Complex medium controllable source electromagnetic multi-parameter three-dimensional inversion method and system

Abstract

The invention discloses a complex medium controllable source electromagnetic multi-parameter three-dimensional inversion method and system, and the method comprises the steps: carrying out the three-dimensional space discretization of an exploration region, and carrying out the multi-time iteration inversion of a discrete exploration region through initializing a multi-parameter geologic model based on a multi-parameter inversion target function, and obtaining a multi-parameter three-dimensional inversion result; wherein the multi-parameter inversion objective function comprises a data fitting constraint, a model constraint and a cross gradient constraint, and the multi-parameter model comprises resistivity, magnetic conductivity, a charging rate, a frequency correlation coefficient and a time constant; and a hexahedral mesh discrete calculation region is adopted in inversion, and a finite element linear equation set is obtained through a vector finite element method, so that multi-parameter three-dimensional forward modeling calculation is realized. According to the method, the induced polarization effect and the magnetic conductivity influence are considered at the same time, multi-parameter three-dimensional inversion is achieved, the application precision and reliability of the controllable source electromagnetic method under the complex geological condition are remarkably improved, and rich geophysical basis is provided.

Description

Technical Field

[0001] This invention belongs to the field of controllable source electromagnetics, specifically relating to a method and system for multi-parameter three-dimensional inversion of controllable source electromagnetics in complex media. Background Technology

[0002] Controlled-source electromagnetic (CSEM) in the frequency domain is an artificial source electromagnetic exploration method based on the principle of electromagnetic induction. It is widely used in oil and gas, mineral resource exploration, and hydrogeology, offering advantages such as high resolution and strong anti-interference capabilities. Geophysical inversion is essentially a mapping problem from data space to model space. Since observational data objectively exist, a solution to the inverse problem must exist. However, errors in the observational data cause instability in the solution, making geophysical inversion an ill-posed problem. Traditional controlled-source electromagnetic forward and inversion methods mostly consider only resistivity. However, the response of actual geological bodies to electromagnetic waves is a comprehensive reflection of multiple electrical parameters. Ignoring the influence of parameters such as polarizability and permeability significantly reduces the accuracy of the inversion interpretation.

[0003] To improve the interpretation of CSEM data in metal mineral clusters, it is crucial to conduct efficient three-dimensional CSEM inversion studies in the frequency domain, considering the effects of induced polarization and magnetic permeability, and applicable to complex geological conditions. Existing technologies mainly focus on the following aspects:

[0004] (1) Using multiple geophysical methods or multi-parameter joint inversion. To accurately characterize complex subsurface media, modern geophysics has developed two joint inversion methods: on the one hand, by combining multiple geophysical methods and utilizing their complementary data to form cross-constraints on the model; on the other hand, by simultaneously extracting multiple intrinsically related physical property parameters from a single dataset and performing multi-parameter joint inversion. Both methods can effectively reduce ambiguity, significantly improve the accuracy of inversion results and the reliability of geological interpretation, and are conducive to improving the level of data interpretation.

[0005] (2) Three-dimensional controlled-source electromagnetic inversion has become an indispensable key technology. To further address the problems that one-dimensional and two-dimensional inversions cannot interpret or may misinterpret, three-dimensional controlled-source electromagnetic inversion reconstructs the real three-dimensional physical structure underground in a way that is closer to physical reality, and restores the three-dimensional propagation characteristics of electromagnetic fields. It can effectively solve the exploration needs in complex geological environments, making the inversion results more unique and reliable.

[0006] Despite significant advancements in frequency-domain CSEM forward and inverse modeling, the frequency-domain controlled-source electromagnetic method faces serious challenges in terms of detection accuracy and interpretation capabilities as geological exploration becomes increasingly sophisticated. Existing techniques, particularly when dealing with complex geological structures, still suffer from the following major deficiencies and shortcomings:

[0007] (1) CSEM three-dimensional inversion has significant multiple solutions, and considering resistivity, polarizability and permeability, the number of unknowns in the inversion increases to several times the original, which will greatly exacerbate the ill-conditioning of the inversion problem; in addition, the volume effect of the electromagnetic method makes the inversion results exhibit divergent characteristics, further increasing the difficulty of interpretation.

[0008] (2) Current research has only independently analyzed the effects of IP effect and permeability parameter in three-dimensional forward modeling and a few inversions, while there are very few frequency domain CSEM three-dimensional inversion studies that consider the effects of induced polarization and permeability simultaneously. In order to comprehensively describe the electromagnetic response characteristics of geological bodies, the inversion interpretation of frequency domain CSEM data must take into account the effects of induced polarization parameters (charge rate, time constant and frequency correlation coefficient) and permeability. Summary of the Invention

[0009] This invention provides a controllable source electromagnetic multi-parameter three-dimensional inversion method and system for complex media. It takes into account the effects of induced polarization and magnetic permeability, and realizes multi-parameter three-dimensional inversion, which significantly improves the accuracy and reliability of the controllable source electromagnetic method under complex geological conditions.

[0010] To achieve the above technical objectives, the present invention adopts the following technical solution:

[0011] A method for three-dimensional electromagnetic multi-parameter inversion of a controllable source in a complex medium includes:

[0012] The exploration area is discretized in three-dimensional space. For the discretized exploration area: first, the multi-parameter model of the controllable source electromagnetic is initialized. Then, through the constructed multi-parameter inversion objective function, the multi-parameter model of the controllable source electromagnetic is iterated and inverted multiple times to obtain the controllable source electromagnetic multi-parameter three-dimensional inversion result of the exploration area.

[0013] The multi-parameter inversion objective function includes data fitting constraints, model constraints, and cross-gradient constraints. The data fitting constraints are used to constrain the electromagnetic response data fitted based on the multi-parameter model obtained from the inversion and the measured electromagnetic response data. The model constraints constrain the multi-parameter model obtained from the inversion relative to the reference model. The cross-gradient constraints refer to cross-constraints between different parameters obtained from the inversion.

[0014] The multi-parameter model includes resistivity, permeability, charge rate, frequency correlation coefficient, and time constant.

[0015] Furthermore, the electromagnetic response data specifically refers to electric and magnetic field data, or apparent parameters calculated from electric and magnetic field data.

[0016] Furthermore, the initialization of the multi-parameter model specifically involves: using existing prior information or a uniform half-space model, assigning initial values ​​to the parameters of each discrete unit in the exploration area.

[0017] Furthermore, the multi-parameter inversion objective function Represented as: ; In the formula, For data fitting constraints, For model constraints, For cross-gradient constraint terms; and As a regularization factor, For the multi-parameter model to be inverted, These represent resistivity, relative permeability, charge rate, frequency correlation coefficient, and time constant, respectively. The constraints are as follows: ; ; ; In the formula, These are measured electromagnetic response data, used as observational data. The electromagnetic response data is for fitting. For observation data with fitted data The covariance matrix between them; As a reference model, For the multi-parameter model to be inverted Compared with the reference model The model covariance matrix between them For multi-parameter models Discrete model vectors with two different parameters. Discrete model vector Spatial gradient vector field.

[0018] Furthermore, each iteration of the inversion uses a nonlinear conjugate gradient algorithm to optimize the objective function, resulting in a multi-parameter model.

[0019] Furthermore, the process of iteratively inverting the multi-parameter model of the controllable source electromagnetics using the constructed multi-parameter inversion objective function includes:

[0020] Using the current multi-parameter model and performing three-dimensional multi-parameter forward modeling based on the differential control equation of electric field intensity, theoretical electromagnetic response data, i.e., fitted data, are calculated. Then, the fitted data and observation data are substituted into the inversion objective function to optimize the calculation and update the multi-parameter three-dimensional model. The process is iterated until the fitted data and observation data meet the convergence condition, and the final multi-parameter three-dimensional model is output as the interpretation result of the physical structure of the underground medium.

[0021] The differential governing equation for the electric field strength is: ; In the formula, Where is curl, and E is electric field strength. The imaginary unit, Angular frequency, Let k be the applied source current density, and k be the wave number in the quasi-static limit. ; To investigate the magnetic permeability of the underground medium in the exploration area, , The permeability of free space, This refers to the relative permeability; when considering the influence of the permeability of the underground medium, Not equal to 1; To determine the electrical conductivity of the subsurface medium in the exploration area, since electrical conductivity and resistivity are reciprocals, when the subsurface medium contains induced polarization effects, the following Cole-Cole model is used to describe the electrical conductivity: ; In the formula, m, c These represent zero-frequency resistivity, charge rate, frequency correlation coefficient, and time constant, respectively.

[0022] Furthermore, the exploration area is discretized in three-dimensional space using a hexahedral mesh, and the differential governing equation of the electric field intensity E is solved using the vector finite element method, specifically including:

[0023] First, the calculation exploration area is divided into multiple mutually exclusive hexahedral units;

[0024] Then, let the edge lengths of any element e in the x, y, and z directions be respectively... The coordinates of the unit center are Then any point inside the unit electric field strength at Represented as: ; In the formula, edge Vector basis functions at the midpoint edge The electric field value at the midpoint;

[0025] Then, using the Galerkin method, we obtain the finite element linear equation system:

[0026] (a1) Calculate the residuals on both sides of the differential governing equation for the electric field strength, and then compare them with the vector basis functions. Take the inner product and set it to 0, that is: (1); (a2) To handle the double curl term, we use the vector identity transformation to obtain: ; Let the variables in the vector identity Substituting and applying the Gaussian divergence theorem, we obtain ; Substitute the above equation back into equation (1) and expand. ,get ; (a3) will Substitute and rearrange to obtain the unit residuals : ; In the formula, Represents any hexahedral element At the edge The weighted residuals; and These are all indices for the degrees of freedom of the edges; The integral range represents the current hexahedral element. ; (a4) Rewrite the above equation in matrix form: ; In the formula, and These are the curl matrix and the mass matrix, respectively; For unit The source term is a vector of size 12×1, determined by the magnitude, position, and direction of the electric source. For any j-th source... Represented as ; (a5) Combine all the element matrices into a global matrix and set the total weighted residual to 0 to obtain a large-scale linear system of finite element equations: ; In the formula, It is a symmetric semidefinite matrix. It is a symmetric positive definite matrix. For all unit field source terms The field source term constituted, denoted as electric field strength.

[0027] A complex medium controllable source electromagnetic multi-parameter three-dimensional inversion system includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor enables the processor to implement the complex medium controllable source electromagnetic multi-parameter three-dimensional inversion method described above.

[0028] Compared with the prior art, the beneficial effects of the present invention are as follows: When considering both polarizability and permeability parameters simultaneously, the electromagnetic field formula becomes more complex and has multiple solutions, making it difficult for existing techniques to simultaneously consider and perform three-dimensional inversion. This invention, based on the vector finite element method, uses a hexahedral mesh for spatial discretization, employs cross-gradient structure constraints to construct a multi-parameter inversion objective function, and optimizes the objective function using a nonlinear conjugate gradient algorithm. This achieves multi-parameter three-dimensional inversion considering resistivity, polarizability, and permeability. It significantly enhances the accuracy of data interpretation and the reliability of inversion using the controlled-source electromagnetic method under complex geological conditions. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the block model and observation device described in an embodiment of the present invention.

[0030] Figure 2 This is the curve showing the decrease in mean square error of the inversion iteration as described in this embodiment of the invention.

[0031] Figure 3 The conductivity inversion result is the result of the method and system described in the embodiments of the present invention.

[0032] Figure 4 This is the charging rate inversion result of the method and system described in the embodiments of the present invention.

[0033] Figure 5 It is the time constant inversion result of the method and system described in the embodiments of the present invention.

[0034] Figure 6 It is the frequency correlation coefficient inversion result of the method and system described in the embodiments of the present invention.

[0035] Figure 7 This is the relative permeability inversion result of the method and system described in the embodiments of the present invention.

[0036] Figures 3-7 The sub-figures (a), (b), (c), and (d) are respectively: based on the established XYZ spatial rectangular coordinate system, they are the pseudo-section diagrams at y=5000m, x=0m, and z=600m, and the schematic diagram of the final three-dimensional inversion result. Detailed Implementation

[0037] The embodiments of the present invention will be described in detail below. These embodiments are based on the technical solutions of the present invention and provide detailed implementation methods and specific operation processes to further explain the technical solutions of the present invention.

[0038] Example 1

[0039] This embodiment sets up a block-shaped model with a resistivity of 10. The charging rate is 0.2, the time constant is 10, the frequency coefficient is 0.1, and the relative permeability is 2. Its background resistivity is 100. There are no anomalies in polarizability and permeability. Two emission sources are set up: the measurement points of source 1 are green dots, and the measurement points of source 2 are red triangles, with some overlap between the measurement points. Figure 1 As shown. The frequencies of field source 1 are 512 / 128 / 32 / 8 Hz, and the frequencies of field source 2 are 8192 / 2048 / 512 / 128 / 32 / 8 / 2 Hz. The initial model is a uniform half-space with a resistivity of 100. The relative permeability is 1, the charge rate is 0.1, the time constant is 10, and the frequency correlation coefficient is 0.1. The electric field x-component E generated by this theoretical model... x It is treated as field observation data and multi-parameter inversion calculation is performed.

[0040] Figure 2 The mean squared error between the observed data and the final predicted data is used as the basis for terminating the inversion after 50 iterations, which yields a lower data fitting result. The inversion results for each parameter are as follows: Figures 3-7 As shown, the inversion results for conductivity, charge rate, time constant, frequency correlation coefficient, and relative permeability are presented in sequence. Each parameter inversion result contains four sub-plots: a pseudo-section plot at y=5000m, x=0m, z=600m, based on the established XYZ Cartesian coordinate system, and a schematic diagram of the final three-dimensional inversion result. Figure 3-7 The multi-parameter inversion results shown indicate that the distributions of resistivity, charge rate, time constant, and frequency coefficients obtained in the final model largely correspond to the positions of the actual model, indicating a good inversion result. The relative permeability inversion effect is slightly worse. This is mainly because the high relative permeability anomaly cancels out the electromagnetic field anomalies generated by resistivity and polarizability. However, the data as a whole still exhibits low resistivity characteristics. Therefore, parameters such as resistivity and polarizability can still invert subsurface electrical properties, while the relative permeability inversion effect is slightly reduced. In summary, this invention, considering multi-parameter inversion of resistivity, polarizability, and permeability, can better recover the parameters of the actual model, and multi-parameter inversion also provides richer geophysical evidence for geological interpretation.

[0041] Ignoring the effects of induced polarization and permeability exacerbates the uncertainty in the interpretation of frequency-domain controllable source electromagnetic data inversion, while simultaneously inverting multiple parameters inevitably leads to a sharp increase in the ambiguity of the inversion. Therefore, this invention provides a multi-parameter three-dimensional inversion method for controllable source electromagnetic data in complex media. The method discretizes the exploration area in three-dimensional space and, for the discrete exploration area, initializes a multi-parameter geological model and performs multiple iterative inversions based on a multi-parameter inversion objective function that simultaneously considers subsurface polarization and permeability, to obtain the multi-parameter three-dimensional inversion results for the exploration area.

[0042] The specific implementation of the controllable source electromagnetic multi-parameter three-dimensional inversion method for complex media includes the following steps: Step 1: Read measurement data from the field measurement equipment. Using the measured electric and magnetic field components, calculate the apparent resistivity, apparent polarizability, and apparent permeability of each parameter. When calculating these apparent parameters, noisy measurement data needs to be processed. Specifically, a median filtering algorithm can be used to remove outlier noise: scan the data with a certain window size, and use the median of the data within the window as the new data value for that point. Additionally, a wavelet transform algorithm is used to decompose the data into multiple scales, thresholding the noise coefficients at each scale to remove high-frequency noise. Finally, the data is reconstructed using inverse wavelet transform, completing the noise data processing.

[0043] Step 2, construct the multi-parameter inversion objective function This includes data fitting constraints, model constraints, and cross-gradient constraints, expressed as: ; In the formula, For data fitting constraints, For model constraints, For cross-gradient constraint terms; and As a regularization factor, it plays a role in balancing the data fitting term, model constraint term, and cross-gradient constraint term. For the multi-parameter model to be inverted, These represent resistivity, relative permeability, charge rate, frequency correlation coefficient, and time constant, respectively. Therefore, the multi-parameter inversion objective function in this embodiment simultaneously considers resistivity, polarizability, and permeability.

[0044] (1) Data fitting constraint: used to constrain the electromagnetic response data fitted by the multi-parameter model obtained from the inversion and the measured electromagnetic response data.

[0045] ; In the formula, For observational data, there is an N-dimensional data vector representing the measured electromagnetic response data. The electromagnetic response data is for fitting. The data covariance matrix is ​​represented as an N×N diagonal matrix, with the reciprocals of the data standard deviations on the diagonal. In this embodiment, the electromagnetic response data can be electromagnetic field components from different measurement points, different field sources, and different frequencies, or it can be apparent parameter values ​​calculated using electromagnetic fields.

[0046] The electromagnetic response data The fitting method is as follows: based on the differential control equation of electric field intensity, and by forward modeling the multi-parameter model obtained from the previous iteration inversion, the fitted electromagnetic response data is obtained.

[0047] In this embodiment, under the quasi-static limit condition ( (The effect of displacement current can be ignored), assuming the time harmonic factor is... Then the differential governing equation for the electric field strength E is: ; Where E is the electric field strength (V / m). and These are the dielectric conductivity (S / m) and magnetic permeability (H / m), respectively. The imaginary unit, ω is the angular frequency (rad / s). For the applied source current density Let k be the wave number under the quasi-static limit. The permeability of the medium is: , The value is the vacuum permeability (H / m). This refers to the relative permeability; when considering the influence of the permeability of the underground medium, The value is not 1. When the underground medium contains induced polarization, its resistivity is a complex number that changes with the power supply frequency. Therefore, the Cole-Cole model is introduced, which can accurately describe the conductivity dispersion phenomenon: ; The above formula contains four parameters m, c and , respectively, represent zero-frequency resistivity (i.e., resistivity as commonly referred to), charge rate, frequency correlation coefficient, and time constant. It is the imaginary unit.

[0048] The differential governing equation for the electric field strength E described above describes the electric field strength at angular frequency E. Under time-harmonic excitation, the electric field With source current density And the logical relationship between the electromagnetic parameters of the medium. The electric field intensity E is the direct calculation object in the electromagnetic forward modeling. After solving the differential governing equation to obtain the distribution of the electric field in the underground space (this embodiment will later introduce the finite element discretization of the governing equation to obtain a large-scale linear finite element equation system, which is used to directly solve the unknown electric field values ​​on all discrete elements in the calculation area), observable electromagnetic response data such as magnetic field H and induced electromotive force can be further derived.

[0049] In this embodiment, the vector finite element method is used to discretize the exploration area in three-dimensional space using a hexahedral mesh, dividing the computational region into multiple mutually exclusive hexahedral elements. Then, it is assumed that the edge lengths of any element e in the x, y, and z directions are respectively... The coordinates of the unit center are Then the electric field intensity at any point (x,y,z) inside the unit can be expressed as: ; in, edge Vector basis functions at the midpoint edge The electric field value at the midpoint.

[0050] Then, the Galerkin method is used to construct a system of linear equations for the finite element method: (a1) Calculate the residuals on both sides of the differential governing equation for the electric field strength, and then compare them with the vector basis functions. Take the inner product and set it to 0, that is: (1); (a2) To handle the double curl term, we use the vector identity transformation to obtain: ; Let the variables in the vector identity Substituting and applying the Gaussian divergence theorem, we obtain ; Substitute the above equation back into equation (1) and expand. ,get ; (a3) will Substitute and rearrange to obtain the unit residuals : ; In the formula, Represents any hexahedral element At the edge The weighted residuals; and These are all indices for the degrees of freedom of the edges; The integral range represents the current hexahedral element. ; (a4) Rewrite the above equation in matrix form: ; In the formula, and These are the curl matrix and the mass matrix, respectively; For unit The source term is a vector of size 12×1, determined by the magnitude, position, and direction of the electric source. For any j-th source... Represented as ;in , The size is 12×12, and the specific calculation formula is as follows: ; ; ; The matrices involved in the above formula are all 4×4 in size, and p represents any one of the x, y, and z directions in three-dimensional space.

[0051] (a5) Combine all the element matrices into a global matrix and set the total weighted residual to 0 to obtain a large-scale linear system of finite element equations: ; In the formula, It is a symmetric semidefinite matrix. It is a symmetric positive definite matrix. For all unit field source terms The field source term constituted, denoted as electric field strength.

[0052] (2) Model constraints are used to constrain the multi-parameter model obtained by inversion relative to the reference model.

[0053] Based on existing geological data, prior information, or previous exploration results in the exploration area, an initial geological model is constructed that includes parameters such as resistivity, polarizability, and magnetic permeability.

[0054] ; In the formula, As a reference model, For the multi-parameter model to be inverted Compared with the reference model The model covariance matrix can be determined by either a first-order gradient operator or a second-order Laplace operator.

[0055] (3) Cross gradient constraints are used to cross-constrain the different parameters obtained by inversion.

[0056] A cross-gradient operator is used for structural similarity constraints. An inversion objective function considering resistivity, polarizability, and permeability is established. During the construction of the cross-gradient structural constraint term, it is assumed that there is structural similarity between different parameters. However, if there is no obvious structural similarity between different parameters, especially between the permeability parameter and the other electrical parameters, a smaller constraint weight can be applied to the corresponding cross-gradient term to weaken its influence.

[0057] ; The above formula represents the sum of the L2 norms of the cross gradients between the two parameters. For multi-parameter models Discrete model vectors with different parameters in the model. Discrete model vector The spatial gradient vector field is the discrete model vector. The gradient.

[0058] Step 3: Set the multi-threaded parallel parameters.

[0059] In this embodiment, the inversion parameters include five: resistivity, relative permeability, charge rate, frequency correlation coefficient, and time constant. Increasing the number of parameters leads to a significant increase in computational load. Furthermore, this invention inverts a three-dimensional space, which also results in a large computational burden. In summary, the computational load for multi-parameter three-dimensional inversion is substantial; therefore, this embodiment employs multi-threaded parallel processing to improve computational efficiency. Accordingly, when setting the inversion parameters, this embodiment needs to determine the number of threads participating in parallel computation, allocating them reasonably based on the number and performance of the computer hardware cores to fully utilize computing resources.

[0060] Furthermore, when performing the inversion using multi-threaded parallelism in this embodiment, a data sharing and synchronization mechanism is set up between threads to ensure data consistency and computational accuracy. To improve the computational efficiency of the inversion, this embodiment uses the MPI-OpenMP hybrid parallel algorithm to accelerate the joint inversion algorithm.

[0061] Step 4: Initialize the multi-parameter model.

[0062] Specifically, by utilizing existing prior information or a uniform half-space model, initial values ​​are assigned to the parameters of each discrete unit in the exploration area.

[0063] Step 5: Initiate iterative inversion based on the multi-parameter inversion objective function.

[0064] This embodiment employs a nonlinear conjugate gradient algorithm to optimize the objective function. During the inversion iteration process, it only requires calculating the gradient of the objective function, offering advantages such as lower computational cost, less memory usage, and faster convergence. In each iteration, based on the iterative updates of the model parameters, the updated model parameters are calculated using the current gradient vector, observation data, forward prediction data, and various constraints. Through continuous iteration, the optimal solution is obtained.

[0065] In this embodiment, in each inversion iteration, the model parameters obtained from the inversion are substituted back into the forward calculation to obtain the theoretical electromagnetic response data corresponding to the inversion model. This data reflects the electromagnetic signals generated by the underground geological body represented by the current inversion model. Subsequently, the theoretical electromagnetic response data is compared in detail with the original observation data, and the degree of data fit is evaluated by calculating the fit difference between the two. Commonly used indicators for evaluating the fit difference include mean squared error (MSE). The specific calculation process is as follows: First, the square of the difference between the theoretical response and the observed data at each data point is calculated. Then, the squared differences of all data points are summed, and then divided by the total number of data points to obtain the mean squared error value. This MSE is compared with a preset threshold to determine whether the degree of data fit meets the requirements. If the MSE is less than or equal to the threshold, the inversion model is considered to fit the actual observation data well, and the inversion result is reliable; otherwise, the inversion iteration needs to continue until the fit difference meets the set conditions.

[0066] Step 6: Export the multi-parameter inversion results and carry out geological interpretation work.

[0067] Once the fitting degree reaches the preset standard, it indicates that the inversion calculation has achieved satisfactory results. At this point, the multi-parameter results obtained from the inversion can be exported. These parameters include the distribution information of various electrical parameters such as resistivity, polarizability, and magnetic permeability of each underground unit. They are usually stored and output in text files or database tables, with the stratigraphic location and specific values ​​of each parameter clearly marked during output to ensure data accuracy and readability. Based on the exported multi-parameter inversion results, combined with the geological background of the mining area, historical exploration data, and relevant geological theories, further geological interpretation work can be carried out. Interpreters analyze the structure, texture, and lithological variations of underground geological bodies based on the distribution characteristics of different electrical parameters. For example, resistivity distribution can identify high-resistivity and low-resistivity regions, inferring lithological differences; polarizability distribution characteristics can identify geological bodies with excitation polarization effects, inferring areas that may contain mineralized deposits. Simultaneously, combined with geological structural theory, the relationship between the distribution of geological bodies and structures is analyzed to determine whether there are geological conditions favorable for mineralization and ore reserves. Based on the above analysis results, geological profile maps, structural maps, and other maps can be compiled to provide intuitive and accurate geological data for mineral exploration and guide subsequent exploration work, such as determining drilling target areas and assessing resource reserves.

[0068] The above embodiments are preferred embodiments of this application. Those skilled in the art can make various changes or improvements based on them. Without departing from the overall concept of this application, these changes or improvements should fall within the scope of protection claimed in this application.

Claims

1. A complex medium controlled source electromagnetic multi-parameter three-dimensional inversion method, characterized in that, The method comprises the following steps: The survey area is discretized in three-dimensional space, and for the discretized survey area: firstly, a multi-parameter model of controlled source electromagnetic is initialized, and then the multi-parameter model of controlled source electromagnetic is iteratively inverted multiple times by constructing a multi-parameter inversion objective function to obtain a multi-parameter three-dimensional inversion result of controlled source electromagnetic of the survey area; The multi-parameter inversion objective function comprises a data fitting constraint, a model constraint and a cross-gradient constraint; the data fitting constraint is used to constrain the electromagnetic response data fitted based on the obtained multi-parameter model and the measured electromagnetic response data; the model constraint is used to constrain the obtained multi-parameter model relative to a reference model; and the cross-gradient constraint is used to cross-constrain different parameters in the obtained multi-parameter model. The multi-parameter model comprises resistivity, magnetic permeability, chargeability, frequency-dependent coefficient and time constant.

2. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 1, wherein, The electromagnetic response data specifically refer to electric field and magnetic field data or apparent parameters calculated from the electric field and magnetic field data.

3. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 1, wherein, The initialization of the multi-parameter model specifically comprises: using existing prior information or a uniform half-space model to assign initial values to each parameter of each discretized unit of the survey area.

4. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 1, wherein, The multi-parameter inversion objective function is represented as: ； wherein is a data fitting constraint, is a model constraint, is a cross-gradient constraint term; and is a regularization factor, is a multi-parameter model to be inverted, respectively represent resistivity, relative permeability, chargeability, frequency dependent coefficient and time constant. The constraints are as follows: ； ； ； wherein is the measured electromagnetic response data, as the observed data; is the fitted electromagnetic response data; is the observed data and the fitted data is the covariance matrix between the observed data is the reference model, is the multi-parameter model to be inverted is the model covariance matrix between the reference model and the multi-parameter model is the multi-parameter model is the discrete model vector of different parameters in the multi-parameter model is the spatial gradient vector field of the discrete model vector .

5. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 1, wherein, Each iteration inversion uses a nonlinear conjugate gradient algorithm to optimize the objective function to obtain the multi-parameter model.

6. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 1, wherein, The multi-parameter model is iteratively inverted multiple times by constructing the multi-parameter inversion objective function, which comprises the following steps: The current multi-parameter model is used to perform three-dimensional multi-parameter forward modeling based on the electric field intensity differential control equation to calculate theoretical electromagnetic response data, i.e., fitting data; then the fitting data and the observation data are substituted into the inversion objective function to optimize and update the multi-parameter three-dimensional model; the iteration is continuously performed until the fitting data and the observation data meet the convergence condition, and the final multi-parameter three-dimensional model is output as the interpretation result of the underground medium physical structure; The differential control equation of the electric field intensity is as follows: ； wherein is the curl, E is the electric field intensity, is the imaginary unit, is the angular frequency, is the impressed field source current density, k is the wave number at the quasi-static limit, ; is the magnetic permeability of the subsurface medium in the survey area, , is the vacuum magnetic permeability, is the relative magnetic permeability; when the effect of the magnetic permeability of the subsurface medium is considered, is not 1; is the electrical conductivity of the subsurface medium in the survey area, the electrical conductivity and the resistivity are reciprocals of each other, when the subsurface medium contains the excitation polarization effect, the following Cole-Cole model is used to describe the electrical conductivity: ； In the formula, , m, c, respectively represent zero frequency resistivity, charging rate, frequency correlation coefficient and time constant.

7. The complex media controlled source electromagnetic multi-parameter 3D inversion method of claim 6, wherein, The survey area is discretized in three-dimensional space by using a hexahedral grid, and the differential control equation of the electric field intensity E is solved by using a vector finite element method, which specifically comprises the following steps: Firstly, the survey area is divided into multiple mutually exclusive hexahedral units; Then, let the edge length of an arbitrary unit e in the x, y, z three directions be , the unit center coordinates be , and the electric field intensity at an arbitrary point inside the unit be , which is expressed as: ； wherein is an edge vector basis function of the midpoint, is an edge electric field value of the midpoint; Then, the Galerkin method is used to obtain a finite element linear equation group: (a1) calculating the residual of the left and right sides of the differential control equation of the electric field intensity, and then taking the inner product with the vector basis function making the inner product equal to zero, i.e.: (1)； (a2) In order to process the double spinor term, the vector identity is transformed to obtain: ； Let the variables in the vector identity Substitute and apply the Gauss divergence theorem to obtain ； Substitute the above equation into equation (1) and expand to obtain ； (a3) adding Substituting and rearranging, the unit residual : ； wherein represents an arbitrary hexahedral element at an edge of the weighted residual error; and are indices of the edge degrees of freedom; denotes the integral range is the current hexahedral element ; (a4) The above formula is written in matrix form: ； wherein and are the curl matrix and the mass matrix, respectively; is the field source term of the element is a 12 x 1 vector determined by the size, location, and direction of the electric source, and the arbitrary jth field source is expressed as ; (a5) All the element matrices are combined into a total matrix, and the total weighted residual error is set to 0 to obtain a large finite element linear equation group: ； wherein is a symmetric semi-definite matrix, is a symmetric positive definite matrix, is a field source term of all unit cells comprising field source terms, is the electric field intensity.

8. A complex medium controlled source electromagnetic multi-parameter 3D inversion system comprising a memory and a processor, the memory having stored therein a computer program, characterized in that, The computer program is executed by the processor to implement the complex medium controlled source electromagnetic multi-parameter three-dimensional inversion method according to any one of claims 1-7.

Images