Comprehensive inversion method and system based on bidirectional magnetic susceptibility weighted constraint

Abstract

The invention relates to the technical field of geophysics, and particularly discloses a comprehensive inversion method and system based on bidirectional magnetic susceptibility weighted constraint, and the method comprises the steps: dividing an underground three-dimensional space, and building two different cuboid models; determining a total magnetic field observed on the ground and gradient anomaly thereof based on the observation point coordinates and the cuboid model position, and establishing a kernel function matrix; determining an inversion objective function reflecting the relationship between the total magnetic field and the gradient anomaly thereof and the physical property parameters; adding the bidirectional magnetic susceptibility weighted constraint function into an inversion objective function; the optimal solution of the comprehensive inversion target function is solved based on the conjugate gradient method, the high-level resolution capability of the magnetic gradient data and the deep detection advantage of the total magnetic field data are fused, the precision of fine depiction of the superficial anomalous body boundary and the reconstruction precision of the deep field source structure are remarkably improved, a vector cross multiplication regularization constraint mechanism is introduced, and the method has the advantages of high resolution and high resolution. The inversion multiplicity under the complex geological condition is greatly inhibited, and the reliability of the model result and the geological rationality are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of geophysics, and in particular to a comprehensive inversion method and system based on bidirectional susceptibility weighted constraint. Background Technique

[0002] As an important means of mineral resource exploration, geophysical magnetic exploration can effectively identify the spatial distribution of ore bodies by inverting the magnetization rate distribution characteristics of underground magnetic bodies. Traditional magnetic inversion methods mainly rely on single total magnetic field anomaly data and solve the distribution of underground physical property parameters by establishing a linearized inversion equation. However, this method has significant technical bottlenecks: on the one hand, the total magnetic field anomaly is greatly affected by the depth of the field source, and the shallow anomaly signal is easy to cover the weak anomaly characteristics in the deep part, resulting in insufficient vertical resolution; on the other hand, there is a serious problem of non-uniqueness in the inversion of single data source, and it is difficult to accurately depict the ore body boundary under complex geological conditions.

[0003] In recent years, with the development of magnetic gradient measurement technology, it has become a research hotspot to improve the inversion resolution by using the high-precision and multi-component characteristics of magnetic gradient tensor data. Theoretically, magnetic gradient data has higher horizontal resolution, but its signal intensity decays faster with distance and is less sensitive to deep field sources. Although existing joint inversion methods attempt to combine total magnetic field and gradient data, there are still the following technical defects in practical applications: (1) lack of an effective data fusion mechanism, and simple data weighted superposition is easy to cause overfitting of shallow information and loss of deep signals; (2) traditional regularization constraint methods (such as smooth constraint, minimum model constraint) are difficult to adapt to complex geological structures, resulting in false anomalies or blurred boundaries in the inversion model; (3) the spatial correlation of physical property parameters under structured grids is insufficiently characterized, and existing gradient operator discretization methods are difficult to accurately depict anisotropic characteristics in hexahedral grids.

[0004] In addition, in the exploration of deep mineral resources, the magnetic differences between buried ore bodies and surrounding rocks often show non-linear variation characteristics, and the fixed constraint weight mechanism adopted by conventional inversion algorithms cannot dynamically adapt to the gradient change law of physical property parameters. Especially in complex structural areas such as fault-developed areas or contact zones, the existing methods have insufficient ability to jointly invert multi-scale magnetic bodies, seriously restricting the construction accuracy of 3D geological models. Therefore, developing a high-precision inversion method that can deeply fuse multi-source magnetic data and has an adaptive constraint mechanism has become a key technical problem for improving the exploration efficiency of mineral resources in complex areas. Summary of the Invention

[0005] The purpose of the present invention is to provide a comprehensive inversion method and system based on bidirectional susceptibility weighted constraint to solve the problems raised in the above background technique.

[0006] To achieve the above object, the present invention provides the following technical solution: A comprehensive inversion method based on bidirectional susceptibility weighted constraint, the method comprising:

[0007] Subdivide the underground three-dimensional space and establish two different cuboid models;

[0008] Based on the observation point coordinates and the cuboid model positions, determine the total magnetic field and its gradient anomalies observed on the ground;

[0009] Based on the total magnetic field and its gradient anomalies, establish a kernel function matrix;

[0010] Based on the kernel function matrix, determine an inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters;

[0011] Construct a bidirectional susceptibility weighted constraint function, add the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function;

[0012] Based on the conjugate gradient method, obtain the optimal solution for the comprehensive inversion objective function.

[0013] Preferably, the step of determining the inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters based on the kernel function matrix specifically includes:

[0014] Determine the relationship between the total magnetic field and its gradient anomalies and physical property parameters;

[0015] Based on the relationship between the total magnetic field and its gradient anomalies and physical property parameters, determine the inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters;

[0016] Introduce a regularization constraint term to update the inversion objective function.

[0017] Preferably, the step of constructing a bidirectional susceptibility weighted constraint function, adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function specifically includes:

[0018] Construct a bidirectional susceptibility weighted constraint function;

[0019] Transform the bidirectional susceptibility weighted constraint function;

[0020] Add the transformed bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function.

[0021] Preferably, in the step of establishing the kernel function matrix based on the total magnetic field and its gradient anomalies, the specific calculation formula of the kernel function matrix is as follows:

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] wherein, is the total magnetic field anomaly, is -direction magnetic gradient anomaly, is -direction magnetic gradient anomaly, is -direction magnetic gradient anomaly, is the permeability in vacuum, is the magnetization intensity vector modulus; , , ; , ; is the magnetic dip angle, is the magnetic declination, is the geomagnetic dip angle, is the angle between magnetic north and axis (true north direction).

[0027] Preferably, the relationship between the determined total magnetic field and its gradient anomaly and physical property parameters is expressed as:

[0028] ;

[0029] wherein, the column vector is the magnetic measurement data, including the total magnetic field anomaly , the magnetic gradient anomaly , and , its dimension is , the column vector is the physical property parameter, here representing the magnetic susceptibility parameter, its dimension is ; the matrix represents the kernel function matrix connecting the observed data and the physical property parameter , its dimension is .

[0030] Preferably, in the step of determining the inversion objective function reflecting the relationship between the total magnetic field and its gradient anomaly and physical property parameters based on the relationship between the total magnetic field and its gradient anomaly and physical property parameters, the inversion objective function is:

[0031] .

[0032] Preferably, the updated inversion objective function in the step of updating the inversion objective function by introducing the regularization constraint term is:

[0033] ;

[0034] where , is the magnetic susceptibility obtained by the joint inversion of the total magnetic field and its gradient, is the regularization coefficient, is the model weighting matrix of the joint inversion of the total magnetic field and its gradient.

[0035] Preferably, the two-way magnetic susceptibility weighted constraint function in the step of constructing the two-way magnetic susceptibility weighted constraint function is:

[0036] ;

[0037] where,

[0038] ;

[0039] ;

[0040] ;

[0041] represents the magnetic susceptibility result obtained by inverting the total magnetic field anomaly ( );

[0042] The specific form of the two-way magnetic susceptibility weighted constraint function in the comprehensive inversion objective function of the total magnetic field and its gradient based on the two-way magnetic susceptibility weighted constraint is as follows:

[0043] ;

[0044] where,

[0045] ;

[0046] ;

[0047] ;

[0048] , represents the magnetic susceptibility result obtained by the joint inversion of the magnetic gradient anomalies in three directions ( , , ).

[0049] Preferably, in the step of adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function, the comprehensive inversion objective function is:

[0050] ;

[0051] where is the susceptibility obtained by the comprehensive inversion of the total magnetic field and its gradient based on the bidirectional susceptibility weighted constraint, is the regularization coefficient, is the regularization factor of the bidirectional susceptibility weighted constraint function.

[0052] The present invention also provides a comprehensive inversion system based on bidirectional susceptibility weighted constraint for implementing a comprehensive inversion method based on bidirectional susceptibility weighted constraint. The system includes:

[0053] A meshing module for meshing the underground three-dimensional space and establishing two different cuboid models;

[0054] An observation module for determining the total magnetic field and its gradient anomalies observed on the ground based on the observation point coordinates and the cuboid model positions;

[0055] A matrix module for establishing a kernel function matrix based on the total magnetic field and its gradient anomalies;

[0056] An objective function determination module for determining an inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the kernel function matrix;

[0057] A weighted constraint module for constructing a bidirectional susceptibility weighted constraint function and adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function;

[0058] A solution module for obtaining the optimal solution of the comprehensive inversion objective function based on the conjugate gradient method.

[0059] Compared with the prior art, the beneficial effects of the present invention are as follows: By integrating the high-level resolution ability of magnetic gradient data and the deep exploration advantage of total magnetic field data, the present invention significantly improves the fine delineation of the boundaries of shallow anomalies and the reconstruction accuracy of deep field source structures, effectively solving the problem of unbalanced shallow and deep resolution in traditional inversion methods; The dynamic weighted constraint design can adaptively adjust the contribution weights of data at different depths, and achieve high-precision analysis of the spatial morphology of ore bodies in complex scenarios such as nappe structures and igneous rock cover; At the same time, based on the collaborative optimization of structured grids and gradient operators, the computational efficiency and stability of 3D inversion are significantly improved, providing technical support for large-scale exploration. In engineering applications, this method can accurately locate the spatial distribution of concealed ore bodies, reduce exploration costs, provide high-precision support for the delineation of deep mineral resource target areas and reserve assessment, and has important practical value and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0061] Figure 1 It is a flowchart of a comprehensive inversion method based on bidirectional susceptibility weighted constraints provided by an embodiment of the present invention.

[0062] Figure 2 It is an overall process diagram of the comprehensive inversion based on bidirectional susceptibility weighted constraints provided by an embodiment of the present invention.

[0063] Figure 3 It is a schematic diagram of the positions of observation points and a cuboid model provided by an embodiment of the present invention.

[0064] Figure 4 It is a flowchart of the steps for determining an inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters based on a kernel function matrix provided by an embodiment of the present invention.

[0065] Figure 5 It is a flowchart of the steps for constructing a bidirectional susceptibility weighted constraint function, adding the bidirectional susceptibility weighted constraint function to the inversion objective function, and generating a comprehensive inversion objective function provided by an embodiment of the present invention.

[0066] Figure 6 It is a map of the total magnetic field and its gradient anomalies provided by an embodiment of the present invention.

[0067] Figure 7 It is an inversion result map provided by an embodiment of the present invention.

[0068] Figure 8Block diagram of the composition structure of a comprehensive inversion system based on bidirectional magnetic susceptibility weighted constraint provided by an embodiment of the present invention. Detailed implementation manners

[0069] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0070] As Figures 1 to 7 shown, in an embodiment of the present invention, a comprehensive inversion method based on bidirectional magnetic susceptibility weighted constraint, the method includes steps S100 to S600:

[0071] Dissect the underground three-dimensional space and establish two different cuboid models;

[0072] Based on the observation point coordinates and the cuboid model positions, determine the total magnetic field observed on the ground and its gradient anomaly;

[0073] Based on the total magnetic field and its gradient anomaly, establish a kernel function matrix;

[0074] Based on the kernel function matrix, determine an inversion objective function reflecting the relationship between the total magnetic field and its gradient anomaly and the physical property parameters;

[0075] Construct a bidirectional magnetic susceptibility weighted constraint function, add the bidirectional magnetic susceptibility weighted constraint function to the inversion objective function, and generate a comprehensive inversion objective function;

[0076] Based on the conjugate gradient method, obtain the optimal solution for the comprehensive inversion objective function.

[0077] In this embodiment, in the step of dissecting the underground three-dimensional space and establishing two different cuboid models, specifically: dissect the underground three-dimensional space into M regularly sized cubic units with a background magnetic susceptibility of 0, and the sizes of the two cuboid models with a magnetic susceptibility difference of 0.25 SI are respectively and , and there are observation points on the ground. The coordinates of the observation point are , and the position of the cuboid unit is , where , , , specifically as Figure 3 shown;

[0078] Calculate the total magnetic field observed on the ground and its gradient anomaly. The total magnetic field anomaly is represented by , and the gradient anomaly refers to the total magnetic field anomaly in - direction, - direction, and the partial derivatives in the - direction, which are represented by , and respectively, as specifically shown in Figure 6 where Figure a is the total magnetic field anomaly ( ), Figure b is the magnetic gradient anomaly in the - direction , Figure c is the magnetic gradient anomaly in the - direction , and Figure d is the magnetic gradient anomaly in the - direction ;

[0079] The specific calculation formula of the kernel function matrix (including the total magnetic field anomaly , the magnetic gradient anomaly in the - direction , the magnetic gradient anomaly in the - direction and the magnetic gradient anomaly in the - direction ) is as follows:

[0080] ;

[0081] ; ;

[0082] ;

[0083] where the magnetic permeability in vacuum is , is the modulus of the magnetization intensity vector ; , , ; , ; is the magnetic dip angle, is the magnetic declination, is the geomagnetic dip angle, is the angle between magnetic north and the axis (true north direction).

[0084] As shown in Figure 4 , as a preferred embodiment of the present invention, the steps of determining the inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters based on the kernel function matrix specifically include:

[0085] Determine the relationship between the total magnetic field and its gradient anomalies and physical property parameters;

[0086] Determine the inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and physical property parameters based on the relationship between the total magnetic field and its gradient anomalies and physical property parameters;

[0087] Introduce a regularization constraint term to update the inversion objective function.

[0088] In this embodiment, the relationship between the total magnetic field and its gradient anomalies and physical property parameters is:

[0089] ;

[0090] Among them, the column vector is magnetic measurement data, including the total magnetic field anomaly , the magnetic gradient anomaly , and , its dimension is , the column vector is the physical property parameter, here representing the magnetic susceptibility parameter, and its dimension is ; the matrix represents the kernel function matrix connecting the observed data and the physical property parameter , and its dimension is , , .

[0091] Write the inversion problem represented by in the form of an objective function:

[0092] ;

[0093] After introducing the regularization constraint term, the joint inversion objective function of the total magnetic field and its gradient can be written as:

[0094] ;

[0095] Among them , is the magnetic susceptibility obtained by the joint inversion of the total magnetic field and its gradient. is the regularization coefficient, and its value is generally between 0 and . is the model weighting matrix for the joint inversion of the total magnetic field and its gradient.

[0096] By integrating the high horizontal resolution ability of magnetic gradient data with the deep exploration advantage of total magnetic field data, the fine delineation of the shallow anomaly body boundary and the reconstruction accuracy of the deep field source structure are significantly improved, effectively solving the problem of unbalanced shallow and deep resolution in traditional inversion methods. The introduction of the vector cross product regularization constraint mechanism greatly suppresses the non-uniqueness of inversion under complex geological conditions and enhances the reliability and geological rationality of the model results. The dynamic weighted constraint design can adaptively adjust the contribution weights of data at different depths, achieving high-precision analysis of the spatial morphology of ore bodies in complex scenarios such as nappe structures and igneous rock cover. At the same time, based on the collaborative optimization of structured grids and gradient operators, the computational efficiency and stability of 3D inversion are significantly improved, providing technical support for large-scale exploration. In engineering applications, this method can accurately locate the spatial distribution of concealed ore bodies, reduce exploration costs, and provide high-precision support for the delineation of deep mineral resource target areas and reserve assessment, with important practical value and promotion prospects.

[0097] The underground space is divided into many grid cells of the same size through a regular grid. The gradient constraint term is calculated using the joint inversion results of the total magnetic field anomaly inversion result and the magnetic gradient anomaly, and the constraint function is constructed and added to the joint inversion objective function of the total magnetic field and its gradient. For the construction of the self-structure constraint function under a structured hexahedron grid, the central difference can be used to replace the derivatives of physical property parameters in each direction, and the second norm of the cross product of the physical property parameter gradient vectors is added to the inversion objective function. By utilizing the higher horizontal resolution of magnetic gradient data and the better reflection of total magnetic field data for deep field source information, the information contained in different data is fully utilized, solving the problem of low inversion resolution, and being able to better depict the distribution of ferromagnetic minerals underground, providing technical support for mineral resource exploration in complex areas.

[0098] As Figure 5 shown, as a preferred embodiment of the present invention, the steps of constructing a two-way susceptibility weighted constraint function and adding the two-way susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function specifically include:

[0099] Construct a two-way susceptibility weighted constraint function;

[0100] Transform the two-way susceptibility weighted constraint function;

[0101] Add the transformed two-way susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function.

[0102] In this embodiment, the specific form of the two-way susceptibility weighted constraint function is as follows:

[0103] ;

[0104] where,

[0105] ;

[0106] ;

[0107] ;

[0108] Here, represents the magnetic susceptibility result obtained by inverting the total magnetic field anomaly ( ).

[0109] In the comprehensive inversion objective function of the total magnetic field and its gradient based on the two-way magnetic susceptibility weighted constraint, the specific form of the two-way magnetic susceptibility weighted constraint function is as follows:

[0110] ;

[0111] Among them,

[0112] ;

[0113] ;

[0114] ;

[0115] Here, , represents the magnetic susceptibility result obtained by jointly inverting the magnetic gradient anomalies in three directions ( , , ).

[0116] Before obtaining the optimal solution of the objective function, the form of the two-way magnetic susceptibility weighted constraint function needs to be transformed. First, perform the following transformation on in the two-way magnetic susceptibility weighted constraint function , , :

[0117] ;

[0118] ;

[0119] ;

[0120] Among them, , , are first-order difference operators, and the forward difference or backward difference is often used to replace the derivative at the boundary.

[0121] Then, for the two-way magnetic susceptibility weighted constraint function in , , perform the following transformations:

[0122] ; ;

[0123] ;

[0124] ;

[0125] Through the above transformations, the bidirectional susceptibility weighted constraint function and can be written in the following form:

[0126] ;

[0127] ;

[0128] Add the bidirectional susceptibility weighted constraint function to the magnetic total field and its gradient inversion objective function to obtain the comprehensive inversion objective function of the magnetic total field and its gradient based on bidirectional susceptibility weighted constraint:

[0129] ;

[0130] where is the susceptibility obtained by the comprehensive inversion of the magnetic total field and its gradient based on bidirectional susceptibility weighted constraint, is the regularization coefficient, is the regularization factor of the bidirectional susceptibility weighted constraint function, and its value in the inversion is . It can make the inversion result more focused and improve the ability to recover the magnetic structure of underground magnetic bodies. Its form is as follows:

[0131] ;

[0132] where is the number of observation points, is the number of regular cubic units for the three-dimensional spatial subdivision underground. , and are the inversion results of the magnetic total field anomaly and the joint inversion result of the magnetic gradient anomaly respectively. is the depth weighting parameter, usually taking a value between 1 and 2, and its value in the inversion is 1.1. means converting the vector in the parentheses into a diagonal matrix.

[0133] Comprehensive inversion objective function of the magnetic total field and its gradient based on bidirectional susceptibility weighted constraint Use the conjugate gradient method to obtain the optimal solution under the bidirectional susceptibility weighted constraint 。

[0134] The results are as Figure 7 shown in the inversion result diagrams. The black dashed boxes and the labels inside correspond to the geological bodies and their true positions. Diagram a is a slice at y = 1900 m of the three-dimensional result inverted from the total magnetic field anomaly ( ), diagram b is a slice at y = 1900 m of the three-dimensional result jointly inverted from the magnetic gradient anomalies ( , , ), diagram c is a slice at y = 1900 m of the three-dimensional result jointly inverted from the total magnetic field and its gradients ( , , , ), and diagram d is a slice at y = 1900 m of the three-dimensional result of the comprehensive inversion of the total magnetic field and its gradients based on the two-way magnetic susceptibility weighted constraint.

[0135] Comparative example: Under the same conditions, the magnetic susceptibility value restored by the comprehensive inversion of the total magnetic field and its gradients based on the two-way magnetic susceptibility weighted constraint is 2 times that of the joint inversion of the total magnetic field and its gradients, and the resolution is improved by 1 time.

[0136] The present invention also provides a comprehensive inversion system based on the two-way magnetic susceptibility weighted constraint for implementing a comprehensive inversion method based on the two-way magnetic susceptibility weighted constraint. The system includes:

[0137] A meshing module 100 for meshing the underground three-dimensional space and establishing two different cuboid models;

[0138] An observation module 200 for determining the total magnetic field and its gradient anomalies observed on the ground based on the observation point coordinates and the positions of the cuboid models;

[0139] A matrix module 300 for establishing a kernel function matrix based on the total magnetic field and its gradient anomalies;

[0140] An objective function determination module 400 for determining an inversion objective function reflecting the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the kernel function matrix;

[0141] A weighted constraint module 500 for constructing a two-way magnetic susceptibility weighted constraint function and adding the two-way magnetic susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function;

[0142] A solution module 600 for obtaining the optimal solution for the comprehensive inversion objective function based on the conjugate gradient method.

[0143] The foregoing are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A comprehensive inversion method based on bidirectional susceptibility weighted constraints, characterized in that, The method includes: Subdividing the underground three-dimensional space and establishing two different cuboid models; Determining the total magnetic field and its gradient anomalies observed on the ground based on the observation point coordinates and the positions of the cuboid models; Establishing a kernel function matrix based on the total magnetic field and its gradient anomalies; Determining an inversion objective function that reflects the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the kernel function matrix; Constructing a bidirectional susceptibility weighted constraint function, adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function; Finding the optimal solution for the comprehensive inversion objective function based on the conjugate gradient method.

2. The integrated inversion method based on bidirectional magnetic susceptibility weighted constraint according to claim 1, wherein, The steps of determining the inversion objective function that reflects the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the kernel function matrix specifically include: Determining the relationship between the total magnetic field and its gradient anomalies and the physical property parameters; Determining the inversion objective function that reflects the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the relationship between the total magnetic field and its gradient anomalies and the physical property parameters; Introducing a regularization constraint term to update the inversion objective function.

3. The integrated inversion method based on bidirectional susceptibility weighted constraint according to claim 1, wherein The steps of constructing a bidirectional susceptibility weighted constraint function, adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function specifically include: Constructing a bidirectional susceptibility weighted constraint function; Transforming the bidirectional susceptibility weighted constraint function; Adding the transformed bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function.

4. A comprehensive inversion method based on bidirectional magnetic susceptibility weighted constraint according to claim 1, characterized in that, The specific calculation formula of the kernel function matrix in the step of establishing the kernel function matrix based on the total magnetic field and its gradient anomalies is as follows: ； ； ； ； Among them, is the total magnetic field anomaly, is the magnetic gradient anomaly in the direction, is the magnetic gradient anomaly in the direction, is the magnetic gradient anomaly in the direction, is the magnetic permeability in vacuum, is the modulus of the magnetization intensity vector , , ; , ; is the magnetic dip angle, is the magnetic declination, is the geomagnetic dip angle, is the angle between magnetic north and the axis (true north direction).

5. The comprehensive inversion method based on bidirectional magnetic susceptibility weighted constraint according to claim 2, wherein, The determination of the relationship between the total magnetic field and its gradient anomalies and the physical property parameters is expressed as: ； Among them, the column vector Magnetic survey data, including total magnetic field anomalies , magnetic gradient anomaly 、 and , whose dimension is , column vector is a physical parameter, which here represents the magnetic susceptibility parameter, and its dimension is ;matrix Indicates connected observation data and physical properties The kernel function matrix has the dimension .

6. The integrated inversion method based on bidirectional magnetic susceptibility weighted constraint according to claim 5, characterized in that, The inversion objective function in the steps of determining the inversion objective function that reflects the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the relationship between the total magnetic field and its gradient anomalies and the physical property parameters is: 。 7. The comprehensive inversion method based on bidirectional susceptibility weighted constraint according to claim 6, wherein The updated inversion objective function in the step of introducing a regularization constraint term to update the inversion objective function is: ； wherein , is the magnetic susceptibility obtained by the joint inversion of the total magnetic field and its gradient, is the regularization coefficient, is the model weighting matrix for the joint inversion of the total magnetic field and its gradient.

8. A comprehensive inversion method based on bidirectional susceptibility-weighted constraints according to claim 3, characterized in that The bidirectional susceptibility weighted constraint function in the step of constructing the bidirectional susceptibility weighted constraint function is: ； Where, ； ； ； indicating the magnetic susceptibility result obtained by inverting the total magnetic field anomaly ( ); Bidirectional magnetic susceptibility weighted constraint function in the comprehensive inversion objective function of total magnetic field and its gradient based on bidirectional magnetic susceptibility weighted constraint The specific form is as follows: ； Where, ； ； ； , represents the magnetic susceptibility results obtained by the joint inversion of three-direction magnetic gradient anomalies ( , , ).

9. The integrated inversion method based on bidirectional magnetic susceptibility weighted constraint according to claim 8, wherein In the step of adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function, the comprehensive inversion objective function is: ； wherein, is the magnetic susceptibility obtained by comprehensive inversion of the total magnetic field and its gradient based on the bidirectional magnetic susceptibility weighted constraint, is the regularization coefficient, is the regularization factor of the bidirectional magnetic susceptibility weighted constraint function.

10. A comprehensive inversion system based on bidirectional magnetic susceptibility weighted constraint is used to implement the comprehensive inversion method based on bidirectional magnetic susceptibility weighted constraint according to any one of claims 1-9, and is characterized in that, The system includes: A subdivision module for subdividing the underground three-dimensional space and establishing two different cuboid models; An observation module for determining the total magnetic field and its gradient anomalies observed on the ground based on the observation point coordinates and the positions of the cuboid models; A matrix module for establishing a kernel function matrix based on the total magnetic field and its gradient anomalies; An objective function determination module for determining an inversion objective function that reflects the relationship between the total magnetic field and its gradient anomalies and the physical property parameters based on the kernel function matrix; A weighted constraint module for constructing a bidirectional susceptibility weighted constraint function, adding the bidirectional susceptibility weighted constraint function to the inversion objective function to generate a comprehensive inversion objective function; A solution module for finding the optimal solution for the comprehensive inversion objective function based on the conjugate gradient method.

Images