Research On The Maximum Detectable Deformation Gradient Model Of InSAR Technology For Deformation Monitoring In Mining Areas

Due to the limitation of the sensor itself and the effects of spatial and temporal decoherence and noise of the SAR image,InSAR technology is prone to phase aliasing in the large deformation area of the mining area,making it difficult to recover the deformation phase from the interference phase,and thus unable to obtain the true surface deformation information which is extremely prone to waste of human and material resources.In view of this problem,to determine whether the InSAR technology is suitable for monitoring the deformation of the mining area and avoid unnecessary waste of manpower and material resources,this thesis combines the geological and mining factors related to mining subsidence,uses numerical simulation and regression analysis to explore the relationship between deformation gradient and the geological and mining factors,and establishes a maximum detectable deformation gradient model of InSAR technology suitable for mining subsidence.The research work and main results of this thesis are as follows:(1)Summarizing the development of SAR satellites and InSAR technology,and the current research status of the application of InSAR technology in deformation monitoring of mining areas,and pointing out the deficiencies in the research of InSAR technology and related algorithms at home and abroad.The basic principles of SAR,InSAR,D-InSAR,and time-series InSAR technology are described,and the key parameters affecting the maximum detectable deformation gradient of InSAR are analyzed.(2)The three kinds of maximum detectable deformation gradient models of InSAR including Massonnet,Baran and Jiang Mi were studied.It is found that the existing models are the general model,which can be applied to a variety of satellite data and various application fields.Moreover,the models are simple and can predict the InSAR detectability conveniently and quickly.However,existing models are established by analyzing the relationship between coherence,SAR image parameters and deformation gradient,without considering large gradient and nonlinear deformation rules of the mining area,so the applicability to the mining area is low.And the models need to obtain the SAR image of the detection area in advance,which will easily cause a waste of time and resources.Therefore,this thesis analyzes the factors that affect surface deformation of mining area,and finds that under the condition that the mining face is determined,mining depth,mining thickness and subsidence coefficient determine the intensity of surface deformation.Therefore,the above parameters are selected to establish the maximum detectable gradient model of InSAR in the mining area.(3)Constructing a static model about the maximum detectable deformation gradient.In this model,90 sets of deformation fields and 1620 differential interferograms are obtained by using the probability integral method to simulate the deformation fields of C-band,X-band and L-band SAR images with different resolutions and different angles of incidence.Then the relationship between mining depth-to-thickness ratio and subsidence coefficient and deformation gradient under different SAR image conditions is analyzed.Based on this,a static model of the deformation gradient under the horizontal coal seam is established using the regression method.At the same time,it is concluded that there is an inverse proportional relationship between the mining depth-to-thickness ratio and the maximum detectable deformation gradient;there is a clear linear relationship between the subsidence coefficient and the maximum detectable deformation gradient.(4)Constructing a dynamic model of the maximum detectable deformation gradient.This model combines knothe time function and probability integral method to simulate the dynamic deformation of the mining process,and uses the probability integral parameters selected by numerical simulation.At the same time,according to the 65 near-level coal seam mining area parameters listed in the "Code for coal pillar reservation and pressure coal mining of buildings,water bodies,railways and main shafts",a total of 2015 simulated images were established.By studying the rules of mining area parameters and deformation gradients under different mining progress,it is found that the mining progress and the maximum detectable deformation gradient have an exponential function relationship.The mining progress and the maximum detectable deformation gradient are subjected to regression analysis to establish a regression equation.Further study is made on the relationship between mining depth,mining thickness and subsidence coefficient and equation coefficients,and using regression analysis method establishs a dynamic function model of TerraSAR data under horizontal coal seams.(5)Using 13 TerraSAR images of Daliuta mining area in Shanxi Province,the static and dynamic models are verified respectively.The experimental results show that:the static model has high detection ability for the isolated points with continuous points or adjacent control points close to detection points,and the accuracy can reach millimeter level;the dynamic model can accurately predict the detection ability of terrasar in the whole mining area,and preliminarily divide the detection boundary,which can provide guidance for the application of InSAR Technology in mining area.Compared with other models,the two models proposed do not need to obtain the coherence of SAR image calculation in advance,only use the geological and mining conditions parameters to predict the maximum detectable value of InSAR Data,improve the detection accuracy of the model,and expand the scope of application of the model.