifically modeled.Based on the established real-time deformation function model and joint estimation method of deformation parameters and atmosphere,a set of ground-based In SAR real-time dynamic processing method is constructed,which realizes real-time high-precision dynamic processing of ground-based In SAR data.A landslide in Danba County,Sichuan Province was observed continuously for 7 hours using ground-based SAR,with a total of 137 SLC images.Using the real-time dynamic processing method proposed in this paper to process it,the cumulative deformation results and average velocity diagram of the landslide during the observation period are obtained.It can be seen from the processing results that many landslides are occurring in the observation area,and the maximum cumulative deformation can reach 20 cm.According to the analysis of the experimental results,90% of the root mean square errors of the residual deformation estimated by the parameters are distributed within 1 mm,and the average value is 0.61 mm.The unwrapping error was about 4% in the first stage;at the beginning of the second stage,due to the sudden decrease in the number of observation targets,which seriously affected the accuracy of phase-space-time three-dimensional unwrapping,the unwrapping error suddenly increased to 20%;The number tends to be stable,and the image unwrapping error is stable at about 1%.As the number of observations increases,the number of unknown parameters to be obtained gradually increases,and the time spent on a single observation gradually increases.The time spent on the "deformation parameter calculation" link gradually becomes the main part of the total observation time,but it still meets the field observation scenario.Requirements for real-time processing of data.(4)In view of the low degree of visualization and weak human-computer interaction of ground-based SAR software,ground-based SAR imaging,D-In SAR and PS-In SAR processing software were developed;comprehensive experiments on ground-based SAR and high-frequency GNSS bridge stability monitoring were carried out.The high-frequency vibration of the bridge is successfully monitored,which verifies the feasibility of the ground-based SAR data processing theory and method developed in this paper in practical application.At present,GBSAR is used for bridge deformation monitoring,and other monitoring equipment is used for external conformity accuracy verification.The work is not perfect,and the process is not perfect.In this thesis,through the joint use of GBSAR and GNSS,bridge vibration under two long-span bridges in Changsha City,Hunan Province for deformation monitoring,a bridge interferometric phase data processing flow was designed for the GPRI-II high-frequency fixed-sweep mode.The results of GPRI-II and GNSS deformation monitoring are compared in time domain and frequency domain.At the same time,in the bridge interferometric phase data processing process,the joint estimation method of deformation and atmospheric model parameters is used to correct the atmospheric phase,thereby obtaining the vertical deformation result of the bridge.Finally,the vertical deformation results of the bridge obtained by GBSAR are compared with the measurement results of the high-frequency GNSS receiver,and the monitoring accuracy of the bridge deformation of the GBSAR is counted.At the same time,combined with the analysis of the deformation external precision and SNR,it is concluded that the signal-to-noise ratio of the observation signal needs to be improved to realize the high-precision observation of GBSAR.In addition,according to the power spectrum comparison results,it is necessary to unify the observation reference when performing GBSAR precision deformation accuracy verification on artificial buildings.