Image Reconstruction Algorithm Of Cross-well Electromagnetic Wave CT Imaging

Cross hole electromagnetic wave CT imaging is a geophysical method to effectively detect the electromagnetic characteristics of underground media.It has a large detection depth,a wide range and a high resolution.It is mainly used in engineering geophysical exploration and the exploration and development of coal,oil and gas,metal minerals and other resources.Through the electromagnetic wave inversion to accurately define the attributes of the underground target（such as absorption coefficient and conductivity）,finally get the high-resolution geological profile image between the two boreholes,so as to judge the location and shape of the underground target.Based on the theoretical basis of cross hole electromagnetic wave image reconstruction technology,Radon transform,aiming at the problem of unknown initial field strength E₀,this paper introduces three methods of processing E₀ in detail:linear fitting method,matrix inversion method and dual frequency electromagnetic wave CT imaging method.For the problem of image reconstruction of inversion algorithm,it introduces back projection algorithm（BPT）,Algebraic reconstruction technique（ART）,simultaneous iterative reconstruction（SIRT）and least square QR decomposition（LSQR）.The linear fitting method can get the initial field strength E₀ before inversion.The matrix inversion method takes the initial field strength E₀ as an unknown parameter to participate in the inversion,and the E₀ can be calculated by solving the large matrix,and the inversion results can be obtained at the same time.The dual frequency electromagnetic wave CT imaging method is used to process the initial field strength E₀,and the ratio of the two frequency data can eliminate its influence,and finally get the conductivity image,but this method is only suitable for good conductors.Different iteration times can improve the imaging results of LSQR algorithm to a certain extent.The measured data are used to analyze the influence of iteration times on the imaging effect of inversion algorithm.The imaging results of 90 iterations are selected for comparison with borehole data,and the imaging inversion algorithm suitable for the study area is obtained.On this basis,the spatial location,scope and size of karst or fracture zone between X1-X2、X2-X3、X4-X5、X6-X5 and X7-X8 boreholes are found out.The results of inversion absorption coefficient contour are basically consistent with the cave information revealed by drilling.According to the inversion of apparent absorption coefficient of abnormal area and surrounding rock,it can be concluded that the apparent absorption coefficient of complete and relatively complete rock is less than 2d B/m;the apparent absorption coefficient of karst or broken rock area is larger,generally greater than 3d B/m.