Research On Identification Method Of Artificial Cavity Based On Frequency Domain Ground-Air Electromagnetic Response

In recent years,the construction of underground artificial cavities to expand urban space resources has become an important trend of urban development,and the number,depth and scale of underground artificial cavities are increasing.The detection and identification method of underground artificial cavity has become the focus of research.In the aspect of military,the detection of underground artificial cavity can not only reduce the blindness of urban attack,but also provide theoretical basis for the design and construction of underground artificial cavity.In the civil aspect,the construction,management and maintenance of underground artificial cavity also depend on geophysical exploration technology.Supported by the research project "Research on Airborne Electromagnetic Detection Method and Implementation Plan Based on Underground Artificial Cavities",this topic studies the detection and identification method of underground artificial cavities based on ground-to-air frequency domain electromagnetic detection technology.The main contents include:1.Based on the electromagnetic field theory,the ground-to-air frequency domain electromagnetic method lays a high-power transmission system on the ground and uses the UAV and other equipment to collect magnetic field signals in the air as a receiving system.It can conduct a large-depth and rapid detection of the area with complex geological structure,and its receiving system will not be affected by ground environmental factors.Due to the complex and diverse application scenarios of the underground artificial cavity,this paper firstly deduces the electromagnetic field propagation mechanism,then analyzes the ground-to-air frequency domain electromagnetic detection method,and finally introduces the principle prototype construction scheme of ground-to-air frequency domain electromagnetic detection.2.The frequency domain electromagnetic detection method is usually to lay a launching system on the ground to emit electromagnetic waves to the detection target area,collect the response magnetic field by the receiving system,and analyze the geological structure of the target area according to the distribution of electromagnetic field.When the geological structure is relatively simple,it can usually be simplified into a stratified earth model.Therefore,this paper establishes a stratified earth model and combines electromagnetic field theory to carry out one-dimensional forward simulation calculation.In order to further study the electromagnetic field distribution in three-dimensional space,combined with the numerical calculation method,the three-dimensional forward simulation calculation of the earth model is carried out,and the electromagnetic response calculation of the earth model with underground artificial cavity is carried out.3.In order to realize efficient,accurate and non-invasive detection and identification of underground artificial cavities,the electromagnetic response obtained by frequencydomain electromagnetic detection method was analyzed in this paper,and the calculation method of apparent resistivity imaging was derived,and the correctness of the apparent resistivity imaging method was verified through simulation experiments and field tests.The electromagnetic response of the 3D geoelectric model is obtained through a large number of simulation calculations.After the apparent resistivity calculation,the apparent resistivity image data set is finally obtained.The Bayesian convolutional neural network is used to classify the image data sets of various underground structures,so as to identify whether there are high resistance anomalies of artificial cavities more efficiently and intelligently,and realize the detection and recognition of artificial cavities.In order to realize efficient,accurate and non-invasive detection and identification of the underground artificial cavity,this paper firstly applies the frequency domain ground-to-air electromagnetic method to establish a one-dimensional stratified earth model and a three-dimensional earth model containing the underground artificial cavity,and analyzes its electromagnetic response.Secondly,the calculation of apparent resistivity imaging is carried out based on 3D simulation data,and the correctness of the apparent resistivity imaging method is verified by field test.Finally,the data set of the apparent resistivity of the earth model is classified by the Bayesian convolutional neural network,which realizes the detection and recognition of the underground artificial cavity.