Gravity inversion method based on unstructured curved surface hexahedral mesh generation

Abstract

The invention discloses a gravity inversion method based on unstructured curved surface hexahedral mesh generation, and solves the problem that three-dimensional density structure characteristics in a complex terrain cannot be accurately calculated in the prior art. According to the method, a regular hexahedron is mapped into an unstructured curved surface hexahedron through isoparametric transformation, topographic relief is accurately fitted, forward modeling results of single curved surface hexahedron units are obtained, a global gravity field distribution model is generated through superposition, and three-dimensional gravity inversion is performed through a nonlinear conjugate gradient method. And acquiring a global optimal solution of the deep density parameter under the complex terrain, and finally testing effectiveness and accuracy through different models to acquire a clear and accurate high-precision geological underground structure model. Through the unstructured curved surface hexahedral mesh generation and nonlinear conjugate gradient inversion technology, the precision and efficiency of deep density parameter inversion under the complex terrain are remarkably improved, the accuracy of deep geophysical exploration is improved, and a reliable basis is provided for mineral resource exploration and geological structure analysis.

Description

Technical Field

[0001] The present invention relates to the field of geophysical exploration, and in particular to a gravity inversion method based on unstructured surface hexahedral grid partitioning. Background Art

[0002] Geophysical inversion is an important technical means in the field of resource exploration. However, complex terrain areas are prone to distortion of the geophysical field, resulting in inaccurate geophysical inversion. Traditional geophysical exploration methods often use regular hexahedral grids for forward modeling. This method cannot simulate high-precision undulating terrain and complex underground structures. Although unstructured tetrahedral grids can simulate terrain, the number of grids is difficult to control and the grid arrangement is irregular, which not only seriously affects the efficiency of the forward modeling process of the gravity method, but also greatly affects the validity of the inversion results. Therefore, rapid forward simulation of terrain with three-dimensional gravity suitable for inversion is a key factor affecting the efficiency and accuracy of the inversion process in complex terrain areas.

[0003] To address the above issues, we propose a gravity inversion method based on unstructured curved hexahedral grids to solve key problems in geophysical exploration, such as simulating undulating terrain, improving the efficiency of forward and inversion, and improving the reliability of results. This provides new technical support for deep underground structure detection and deep mineral resource evaluation. Summary of the Invention

[0004] The purpose of the present invention is to provide a gravity inversion method based on unstructured surface hexahedral meshing to solve the technical problem that the existing technology cannot accurately calculate the three-dimensional density structure characteristics under complex terrain.

[0005] To achieve the above objectives, the present invention provides the following technical solutions:

[0006] The present invention provides a gravity inversion method based on unstructured surface hexahedral meshing, comprising the following steps:

[0007] Step 1: Obtain gravity observation data of the target area, preprocess and calibrate the data to form standardized input data;

[0008] Step 2: Based on the terrain undulation characteristics of the surface, regular grid generation and unstructured surface hexahedron construction are carried out;

[0009] Step 3: Assign initial density parameters to each unstructured surface hexahedral element, calculate the gravity anomaly value of a single element based on potential field theory, superimpose the forward responses of all elements, and generate a global forward prediction model;

[0010] Step 4: Based on the global forward prediction model, the gravity observation data are inverted using the nonlinear conjugate gradient method;

[0011] Step 5: Perform inversion and solve the problem. During the iteration of the nonlinear conjugate gradient method, the density parameters are optimized synchronously to ensure that the final model meets the gravity fitting requirements. After the iteration converges, the optimal density parameter solution of each unit in the global unstructured surface hexahedral grid is extracted to construct a three-dimensional spatial physical property distribution model.

[0012] Step 6: Test the inversion results by setting models of different shapes and densities.

[0013] Furthermore, step 1 includes: generating a cross-shaped gravity observation point for the observation target, obtaining gravity observation data through ground gravimeters, aerial gravity or satellite gravity, which reflects the spatial distribution characteristics of the density of the underground medium, and performing data preprocessing, terrain correction, dimension correction, intermediate layer correction and height correction on the obtained raw data to eliminate the systematic deviation of the observation results caused by environmental interference and terrain undulations, and form standardized input data.

[0014] Furthermore, step 2 includes: adjusting the grid unit size according to the surface elevation data and the complexity of the geological structure, encrypting the grid units in the area with steep terrain changes, and performing sparse processing in the area with flat terrain; using the mesh partitioning method of unstructured surface hexahedron for the steep terrain and complex structure areas, and using regularized hexahedral mesh sections for the flat areas; converting the regular hexahedral rectangular units into unstructured surface hexahedral units through the mathematical mapping relationship of isoparametric transformation, and performing nonlinear mapping on the vertex coordinates of the regular hexahedron according to the actual spatial position of the terrain surface, so that the boundary surfaces of the surface hexahedron accurately fit the surface undulations; ensuring seamless connection between adjacent units and avoiding grid distortion by iteratively optimizing the vertex coordinates and connection relationship of the surface hexahedral units, and finally constructing an unstructured surface hexahedral three-dimensional mesh model that fits the actual terrain.

[0015] Furthermore, step 3 includes: for a single unstructured surface hexahedral unit, performing a forward modeling response of the gravity field, assigning an initial density value parameter to each surface hexahedral unit, and establishing a mapping relationship between the unit physical property parameters and the field response; based on the potential field theory, forming the gravity anomaly value generated by a single surface hexahedral unit at its corresponding spatial position, fully considering the influence of the unit geometric deformation on the field value; by superimposing the forward modeling response results of all unstructured surface hexahedral units, the gravity field distribution of the entire three-dimensional underground space is obtained, and a global forward modeling prediction model consistent with the spatial resolution of the observation data is formed.

[0016] Furthermore, step 4 includes: constructing an objective function, defining the objective function of the inversion problem, which includes gravity data fitting residuals and physical property parameter regularization constraints, and balancing data matching and model smoothness requirements through weight coefficients; optimizing iteration parameters, successively updating the density parameters of each surface hexahedral unit based on the nonlinear conjugate gradient method, using the linear combination of the current gradient direction and the historical search direction to determine the parameter update step size, gradually reducing the objective function value, and approaching the global optimal solution; adjusting the adaptive step size, dynamically adjusting the line search step size in combination with the Wolfe condition, to ensure the stable convergence of the inversion process and avoid falling into local extreme values.

[0017] The present invention provides an electronic device comprising a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory to implement the steps of the gravity inversion method based on unstructured surface hexahedral meshing.

[0018] The present invention provides a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the steps of the gravity inversion method based on unstructured surface hexahedral meshing.

[0019] Based on the above technical solution, the embodiments of the present invention can produce at least the following technical effects:

[0020] (1) The gravity inversion method based on unstructured surface hexahedron meshing provided by the present invention simulates terrain by using unstructured surface hexahedrons, uses the Gauss-Legendre integral method and the integral equation method for fast gravity forward modeling, and adopts the finite element method for gravity forward modeling to achieve a high-resolution and high-efficiency three-dimensional terrain gravity inversion method; through the unstructured surface hexahedron meshing and nonlinear conjugate gradient inversion technology, the accuracy and efficiency of deep density parameter inversion under complex terrain are significantly improved, the accuracy of deep geophysical exploration is improved, and a reliable basis is provided for mineral resource exploration and geological structure analysis.

[0021] (2) The gravity inversion method based on unstructured surface hexahedral grid partitioning provided by the present invention is upgraded from three aspects: simulating undulating terrain, constructing the physical coupling relationship between gravity and density, and the reliability of the inversion results. Finally, a three-dimensional terrain gravity data inversion method with fast calculation speed and small memory usage is developed. It is of great significance for improving the inversion theoretical research and actual production of three-dimensional density parameters, significantly improving the resolution and efficiency of gravity inversion under complex terrain, and providing reliable technical support for deep mineral resource positioning and structural analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.

[0023] Figure 1 This is a flow chart of the gravity inversion method of the present invention;

[0024] Figure 2 Schematic diagram of the isoparametric transformation from a regular hexahedron to an unstructured curved hexahedron. In the figure, (a) is a regular hexahedron, and (b) is a curved hexahedron.

[0025] Figure 3 Obtain gravity data for regular gridding and the red high-density forward model map;

[0026] Figure 4 This is the inversion result diagram of the unstructured surface hexahedron of the red high-density monomer model;

[0027] Figure 5 Forward model diagram for regular grid division of plains with small terrain relief and mountainous areas with large terrain relief;

[0028] Figure 6 To test the inversion results of hexahedron of unstructured surface in plain and mountainous areas;

[0029] Figure 7 The forward model diagram of the two-body model with high-density and low-density bodies divided into regular grids;

[0030] Figure 8 The inversion results of the unstructured surface hexahedron for testing the red high-density and blue low-density models are shown;

[0031] Figure 9 Forward model diagram of high-density model with arbitrary shape divided into regular grids;

[0032] Figure 10 This is the inversion result diagram of the unstructured surface hexahedron for testing the high-density model of arbitrary shape. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in this field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0034] The object of the present invention is achieved through the following technical solutions:

[0035] like Figure 1 As shown, a gravity inversion method based on unstructured surface hexahedral meshing includes the following steps:

[0036] Step 1: Obtain gravity observation data of the target area;

[0037] In order to obtain the density information of the target underground medium, standard grid data is developed, and cross-shaped gravity observation points are generated for the observation target. Gravity observation data are obtained through ground gravimeters, aerial gravity or satellite gravity. The data reflects the spatial distribution characteristics of the density of the underground medium. The obtained raw data is subjected to data preprocessing, terrain correction, dimension correction, intermediate layer correction and height correction to eliminate the systematic deviation of the observation results caused by environmental interference and terrain undulation, and form standardized input data.

[0038] Step 2: Construct an unstructured surface hexahedral mesh;

[0039] Regular grid generation and hexahedron construction of unstructured surfaces are carried out based on the terrain undulation characteristics of the surface: For target areas with large terrain undulations, especially areas with distorted gravity fields, the hexahedron grid generation method of unstructured surfaces is used to discretize and model the undulating terrain.

[0040] First, the grid subdivision design is carried out: the size and shape of the grid cells are dynamically adjusted according to the surface elevation data and the complexity of the geological structure. The size and shape of the grid cells of the unstructured surface hexahedron should be relatively dense. The grid cells should be encrypted in areas with steep terrain changes and sparse in areas with flat terrain to ensure that the grid subdivision matches the terrain undulation and the number of grid cells is minimized.

[0041] Secondly, an isoparametric transformation is performed: through the mathematical mapping relationship of isoparametric transformation, the regular hexahedron rectangular unit is converted into an unstructured surface hexahedron unit. The vertex coordinates of the regular hexahedron are nonlinearly mapped according to the actual spatial position of the terrain surface, so that the boundary surfaces of the surface hexahedron accurately fit the surface undulations.

[0042] It should be noted that regular hexahedral rectangular cells are converted into unstructured curved hexahedral cells through the mathematical mapping relationship of isoparametric transformation. Specifically, the vertex coordinates of the regular hexahedrons are nonlinearly mapped according to the actual spatial positions of the terrain surface, so that the boundary surfaces of the curved hexahedrons continuously conform to the surface morphology, while preserving the topological connectivity and geometric regularity of the hexahedral cells, achieving accurate geometric fitting of complex terrain. Unstructured curved hexahedral meshes can simulate undulating terrain with high precision, and the number of subdivided cells is small, with a regular and simple arrangement. The meshing method directly affects the accuracy and speed of gravity forward and inversion. Therefore, the meshing problem is the primary issue in solving the three-dimensional forward and inversion problems of gravity belt terrain.

[0043] The schematic diagram of isoparametric transformation of curved hexahedron is as follows Figure 2 As shown, the parent unit of the isoparametric unit is Figure 2 (a) shows a cube, and the curved hexahedron unit is as follows Figure 2 (b) The cube contains 27 nodes, and the constructed shape function is as follows:

[0044]

[0045] in:

[0046]

[0047] The expanded form of the shape function is as follows:

[0048]

[0049] The conversion from isoparametric elements to surface elements can be written as:

[0050]

[0051] The mapping relationship in different coordinate systems required for partial derivative calculation is expressed as follows:

[0052]

[0053] The Jacobian matrix J of the isoparametric transformation is expressed as:

[0054]

[0055] Let i, j, k be the unit vectors in the x, y, z directions of the Cartesian coordinate system. The isoparametric transformation can be rewritten as:

[0056]

[0057] By vector The volume element dv composed of is expressed as:

[0058]

[0059] The triple integral equivalent transformation expression based on isoparametric transformation is:

[0060]

[0061] Finally, an unstructured mesh is generated: by iteratively optimizing the vertex coordinates and connection relationships of the curved hexahedral units, seamless connection between adjacent units is ensured, mesh distortion is avoided, and ultimately an unstructured curved hexahedral three-dimensional mesh model that fits the actual terrain is constructed.

[0062] Step 3: Gravity forward response process;

[0063] For a single unstructured surface hexahedral unit, the forward response of the gravity field is performed separately to assign an initial density value parameter to each surface hexahedral unit, and a mapping relationship between the unit physical parameters and the field response is established; based on the potential field theory, the gravity anomaly value generated by a single surface hexahedral unit at its corresponding spatial position is formed, and the influence of the unit's geometric deformation on the field value is fully considered; by superimposing the forward response results of all unstructured surface hexahedral units, the gravity field distribution of the entire three-dimensional underground space is obtained, forming a global forward prediction model consistent with the spatial resolution of the observation data.

[0064] It should be noted that 3D gravity forward numerical simulation involves dividing the 3D underground space with terrain into multiple curved hexahedron elements. By forward modeling a single curved hexahedron element, the forward modeling results for all curved hexahedrons in the 3D underground space are ultimately obtained. Therefore, the core of 3D gravity forward numerical simulation is the forward modeling of curved hexahedron elements.

[0065] Specifically, in the Cartesian coordinate system, let a point O in free space be the origin of the coordinate system, the Z axis is vertically downward, there is a unit with volume v, density ρ, and force intensity M, and the gravitational potential V and force potential U generated at any point p in the external space are:

[0066]

[0067] Among them, r, r′ are the space vectors of the observation point p and the abnormal point Q respectively, and G is the universal gravitational constant.

[0068] By taking the first-order derivative of the gravitational potential V and the force potential U and combining them with formula (9), we can obtain the forward calculation expression of the isoparametric unit gravity of an arbitrary curved hexahedron:

[0069]

[0070] Where μ0 is the vacuum conductivity. The Gauss-Legendre integral method is used to numerically solve Equations (12) and (13), and the forward gravity response of an arbitrary curved hexahedron is finally obtained.

[0071] Step 4: nonlinear conjugate gradient method inversion;

[0072] Based on the global forward prediction model, the gravity observation data are inverted using the nonlinear conjugate gradient method. The objective function of the inversion problem is constructed and defined, which includes the gravity data fitting residual and the regularization constraint terms of the physical parameters. The data matching and model smoothness requirements are balanced through the weight coefficient. The iteration parameters are optimized and the density parameters of each surface hexahedral unit are updated successively based on the nonlinear conjugate gradient method. The parameter update step size is determined by the linear combination of the current gradient direction and the historical search direction, and the objective function value is gradually reduced to approach the global optimal solution. The adaptive step size is adjusted and the line search step size is dynamically adjusted in combination with the Wolfe condition to ensure the stable convergence of the inversion process and avoid falling into local extreme values.

[0073] Step 5: Output the inversion results;

[0074] An inversion solution is performed, optimizing density parameters during the iterative process of the nonlinear conjugate gradient method to ensure that the final model meets the gravity fitting requirements. After iterative convergence, the optimal density parameter solution for each cell in the global unstructured surface hexahedral grid is extracted to construct a three-dimensional spatial physical property distribution model. This model accurately depicts the density interface and geological structural boundaries of deep media under the influence of complex terrain, providing a high-resolution physical property structure basis for mineral resource exploration and geological hazard assessment.

[0075] Step 6: Test the inversion validity and accuracy;

[0076] By setting models of different shapes and densities, the inversion results are tested. A regular single high-density body, a high-density and low-density dual-body model, and a high-density model body of arbitrary shape are set to test this method, proving its effectiveness and accuracy.

[0077] Example:

[0078] 1. Obtain gridded gravity and magnetic observation data for the target area. Obtain gravity and magnetic measurement point data in a cross-grid pattern, evenly distributed in the target area. This gridded measurement point pattern can effectively reflect gravity data representing density differences in the underground medium and magnetic data representing the distribution of magnetic parameters.

[0079] 2. Carry out unstructured surface hexahedron fitting of terrain undulations: For steep terrain and complex structural areas, the unstructured surface hexahedron meshing method is used, and for flat areas, the regularized hexahedron mesh section is used. This ensures that the grid distribution is highly consistent with the terrain characteristics and reduces the workload of meshing. The regular hexahedron unit is projected into an unstructured surface hexahedron through isoparametric transformation. The schematic diagram of the isoparametric transformation of the surface hexahedron is shown as follows: Figure 2 As shown, the parent unit of the isoparametric unit is Figure 2 (a) shows a cube, and the curved hexahedron unit is as follows Figure 2 (b), the cube contains 27 nodes. The terrain undulation is obtained by isoparametric transformation, so that the top surface of the unit fits the surface undulation accurately, such as Figure 3 As shown, an unstructured three-dimensional grid model matching the complex terrain is formed, and the following is obtained: Figure 4 The unstructured hexahedron shown fits the complex terrain map.

[0080] 3. To simulate the abnormal distortion effect under complex terrain, the high-density anomaly body was moved to the plain area. No density anomaly body was set in the mountainous area. The gravity response of the high-density anomaly was forward calculated, and the following results were obtained: Figure 5 The gravity response shown in the figure is obtained by inverting the unstructured conjugate gradient. The results show that the plain area has high density, and the mountain area has no density anomaly. This shows that the inversion of the unstructured surface hexahedral grid section is correct and can eliminate the potential field distortion caused by the terrain. Figure 6 The three-dimensional spatial distribution results of high-density and low-density anomalies.

[0081] 4. In order to further illustrate the correctness of the calculation results of this method, simulations are carried out for high-density anomalies and low-density anomalies under complex terrain, such as Figure 7 As shown in Figure 2, high-density anomalies and low-density anomalies are set up simultaneously in complex terrain. The inversion of the unstructured surface hexahedron is carried out, as shown in Figure 2. Figure 8 As shown in the figure, the results show high-density anomalies and low-density anomalies in the mountainous area, which shows that the inversion is correct and can effectively identify high-density and low-density anomalies.

[0082] 6. To further demonstrate the application of this method in reality, Figure 9 As shown in the figure, a simulation similar to the actual underground geological conditions was carried out, and a high-density anomaly nonlinear conjugate gradient inversion of arbitrary shapes was carried out. High-density anomalies of arbitrary shapes were set under complex terrain, such as Figure 10 As shown in the figure, the inversion results show that the mountainous area has high-density anomalies with arbitrary shapes that are similar to the actual underground geology, indicating that the three-dimensional terrain gravity inversion method can effectively identify high-density anomalies with arbitrary shapes.

[0083] 7. Pass Figure 3-10The results show that the gravity-magnetic joint inversion method based on unstructured surface hexahedral grid partitioning is correct and effective for regular high-density anomalies, regular high-density and low-density anomalies, and high-density anomalies of arbitrary shapes. It can be widely used in the field of geophysical exploration, especially in the field of gravity and magnetic exploration. It can solve the technical problems of three-dimensional density and magnetic parameter structure characteristics under complex terrain, and provide a reliable basis for mineral resource exploration and geological structure analysis.

[0084] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A gravity inversion method based on hexahedral meshing of unstructured surfaces, characterized in that: The following steps are involved: Step 1: Obtain gravity observation data of the target area, preprocess and calibrate the data to form standardized input data; Step 2: Based on the terrain undulation characteristics of the surface, regular grid generation and unstructured surface hexahedron construction are carried out; Step 3: Assign initial density parameters to each unstructured surface hexahedral element, calculate the gravity anomaly value of a single element based on potential field theory, superimpose the forward responses of all elements, and generate a global forward prediction model; Step 4: Based on the global forward prediction model, the gravity observation data are inverted using the nonlinear conjugate gradient method; Step 5: Perform inversion and solve the problem. During the iteration of the nonlinear conjugate gradient method, the density parameters are optimized synchronously to ensure that the final model meets the gravity fitting requirements. After the iteration converges, the optimal density parameter solution of each unit in the global unstructured surface hexahedral grid is extracted to construct a three-dimensional spatial physical property distribution model. Step 6: Test the inversion results by setting models of different shapes and densities.

2. The gravity inversion method based on unstructured surface hexahedral meshing according to claim 1, characterized in that: The step 1 comprises: A cross-shaped gravity observation point is generated for the observation target, and gravity observation data is obtained through ground gravimeters, aerial gravity or satellite gravity. The data reflects the spatial distribution characteristics of the density of the underground medium. The obtained raw data is subjected to data preprocessing, terrain correction, dimension correction, intermediate layer correction and height correction to eliminate the systematic deviation of the observation results caused by environmental interference and terrain undulation, and form standardized input data.

3. The gravity inversion method based on unstructured surface hexahedral meshing according to claim 1, characterized in that: The step 2 includes: Adjust the grid cell size based on surface elevation data and geological structure complexity, densify the grid cells in areas with steep terrain changes, and perform sparse processing in areas with gentle terrain; For steep terrain and complex structural areas, the unstructured surface hexahedron meshing method is used, and for flat areas, the regularized hexahedron mesh section is used; Through the mathematical mapping relationship of isoparametric transformation, regular hexahedral rectangular units are converted into unstructured surface hexahedral units. The vertex coordinates of the regular hexahedron are nonlinearly mapped according to the actual spatial position of the terrain surface, so that the boundary surfaces of the surface hexahedron accurately fit the surface undulations. By iteratively optimizing the vertex coordinates and connection relationships of the curved hexahedral units, seamless connection between adjacent units is ensured and mesh distortion is avoided. Ultimately, an unstructured curved hexahedral three-dimensional mesh model that fits the actual terrain is constructed.

4. The gravity inversion method based on unstructured surface hexahedral meshing according to claim 1, characterized in that: The step 3 comprises: For a single unstructured surface hexahedron element, the forward response of the gravity field is performed to assign initial density value parameters to each surface hexahedron element, and the mapping relationship between the element physical parameters and the field response is established; Based on the potential field theory, the gravity anomaly value generated by a single curved hexahedral unit at its corresponding spatial position is formed, fully considering the impact of the unit's geometric deformation on the field value; By superimposing the forward response results of all unstructured surface hexahedral elements, the gravity field distribution of the entire three-dimensional underground space is obtained, forming a global forward prediction model consistent with the spatial resolution of the observation data.

5. The gravity inversion method based on unstructured surface hexahedral meshing according to claim 1, characterized in that: The step 4 comprises: Construct an objective function and define the objective function of the inversion problem, which includes the gravity data fitting residual and physical property parameter regularization constraints, and balances the data matching and model smoothness requirements through weight coefficients; Optimize the iterative parameters and update the density parameters of each surface hexahedral unit based on the nonlinear conjugate gradient method. Use the linear combination of the current gradient direction and the historical search direction to determine the parameter update step size, gradually reduce the objective function value, and approach the global optimal solution. Adjust the adaptive step size and dynamically adjust the line search step size in combination with the Wolfe condition to ensure the stable convergence of the inversion process and avoid falling into local extreme values.

6. An electronic device comprising a processor and a memory, characterized in that: The memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory to implement the steps of the gravity inversion method based on unstructured surface hexahedral grid partitioning as described in any one of claims 1 to 5.

7. A computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the steps of the gravity inversion method based on unstructured surface hexahedral meshing according to any one of claims 1 to 5.

Images