Research On In Sar Monitoring And Prediction Method Of Mine Subsidence

With the occurrence of geological disasters in mining areas,the monitoring and prevention of surface subsidence has become a hot topic in recent years.When coal is mined from underground,goaf will be formed.Monitoring the surface subsidence above the goaf can understand the damage degree of the surface settlement,explore the mining subsidence mechanism,and provide data for the geological disaster prevention in the mining area.The traditional subsidence monitoring in mining area adopts the form of point-like monitoring stations,which has been unable to meet the needs of mining subsidence monitoring,these methods have high consumption,low efficiency and limited coverage.In recent years,with the development of computer,space satellite and other technologies,it has become a new trend to use satellite remote sensing image to monitor surface changes.In SAR,in particular,has become a hot spot technology for surface subsidence monitoring,especially for mining area subsidence monitoring.However,due to the influence of mining mode and geological structure,the subsidence of the mining area presents characteristics of nonlinear,complexity and large gradient,etc.,In SAR technology is more sensitive to the surface subsidence;while it can only tell us what happened accurately,but can't tell us what will happen in future.So,some improved algorithms of In SAR and a prediction algorithm combined with DIn SAR data needs to be found.The main research contents of this paper are as follows:(1)In this paper,a SAR image registration method based on optimal matching point strategy is proposed.The steps of this strategy are as follows: First,the SRTM DEM is simulated as a SAR image,then feature matching points are searched on simulated SAR images and master/slave SAR images,they are filter using vector field consistency algorithm.At last the exact matching points are found and involved in calculating polynomial parameters,the SAR image registration progress is completed.(2)In this paper,an adaptive Goldstein interferogram filtering method combined with bidimensional empirical mode decomposition(BEMD)algorithm is proposed.First,interferogram is decomposed into four inherent mode function(IMF)components using BEMD.The first three components contain more than 95% of the noise,they are transformed by the Fourier transform and divided into a large number of image Windows.These image Windows are filtered by Goldstein algorithm which was modified with the filter factor of SNR.The IMF components which are filtered are transformed back to interferogram.Experimental results show that this method can improve the filtering quality and reduce the loss of detail information.(3)Surface subsidence in Ningdong mining area was monitored using SBAS In SAR technology.The method optimizes the SAR image interference pair combination by setting the temporal and spatial baseline threshold.The cumulative subsidence data and rates of high coherence points are obtained.The degree of spatial aggregation of high coherence points on the subsidence values and rates is calculated using Getis-Ord Gi* algorithm.Dozens of strong deformation zones have been found in Ningdong mining area.In SAR monitoring values and GPS monitoring values are compared and analysed.It shows that the In SAR monitoring results are accurate and reliable.(4)Combining with DIn SAR data and SVM regress,mining subsidence prediction model was established.DIn SAR data is used as training data of SVM regress algorithm.The nonlinear function relationship between the DIn SAR data and the unknown predicted values is established.In order to realize dynamic prediction,fuzzy information granulation algorithm is introduced into prediction model.Two different sets of SAR data were used for predictive experimental research.The average relative error of the predicted results is less than 5.6% and the WIA of the predictive model reached 0.994.The results show that the prediction model proposed in this paper can be effectively used in mining area subsidence prediction.It provides a new idea for the promotion and application of In SAR technology in mining area.