Estimation And Optimization Of Three-Dimensional Multi-Level Deformation Of Mining Area Based On Single-Band OTD-InSAR

China is the world's largest coal-producing country and the coal will still be China's main energy source for a long time under the background of carbon neutrality.Coal mining will not only severely damage the surface environment of the mining area and cause the surface collapse and formed a pit,but also cause disasters in severe cases.Therefore,monitoring ground deformation in mining areas has important practical significance for disaster assessment.Because the surface deformation caused by coal mining often has the characteristics of fast subsidence rate and large deformation gradient.Combining Differential Interferometric Synthetic Aperture Radar(DInSAR)and Offset-Tracking(OT)technology based on singleband SAR image can extract two-dimensional multi-level discontinuous deformation in the line of sight(LOS)direction and azimuth direction.This cannot extract the multi-level deformation in the three-dimensional(3-D)direction and cannot reflect the true surface deformation of the mine.Regarding the issue above,firstly,this thesis establishes a multilevel mining deformation fusion model based on the single-band OTD-In SAR technology and obtains the multi-level continuous deformation in the LOS and azimuth directions.Then building the probability integral model to estimate the 3-D multi-level mining deformation.Finally,the genetic algorithm is used to optimize the 3-D deformation.The main study work are as follows:(1)Using DInSAR and OT technology to monitor time series deformation in the 52303 workface of Daliuta coal mine.The subsidence distribution,coherence map,interferogram and subsidence map with the large-level and small-level deformation were obtained;The priori fusion model of multi-level deformation was established and the large-level and smalllevel deformations were fusing.And then in order to obtain continuous multi-level LOS deformation of the 52303 workface,the Kriging interpolation technique was used to achieve interpolation.(2)The estimation model of 3-D deformation based on the probability integral model was established,and the feasibility of the model was verified by simulation;The time-series3-D multi-level deformation in mining area was estimated by using multi-level LOS deformation.Comparing with the leveling data to verify the results and the Root Mean Squared Errors(RMSE)are 0.1880 m,0.2243 m,0.2074 m,respectively.(3)Using genetic algorithm to invert the probability integral parameters in the model,and the parameter error obtained by inversion does not exceed 0.6%;To obtain more accurate parameters,the LOS and azimuth deformations are combined to optimize the parameters based on the genetic algorithm,and then using this parameter to estimate the 3-D deformation.The effectiveness of the optimization method is verified by theoretical simulation and measured data;For a large number of errors caused by ground fissures,an empirical model considering the vertical deformation gradient was established to eliminate the errors.The experimental results show that the model can effectively remove the errors.