A method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection

Abstract

The present invention discloses a method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection, which mainly solves the technical problem that it is difficult to make an intuitive and accurate judgment on the spatial position and distribution of anomalies. It includes: obtaining six-component magnetic field detection data; preprocessing the six-component magnetic field detection data, and performing preliminary dynamic weight allocation of the preprocessed six-component magnetic field data; establishing an electromagnetic field forward modeling equation, using an electromagnetic field forward model to simulate the six-component magnetic field response, generating corresponding simulated six-component magnetic field response data, and comparing it with the six-component magnetic field detection data to form constraints for the inversion process; combining the Laplace operator and the second-order derivative of the magnetic field gradient component after the preliminary dynamic weight allocation to perform spatial enhancement of the anomalies of the three-dimensional complex geological body; building a multimodal network; inputting the preprocessed six-component magnetic field data into the multimodal network to obtain a three-dimensional resistivity model or geological body probability distribution corresponding to the three-dimensional complex geological body.

Description

Technical Field

[0001] The invention relates to the technical field of airborne electromagnetic data processing, in particular to an airborne electromagnetic multi-component data processing method for three-dimensional complex geological body detection. Background Art

[0002] Airborne electromagnetic detection is a technology that uses an aerial platform (such as a helicopter or fixed-wing aircraft) to carry electromagnetic detection equipment for geological exploration. At present, the mainstream airborne electromagnetic detection technology only collects vertical component data that is sensitive to the vertical structure of the stratum, and ignores the horizontal component data that is more sensitive to the lateral structure. This results in limited accuracy in restoring the spatial position and shape of underground three-dimensional geological bodies, and it is difficult to make intuitive and accurate judgments on the spatial position and distribution of abnormal bodies. Among them, the horizontal component data is more sensitive to the changes in the horizontal electrical structure than the vertical component data, and has the potential to improve the lateral resolution of the inversion results. Increasing the collection and application of horizontal component data can enhance the detection of underground three-dimensional geological bodies.

[0003] In addition, the existing technology of multi-component airborne electromagnetic data processing, such as Huang Xin's article "Study on Three-dimensional Inversion of Multi-component Frequency Domain Airborne Electromagnetic Data", uses the finite element method of unstructured tetrahedral meshes combined with the quasi-Newton method with limited memory to carry out the fusion of multi-component airborne electromagnetic data. Its disadvantages are: First, high computational complexity: although the unstructured tetrahedral mesh is suitable for irregular geological structures, it has many degrees of freedom, which increases the scale of linear algebra problems and leads to long iteration time for each step. Second; Limited memory quasi-Newton method limitation: although L-BFGS saves memory, for ultra-large-scale data sets, the number of stored correction elements will be limited, affecting the convergence speed and accuracy. Third, grid quality and parameter tuning are difficult: grid generation requires fine processing and is difficult to automatically optimize, resulting in high labor costs; and the complex tuning process of multiple algorithm parameters such as step size and damping coefficient is time-consuming. Fourth, the inversion stability and reliability are insufficient: when the data is sparse or noisy, the model of the unstructured grid may have unstable solutions; when the actual geological structure deviates greatly from the grid assumption, the accuracy of the results is affected. Fifth, parallel computing capabilities are limited: Finite element methods usually have poor symmetry and are difficult to parallelize efficiently, which limits their application efficiency in high-performance computing clusters.

[0004] Therefore, it is urgent to propose a logically simple, accurate and reliable airborne electromagnetic multi-component data processing method for three-dimensional complex geological body detection. Summary of the invention

[0005] In view of the above problems, the purpose of the present invention is to provide an airborne electromagnetic multi-component data processing method for three-dimensional complex geological body detection. The technical solution adopted by the present invention is as follows:

[0006] A method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection, comprising the following steps:

[0007] Acquire six-component magnetic field detection data of airborne electromagnetic detection; the six-component magnetic field detection data includes a magnetic field intensity component , magnetic field intensity component , magnetic field intensity component , magnetic field gradient component , magnetic field gradient component and the magnetic field gradient component ;

[0008] Preprocessing the six-component magnetic field detection data to obtain preprocessed six-component magnetic field data; preliminarily allocating dynamic weights to the six-component magnetic field data preprocessed according to geological prior knowledge and exploration scenarios to obtain six-component magnetic field data after preliminary allocation of dynamic weights;

[0009] Establish an electromagnetic field forward modeling equation, input the six-component magnetic field data after preliminary allocation of dynamic weights into the electromagnetic field forward model to simulate the six-component magnetic field response, generate the corresponding simulated six-component magnetic field response data, and form the constraint conditions of the inversion process by comparing with the six-component magnetic field detection data;

[0010] Obtain the magnetic field gradient components after preliminary allocation of dynamic weights , magnetic field gradient component and the magnetic field gradient component , the magnetic field gradient component after the initial allocation of the Laplace operator and dynamic weights The second derivative of the magnetic field gradient component The second derivative and magnetic field gradient component The second-order derivative of is used to perform spatial enhancement of abnormal bodies in three-dimensional complex geological bodies;

[0011] A multimodal network is constructed; the multimodal network comprises an input layer, a feature extraction layer, a fusion layer and an output layer connected in sequence; the feature extraction layer is composed of a parallel intensity component branch network and a gradient component branch network; the intensity component branch network is composed of a three-layer convolutional neural network CNN, and extracts vertical hierarchical features; the gradient component branch network is composed of a three-layer convolutional neural network CNN and an edge detection layer, and performs boundary response enhancement of the edge of the abnormal body of a three-dimensional complex geological body; the fusion layer uses an attention mechanism to dynamically update the weights of the preprocessed six-component magnetic field data;

[0012] The preprocessed six-component magnetic field data is input into the multimodal network, and the three-dimensional resistivity model or geological body probability distribution corresponding to the three-dimensional complex geological body is obtained.

[0013] Furthermore, the magnetic field gradient component The expression is:

[0014] ;

[0015] The magnetic field gradient component The expression is:

[0016] ;

[0017] The magnetic field gradient component The expression is: .

[0018] Furthermore, the magnetic field intensity component , magnetic field intensity component , magnetic field intensity component An electromagnetic detection device is mounted on a drone platform for data collection; the electromagnetic detection device comprises a transmitting coil and a multi-component electromagnetic receiving coil arranged at the bottom of the drone platform, and an aerial photography positioning and attitude determination module arranged at the bottom of the multi-component electromagnetic receiving coil; the aerial photography positioning and attitude determination module collects the pitch angle, roll angle and yaw angle of the multi-component electromagnetic receiving coil; the multi-component electromagnetic receiving coil comprises a multi-component electromagnetic connection system X coil, a multi-component electromagnetic connection system Y coil and a multi-component electromagnetic connection system Z coil which together form a spherical shape; the multi-component electromagnetic connection system X coil, the multi-component electromagnetic connection system Y coil and the multi-component electromagnetic connection system Z coil have the same structure, and the multi-component electromagnetic connection system X coil comprises a main receiving coil of a magnetic field intensity component, a primary receiving coil of a magnetic field gradient component, and a secondary receiving coil of a magnetic field gradient component which are arranged in sequence from the outside to the inside, a first gap is arranged between the main receiving coil of the magnetic field intensity component and the primary receiving coil of the magnetic field gradient component, and a second gap is arranged between the primary receiving coil of the magnetic field gradient component and the secondary receiving coil of the magnetic field gradient component.

[0019] Furthermore, the six-component magnetic field detection data is preprocessed, including:

[0020] Combine the pitch angle, roll angle and yaw angle, and use the rotation matrix to respectively transform the magnetic field gradient components in the six-component magnetic field detection data , magnetic field gradient component and the magnetic field gradient component Correction is performed, and its expression is:

[0021] ;in, represents the corrected magnetic field gradient, represents the attitude correction rotation matrix, represents the original magnetic field gradient, represents the pitch angle, represents the flip angle, represents the yaw angle;

[0022] The wavelet threshold is used to analyze the magnetic field intensity components in the six-component magnetic field detection data. , magnetic field intensity component , magnetic field intensity component Perform denoising;

[0023] Adaptive Kalman filtering is used to correct the magnetic field gradient component , magnetic field gradient component and magnetic field gradient components Perform denoising.

[0024] Furthermore, according to the preliminary allocation of dynamic weights of the six-component magnetic field data preprocessed according to geological prior knowledge and exploration scenarios, the six-component magnetic field data after the preliminary allocation of dynamic weights is obtained, and its expression is:

[0025] ;

[0026] ;in, represents the dynamic weight corresponding to the magnetic field intensity component, express Take separately , and The corresponding magnetic field intensity component , magnetic field intensity component , magnetic field intensity component ; represents the dynamic weight corresponding to the magnetic field gradient component, express Take separately , and The corresponding magnetic field gradient component , magnetic field gradient component and the magnetic field gradient component .

[0027] Furthermore, an electromagnetic field forward modeling equation is established, and the six-component magnetic field data after the preliminary allocation of dynamic weights are jointly input into the electromagnetic field forward model to simulate the six-component magnetic field response; the electromagnetic field forward modeling equation is expressed as:

[0028] ; Wherein, m represents underground electrical parameters; represents simulated detection data;

[0029] Regularization is used to optimize the underground electrical parameter m, and its expression is:

[0030] ;in, Represents detection data containing six-component magnetic field data; represents a sparse constraint; Represents a smoothness constraint.

[0031] Furthermore, the fusion layer uses the attention mechanism to dynamically update the weights of the preprocessed six-component magnetic field data, and its expression is:

[0032] ;in, Represents the Sigmoid function, which maps the output of the fully connected layer to the interval (0, 1) and generates normalized weights; represents the activation function of the rectified linear unit; Represents the weight matrix of the first fully connected layer; Represents the weight matrix of the second fully connected layer; represents the global average pooling function; Represents the feature map of the Cth channel.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] The present invention combines the pitch angle, roll angle and yaw angle, and uses a rotation matrix to respectively calculate the magnetic field gradient components. , magnetic field gradient component and the magnetic field gradient component It uses fast and accurate attitude solution to perform attitude correction on electromagnetic signals, making the subsequent data interpretation results more accurate.

[0035] The present invention performs preliminary allocation of dynamic weights of six-component magnetic field data based on geological prior knowledge and exploration scenarios, which has the advantage of automatically adjusting data processing priorities according to geological conditions to optimize computing resources and improve the accuracy of inversion results.

[0036] The present invention establishes an electromagnetic field forward modeling equation, and jointly inputs the six-component magnetic field data after preliminary allocation of dynamic weights into the electromagnetic field forward model to simulate the six-component magnetic field response. It verifies the correctness of the data system by simulating the theoretical electromagnetic response and provides a reliable benchmark for the inversion process, thereby improving the rationality and credibility of the solution.

[0037] The present invention combines the Laplace operator and the magnetic field gradient component after the dynamic weight is initially allocated The second derivative of the magnetic field gradient component The second derivative and magnetic field gradient component The second-order derivative of is used to perform spatial enhancement of anomalies in three-dimensional complex geological bodies. It uses mathematical tools to strengthen target features and reduce the impact of interference signals, significantly improving the detection capability under complex geological backgrounds and making anomalies clearer and easier to identify.

[0038] The present invention constructs a multimodal network and fuses the six-component magnetic field data after spatial enhancement and denoising. It provides a more comprehensive geological perspective by integrating information from different data sources (such as multi-component and multi-frequency), and enhances the stability of understanding and interpretation of stratigraphic structure through information complementarity.

[0039] In summary, the present invention has the advantages of simple logic, accuracy and reliability, and has high practical value and promotion value in the field of aviation electromagnetic data processing technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope of protection. For those skilled in the art, other related drawings can be obtained based on these drawings without creative work.

[0041] Figure 1 It is a logic flow chart of the present invention.

[0042] Figure 2 It is a schematic diagram of the structure of the multimodal network of the present invention.

[0043] Figure 3 It is a schematic diagram of the structure of the airborne electromagnetic data acquisition system in the present invention.

[0044] Figure 4 for Figure 3 Schematic diagram of the structure of the multi-component electromagnetic receiving coil.

[0045] Figure 5 for Figure 4 Schematic diagram of the structure of the X coil in the multi-component electromagnetic connection system.

[0046] In the above drawings, the component names corresponding to the reference numerals are as follows:

[0047] 1. UAV platform; 2. UAV to data acquisition system connection line; 3. Multi-component electromagnetic data storage device; 4. Transmitter coil connection line; 5. Wireframe carbon fiber rigid straight connecting rod; 6. Transmitter coil; 7. Aerial photography positioning and attitude module; 8. Multi-component electromagnetic receiving coil; 9. Multi-component data transmission data line; 10. Data acquisition system cable; 11. Transmitter coil power supply line; 12. Multi-component electromagnetic connection system X coil; 13. Multi-component electromagnetic connection system Y coil; 14. Multi-component electromagnetic connection system Z coil; 15. Magnetic field intensity component main receiving coil; 16. Magnetic field gradient component primary receiving coil; 17. Magnetic field gradient component secondary receiving coil; 18. First gap; 19. Second gap. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical scheme and advantages of this application clearer, the present invention is further described below in conjunction with the accompanying drawings and embodiments, and the embodiments of the present invention include but are not limited to the following embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0049] In this embodiment, the term "and / or" is merely a term used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist at the same time, and B exists alone.

[0050] The terms "first" and "second" in the description and claims of this embodiment are used to distinguish different objects rather than to describe a specific order of objects. For example, a first target object and a second target object are used to distinguish different target objects rather than to describe a specific order of target objects.

[0051] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0052] like Figures 1 to 5 As shown, this embodiment provides a method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection, which includes the following steps:

[0053] The first step is to use the electromagnetic detection equipment on the drone platform to collect and obtain the six-component magnetic field data (i.e. detection data) of the electromagnetic detection in the application scenario of the airborne electromagnetic method. Among them, the six-component magnetic field data includes the magnetic field intensity component , magnetic field intensity component , magnetic field intensity component , magnetic field gradient component ( ), magnetic field gradient component ( ) and magnetic field gradient components ( ). In this embodiment, the magnetic field intensity component , magnetic field intensity component , magnetic field intensity component It can reflect the electromagnetic response characteristics of the underground medium as a whole. , magnetic field gradient component and the magnetic field gradient component It can characterize the rate of change of the magnetic field in space and provide more detailed information about the boundaries, shapes and depths of underground anomalies. When detecting complex three-dimensional geological bodies underground, changes in magnetic field intensity can reflect possible areas of electrical differences, while changes in magnetic field gradients can help more accurately locate the boundaries and depth ranges of these abnormal areas.

[0054] Here, the electromagnetic detection equipment includes a transmitting coil 6 and a multi-component electromagnetic receiving coil 8 arranged at the bottom of the UAV platform, and an aerial photography positioning and attitude module 7 arranged at the bottom of the multi-component electromagnetic receiving coil 8. Among them, the aerial photography positioning and attitude module 7 collects the pitch angle, roll angle and yaw angle of the multi-component electromagnetic receiving coil 8. In addition, the multi-component electromagnetic receiving coil includes a multi-component electromagnetic connection system X coil 12, a multi-component electromagnetic connection system Y coil 13 and a multi-component electromagnetic connection system Z coil 14 that together form a spherical shape. The structures of the multi-component electromagnetic connection system X coil 12, the multi-component electromagnetic connection system Y coil 13 and the multi-component electromagnetic connection system Z coil 14 are the same. Then, taking the multi-component electromagnetic connection system X coil 12 as an example, it includes a magnetic field intensity component main receiving coil 15, a magnetic field gradient component primary receiving coil 16, and a magnetic field gradient component secondary receiving coil 17 arranged from the outside to the inside, a first gap 18 arranged between the magnetic field intensity component main receiving coil 15 and the magnetic field gradient component primary receiving coil 16, and a second gap 19 arranged between the magnetic field gradient component primary receiving coil 16 and the magnetic field gradient component secondary receiving coil 17. The magnetic field intensity component main receiving coil 15 is used to receive the magnetic field intensity component data, and the magnetic field gradient component primary receiving coil 16 and the magnetic field gradient component secondary receiving coil 17 are used to receive the magnetic field gradient component data.

[0055] by For example, the shape, number of turns and area parameters of the two coils of the magnetic field gradient component primary receiving coil 16 and the magnetic field gradient component secondary receiving coil 17 are the same. By measuring the difference of the induced electromotive force of the two coils and combining the distance between them ,use Calculate. Among them, represents the induced electromotive force of the primary receiving coil 16 of the magnetic field gradient component, The induced electromotive force of the secondary receiving coil 17 represents the magnetic field gradient component.

[0056] The second step is to preprocess the six-component magnetic field data as follows:

[0057] The aerial photography positioning module 7 collects the pitch angle, roll angle and yaw angle of the multi-component electromagnetic receiving coil 8, and uses a rotation matrix to respectively transform the magnetic field gradient components , magnetic field gradient component and the magnetic field gradient component Correction is performed, and its expression is:

[0058] ;in, represents the corrected magnetic field gradient, represents the attitude correction rotation matrix, represents the original magnetic field gradient, represents the pitch angle, represents the flip angle, Indicates the yaw angle.

[0059] Then, adaptive Kalman filtering is used to correct the magnetic field gradient component , magnetic field gradient component and the magnetic field gradient component Perform denoising.

[0060] In addition, wavelet threshold is used to analyze the magnetic field intensity components in the six-component magnetic field detection data. , magnetic field intensity component , magnetic field intensity component Perform denoising.

[0061] The third step is to preliminarily allocate the dynamic weights of the six-component magnetic field data based on geological prior knowledge and exploration scenarios. The expression is:

[0062] ;

[0063] ;in, represents the dynamic weight corresponding to the magnetic field intensity component, express Take separately , and The corresponding magnetic field intensity component , magnetic field intensity component , magnetic field intensity component ; represents the dynamic weight corresponding to the magnetic field gradient component, express Take separately , and The corresponding magnetic field gradient component , magnetic field gradient component and the magnetic field gradient component .

[0064] For example, in the area with obvious vertical stratification, the magnetic field intensity component Assign a higher weight (0.5-0.7) and increase the magnetic field gradient component in the lateral complex area Weight (0.3-0.4), the rest of the weight values ​​0-0.1. For example, in areas with complex lateral distribution of geological structures, the magnetic field gradient component , magnetic field gradient component The weight is increased to (0.4-0.5), the magnetic field strength component , magnetic field intensity component The weight of is between (0.1-0.2), and the rest of the weights are between (0-0.1).

[0065] The fourth step is to establish the electromagnetic field forward equation, and input the six-component magnetic field data after the preliminary allocation of dynamic weights into the electromagnetic field forward model to simulate the six-component magnetic field response, generate the corresponding simulated six-component magnetic field response data, and form the constraint conditions of the inversion process by comparing with the detection data. Here, the expression of the electromagnetic field forward equation is:

[0066] ; Wherein, m represents underground electrical parameters; Represents simulated probe data.

[0067] Regularization is used to optimize the underground electrical parameter m, and its expression is:

[0068] ;in, Represents detection data containing six-component magnetic field data; represents a sparse constraint; Represents a smoothness constraint.

[0069] Step 5: Obtain the magnetic field gradient components after preliminary allocation of dynamic weights , magnetic field gradient component and the magnetic field gradient component , the magnetic field gradient component after the initial allocation of the Laplace operator and dynamic weights The second derivative of the magnetic field gradient component The second derivative and magnetic field gradient component The second-order derivative of the gradient component (such as ) to further extract boundary features and combine with the Laplacian operator ( ) to enhance the edge response of the anomaly. For example, in salt dome detection, The high value areas of may correspond to the top of the salt dome, while The zero crossing points of may indicate boundaries.

[0070] The sixth step is to build a multimodal network and perform fusion processing.

[0071] The multimodal network includes an input layer, a feature extraction layer, a fusion layer and an output layer connected in sequence. The feature extraction layer consists of a parallel intensity component branch network and a gradient component branch network. The intensity component branch network consists of a three-layer convolutional neural network CNN and extracts vertical stratified features. The gradient component branch network consists of a three-layer convolutional neural network CNN and an edge detection layer, and performs boundary response enhancement of the edge of the abnormal body of a three-dimensional complex geological body. The fusion layer uses an attention mechanism to dynamically update the weights of the preprocessed six-component magnetic field data, and its expression is:

[0072] ;in, represents the Sigmoid function, , map the fully connected layer output to the (0, 1) interval to generate normalized weights; represents the activation function of the rectified linear unit. ; Represents the weight matrix of the first fully connected layer; Represents the weight matrix of the second fully connected layer; represents the global average pooling function; Represents the feature map of the Cth channel.

[0073] Here, the preprocessed six-component magnetic field data are input into the multimodal network, and the three-dimensional resistivity model or geological body probability distribution corresponding to the three-dimensional complex geological body is obtained.

[0074] The above embodiments are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any changes that adopt the design principles of the present invention and are made through non-creative work on this basis should fall within the protection scope of the present invention.

Claims

1. A method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection, characterized in that: The following steps are involved: Acquire six-component magnetic field detection data from airborne electromagnetic detection; The six-component magnetic field detection data includes a magnetic field intensity component , magnetic field intensity component , magnetic field intensity component , magnetic field gradient component , magnetic field gradient component and magnetic field gradient components ; Preprocessing the six-component magnetic field detection data to obtain preprocessed six-component magnetic field data; preliminarily allocating dynamic weights to the preprocessed six-component magnetic field data according to geological prior knowledge and exploration scenarios to obtain six-component magnetic field data after the dynamic weights are preliminarily allocated; Establish an electromagnetic field forward modeling equation, input the six-component magnetic field data after preliminary allocation of dynamic weights into the electromagnetic field forward model to simulate the six-component magnetic field response, generate the corresponding simulated six-component magnetic field response data, and form the constraint conditions of the inversion process by comparing with the six-component magnetic field detection data; Obtain the magnetic field gradient components after preliminary allocation of dynamic weights , magnetic field gradient component and the magnetic field gradient component , the magnetic field gradient component after the initial allocation of the Laplace operator and dynamic weights The second derivative of the magnetic field gradient component The second derivative and magnetic field gradient component The second-order derivative of is used to perform spatial enhancement of abnormal bodies in three-dimensional complex geological bodies; Building a multimodal network; the multimodal network includes an input layer, a feature extraction layer, a fusion layer and an output layer connected in sequence; the feature extraction layer is composed of a parallel intensity component branch network and a gradient component branch network; the intensity component branch network is composed of a 3-layer convolutional neural network CNN, and extracts vertical hierarchical features; The gradient component branch network is composed of a three-layer convolutional neural network (CNN) and an edge detection layer, and performs boundary response enhancement of the edge of the abnormal body of the three-dimensional complex geological body; the fusion layer uses an attention mechanism to dynamically update the weights of the preprocessed six-component magnetic field data; The preprocessed six-component magnetic field data is input into the multimodal network, and the three-dimensional resistivity model or geological body probability distribution corresponding to the three-dimensional complex geological body is obtained.

2. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 1, characterized in that: The magnetic field gradient component The expression is: ; The magnetic field gradient component The expression is: ; The magnetic field gradient component The expression is: .

3. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 1, characterized in that: The magnetic field intensity component , magnetic field intensity component , magnetic field intensity component An electromagnetic detection device is mounted on a drone platform for data collection; the electromagnetic detection device comprises a transmitting coil and a multi-component electromagnetic receiving coil arranged at the bottom of the drone platform, and an aerial photography positioning and attitude determination module arranged at the bottom of the multi-component electromagnetic receiving coil; the aerial photography positioning and attitude determination module collects the pitch angle, roll angle and yaw angle of the multi-component electromagnetic receiving coil; the multi-component electromagnetic receiving coil comprises a multi-component electromagnetic connection system X coil, a multi-component electromagnetic connection system Y coil and a multi-component electromagnetic connection system Z coil which together form a spherical shape; the multi-component electromagnetic connection system X coil, the multi-component electromagnetic connection system Y coil and the multi-component electromagnetic connection system Z coil have the same structure, and the multi-component electromagnetic connection system X coil comprises a main receiving coil of a magnetic field intensity component, a primary receiving coil of a magnetic field gradient component, and a secondary receiving coil of a magnetic field gradient component which are arranged in sequence from the outside to the inside, a first gap is arranged between the main receiving coil of the magnetic field intensity component and the primary receiving coil of the magnetic field gradient component, and a second gap is arranged between the primary receiving coil of the magnetic field gradient component and the secondary receiving coil of the magnetic field gradient component.

4. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 3, characterized in that: Preprocess the six-component magnetic field detection data, including: Combine the pitch angle, roll angle and yaw angle, and use the rotation matrix to respectively transform the magnetic field gradient components in the six-component magnetic field detection data , magnetic field gradient component and magnetic field gradient components Correction is performed, and its expression is: ;in, represents the corrected magnetic field gradient, represents the attitude correction rotation matrix, represents the original magnetic field gradient, represents the pitch angle, represents the flip angle, represents the yaw angle; The wavelet threshold is used to analyze the magnetic field intensity components in the six-component magnetic field detection data. , magnetic field intensity component , magnetic field intensity component Perform denoising; Adaptive Kalman filtering is used to correct the magnetic field gradient component , magnetic field gradient component and magnetic field gradient components Perform denoising.

5. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 4, characterized in that: According to the preliminary allocation of dynamic weights of the six-component magnetic field data after preprocessing based on geological prior knowledge and exploration scenarios, the six-component magnetic field data after preliminary allocation of dynamic weights are obtained, and its expression is: ; ;in, represents the dynamic weight corresponding to the magnetic field intensity component, express Take separately , and The corresponding magnetic field intensity component , magnetic field intensity component , magnetic field intensity component ; represents the dynamic weight corresponding to the magnetic field gradient component, express Take separately , and The corresponding magnetic field gradient component , magnetic field gradient component and the magnetic field gradient component .

6. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 4, characterized in that: An electromagnetic field forward modeling equation is established, and the six-component magnetic field data after preliminary allocation of dynamic weights are jointly input into the electromagnetic field forward model to simulate the six-component magnetic field response; the electromagnetic field forward modeling equation is expressed as: ; Wherein, m represents underground electrical parameters; represents simulated detection data; Regularization is used to optimize the underground electrical parameter m, and its expression is: ;in, Represents detection data containing six-component magnetic field data; represents a sparse constraint; Represents a smoothness constraint.

7. The method for processing airborne electromagnetic multi-component data for three-dimensional complex geological body detection according to claim 4, characterized in that: The fusion layer uses the attention mechanism to dynamically update the weights of the preprocessed six-component magnetic field data, and its expression is: ;in, Represents the Sigmoid function, which maps the output of the fully connected layer to the interval (0, 1) and generates normalized weights; represents the activation function of the rectified linear unit; Represents the weight matrix of the first fully connected layer; Represents the weight matrix of the second fully connected layer; represents the global average pooling function; Represents the feature map of the Cth channel.

Images