Basic Research On Azimuth Electromagnetic Wave Detection In Drilling Hole Of Coal Mine

In the process of underground coal mine production,the advance and accurate identification of coal-rock interface is of great significance to the geological guidance of unmanned coal mining and drilling,and it is one of the key technologies to realize intelligent mining of coal mine.Existing geophysical prospecting and borehole measurement methods in coal mine have low detection accuracy for coal-rock interface,which cannot meet the needs of constructing precision of digitization model of coal face.Azimuth electromagnetic wave detection instrument has azimuth resolution ability and large detection depth,which has great application potential for coal-rock interface detection.Due to the different measurement environments,the existing commercial logging instruments in oil fields cannot be directly used for the measurement of borehole along coal seam in coal mine.Therefore,it is of great significance to carry out the basic research of azimuth electromagnetic wave detection of borehole along coal seam,which is suitable for the identification of coal-rock interface in coal mine.In this paper,the coal-rock interface in coal mine is taken as the research object,and the along coal seam calculation model of azimuth electromagnetic wave detection is established by combining theoretical research and numerical simulation,developed the horizontal layered strata model calculation software.The working principle,influencing factors of measurement response,typical response characteristics,detection range,resolution of azimuth electromagnetic wave detection method and other basic problems are studied.Aiming at the measurement environment of underground coal mine,the parameters of azimuth electromagnetic wave coil system are optimized,and the response characteristics and detection ability of the optimized coil system in horizontal hole of underground coal mine and complex 3D geological model such as undulating stratum and goaf are analyzed.Finally,a new method of azimuth electromagnetic wave detection data inversion based on artificial neural network is proposed.The main research results of this paper are as follows:1)Based on Maxwell’s equations of static medium,the expressions of reflected and transmitted wave field of plane electromagnetic wave at coal-rock interface are derived by introducing complex permittivity.On this basis,the influence of resistivity and dielectric constant on electromagnetic wave propagation at coal-rock interface is investigated,which lays a theoretical foundation for the design of azimuth electromagnetic wave instrument.Established the coal rock interface bearing typical electromagnetic wave measuring response forward model,using the generalized reflection coefficient method and the finite element method to calculate the coil system structure,transmission frequency,hole spacing,radius of coil,Angle,layer thickness,the resistivity value,the resistivity contrast,hole diameter and drilling fluid,formation resistivity anisotropy response factors such as the change of measurement,The main and secondary factors affecting edge detection distance of azimuth electromagnetic wave instrument are clarified and their influence law on measurement response is revealed.2)Based on the theoretical research and numerical simulation results,the structure and parameters of azimuth electromagnetic wave coil system used for coal-rock interface recognition in underground coal mine are optimized,and the double inclined orthogonal coil system has better applicability in horizontal drilling coal-rock interface recognition.The response characteristics and resolution ability of the optimized coil system,such as frequency,source distance and resistivity contrast,were analyzed.The detection characteristics of the optimized coil system in complex geological models,such as undulating strata and goaf,and the detection ability of coal-rock interface were investigated.The results show that the preferred azimuth electromagnetic coil system has better detection ability for undulating strata and low resistance goaf.3)In view of the shortcoming that gradient data inversion method cannot obtain high precision and fast speed at the same time,a new azimuth electromagnetic wave detection data inversion method based on multi-task deep convolutional neural network and end-to-end full convolutional neural network is proposed and established.The effects of each component of convolutional neural network on electromagnetic wave data inversion are analyzed,and the inversion parameters of two deep convolutional neural networks are optimized.The results show that this method is fast and robust,and can accurately map azimuthal electromagnetic wave forward modeling data into coal seam resistivity and coal-rock interface edge exploration distance to be inverted.