Research On Subsidence Monitoring Method Of InSAR In Mining Area For Subsidence Funnel

The surface subsidence caused by exploitation of coal resources has brought a series of problems to the local government,coal enterprises and the people in mining area.As the basis of solving these problems,mining subsidence monitoring has attracted many scholars' attentions.With the development of survey technology,more and more new technologies are applied to mining subsidence monitoring.Among those technologies,Synthetic aperture radar interferometry(InSAR)technology which can monitor large area deformation with all-weather has become a research hotspot.However,due to the limitations of InSAR technology,there are many problems in monitoring the surface subsidence caused by a single working face in a short time(called subsidence funnel),such as the location of the subsidence funnel is difficult to determine,and the accuracy of phase unwrapping is not well.In view of these problems,the following research works were carried out in this paper:1)The principle of InSAR deformation monitoring was introduced in detail,and the principle of several commonly used phase unwrapping methods were introduced,and the influencing factors of InSAR Technology in mining subsidence monitoring were analyzed.2)In order to solve the problem that it is difficult to determine the location of the subsidence funnel in the large-scale InSAR interferograms,a subsidence funnel detection model based on the HOG(histogram of oriented gradient)feature was proposed.The preliminary detection model was established by SVM(support vector machine)using the sample set constructed by the subsidence funnel collected in different interferograms.On this basis,the low coherence value of the subsidence funnel was used to optimize the test results and improve the reliability of the test results.The experimental results show that the detection model can effectively detect the settlement funnel in the interferograms,and the threshold of coherence value can improve the detection results.3)In order to solve the problem of discontinuity or missing of interference fringes caused by surface conditions and time incoherence in InSAR mining area subsidence monitoring,a phase unwrapping algorithm based on U-Net neural network was proposed.First,the trained network was used to extract the edge of the interferogram to get the boundary information of the interference fringes.Then,the edge thinning and edge connection were carried out to obtain the complete interference fringe boundary.The phase jump count n is calculated inside the fringe,and 2n? is added or subtracted from the corresponding position in the interferogram to realize the phase unwrapping.Different quality interferograms were selected for experiments,and the experimental results were compared with other commonly used phase unwrapping methods.It is proved that this method can solve the problem that the interferometric fringes cannot be accurately unwrapped due to the discontinuity or missing of interference fringes,and can improve the reliability of InSAR mining subsidence monitoring.