| Coal is China's main energy source and an important industrial raw material.As an important basic industry,the coal industry strongly supports the steady and rapid development of China's national economy and society.The development of national social economy cannot be separated from coal resources,but coal mining may cause damage to ecological environment,industrial and civil buildings and structures,facilities of transportation,electricity,and communication.Obtaining mining subsidence rules through real-time monitoring of surface subsidence in mining areas,has important theoretical and practical value for preventing and controlling the impact of mining subsidence on the ecological environment,guiding the green and safe production of mining areas and the governance and restoration of the ecological environment.Synthetic Aperture Radar(SAR)technology is a non-contact and green monitoring method,which has been widely used in deformation monitoring in mining areas.Interferometric SAR(InSAR)technology has high accuracy in monitoring small-scale deformations,but it cannot obtain large-scale deformations in mining areas.Sub-band InSAR technology can reduce the difficulty of unwrapping and improve the measurement accuracy,but there is very little research on this technology at home and abroad.Offset-tracking method can be used to extract large-scale deformations in mining areas,but its monitoring accuracy is relatively low.How to combine the advantages of various methods and obtain large gradient deformation information with high accuracy in mining areas is a scientific problem at present.For this reason,this paper focuses on the suitability of sub-band InSAR technology for deformation monitoring in mining areas,the selection of the cross-correlation windows of offsettracking methods in mining-induced deformation monitoring,and the fusion of InSAR,sub-band InSAR,offset-tracking method and probability integral model.The reliability of these methods are proved by comparison with the measured data.The main work and results are as follows:(1)This paper summarizes the current research status of SAR technology in mining-induced deformation monitoring,describes the principles of InSAR technology,sub-band InSAR technology,and offset-tracking method for deformation monitoring,analyzes the errors sources of the three methods,and briefly introduces the principle of probability integral model,and simply analyzes the laws of subsidence and horizontal displacements on the main section in the strike direction of the horizontal coal seam.(2)To explore the relationship between the monitoring accuracy of sub-band InSAR technology and sub-band bandwidth parameters,based on the analysis and study of the influence of different resolutions and different sub-band bandwidths on monitoring different surface deformations,the optimal sub-band bandwidth selection method under different image resolution conditions is determined,and the method of fusing sub-band InSAR technology and probability integral model is proposed to extract large-scale deformations in mining areas.The experimental results show that:the wider the full-bandwidth of the image,the higher the monitoring accuracy of the sub-band InSAR technology;the monitoring accuracy of sub-band InSAR technology is related to the full-bandwidth of the image,sub-band bandwidth parameters and deformation magnitude;and there is an optimal full-frequency bandwidth and sub-band bandwidth to monitor deformation of different magnitudes.So,the appropriate subband bandwidth parameters should be selected according to the image full-bandwidth and deformation magnitude to achieve the purpose of optimal monitoring.The subband InSAR technology based on probability integral model can effectively remove the noise in the sub-band interferometric phase.The proposed method is used to extract the surface deformation information of the working face 52,304 in the Daliuta mining area,in Shaanxi Province.By comparing the experimental results with the deformation results obtained by the conventional Differential InSAR(D-InSAR)and global positioning system(GPS),it is found that: the conventional D-InSAR yields higher precision than sub-band InSAR technology in regions with small deformations;in the edge region of the mined-out area,the conventional D-InSAR has been completely unable to monitor the accurate deformation information,while the proposed method can better obtain surface deformation information with a standard deviation of 0.108 m.(3)Aiming at the problem that the deformation gradient within the crosscorrelation window and the noise between images have a greater impact on the monitoring accuracy of the offset-tracking method,the influence of the crosscorrelation window on the accuracy of monitoring different deformations by the offsettracking method is systematically studied,and the results show that: using small crosscorrelation windows can ensure the accuracy of the deformation results,and using large cross-correlation windows can suppress noise better;in areas with large deformation gradients,using large cross-correlation windows will cause deformation compression.Therefore,when there is noise between the main and slave images,larger crosscorrelation windows should be selected for the stable area and areas with smaller deformation gradients to suppress the influence of noise on the deformation results;for areas with larger deformation gradients,smaller cross-correlation windows should be selected to ensure the accuracy of the deformation information.For this reason,an offset-tracking method of the adaptive cross-correlation window based on the deformation gradient is proposed.This method is used to extract the deformation of two TSX images with a time interval of 143 days.The experimental results show that the method is reliable and effective.(4)In view of the problem that InSAR technology and offset-tracking method are difficult to obtain the complete deformation field of the mining area due to the influence of the vegetation in mining areas,the IOP method based on probability integral model combining InSAR technology and offset-tracking method is proposed to obtain the complete deformation field with high accuracy in mining areas.The IOP method uses the deformation information of some discrete points obtained by the InSAR technology and the offset-tracking method to obtain the probability integral parameters and calculate the not obtained deformation information in vegetation area through the probability integral model,and to fill in the deformation information not obtained by the InSAR technology or the offset-tracking method.Thus,the complete deformation field in the vegetation area can be obtained.By comparing the deformation information obtained by the IOP method with the GPS measured data,the monitoring accuracies of the IOP method in the strike and dip directions are 0.122 m and 0.103 m,respectively,and which are better than that of InSAR technology and offset-tracking method.(5)In view of the adaptability and accuracy of different deformations monitored by InSAR(small gradient),sub-band InSAR(medium gradient),and offset-tracking(large gradient),the method(ISOP)based on probability integral model by fusing InSAR technology,sub-band InSAR technology and offset-tracking method is proposed to obtain the complete deformation field in mining areas with high precision.This method uses InSAR to obtain small deformation information at the edge of the deformation basin,sub-band InSAR to obtain medium deformation information at the edge of the mined-out area,and offset-tracking to obtain large gradient deformation information at the center of the deformation basin.Then,this method uses probability integral model to remove the noise,fuse and invert the results of the three methods,so as to obtain the high-precision and complete deformation information field in mining areas.The experimental results of the working face 52,304 in Daliuta mining area show that: the average maximum surface deformation speed in the SAR LOS direction reaches 140 mm/day,and the maximum cumulative deformation in the SAR LOS direction reaches 3.605 m;by comparing the experimental results with GPS measured data,the root mean square errors in the strike and dip directions of the ISOP method are 0.161 m and 0.105 m,respectively.There are 78 figures,31 tables and 173 references in this paper. |
PDF Full Text Request