A high-precision forward modeling method and equipment for the spatial domain of magnetic field effects of spherical prisms
By adaptively selecting Gaussian nodes and using an equal-interval subdivision method with the longest side length, the problem of excessive subdivision in the forward modeling of magnetic source magnetic effects in spherical coordinates is solved, achieving efficient and accurate calculation of magnetic field effects.
Patent Information
- Application Number
- CN202410408352.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-07
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2044-04-07
AI Technical Summary
When performing forward modeling of magnetic source magnetic effects in spherical coordinates, existing recursive and "stack" subdivision methods may lead to over-subdivision or stack overflow, and fixed Gaussian nodes may not be the best choice under different circumstances, affecting computational efficiency and accuracy.
An adaptive selection of Gaussian nodes (1-1-1 or 2-2-2) is adopted. The total number of subdivided spherical prisms is determined by a single subdivision. Based on the method of equal interval subdivision according to the longest side length, subdivision is only performed when necessary, ensuring that the average DSR of all subdivided blocks is greater than the required accuracy, thereby reducing computation time.
While meeting accuracy requirements, it significantly improves computational efficiency, reduces computation time, balances forward modeling errors, and avoids the computational burden caused by excessive subdivision.
Smart Images

Figure CN118191956B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of forward modeling of magnetic field effects, and more particularly to a high-precision forward modeling method and equipment for the spatial domain of magnetic field effects on a spherical prism. Background Technology
[0002] Forward modeling of the magnetic effects of magnetic sources (i.e., magnetic potential, magnetic vector, and magnetic gradient tensor) is one of the most fundamental problems in the geophysical and geological interpretation of lithospheric magnetic anomalies. Traditionally, this problem has been solved in Cartesian coordinates, where the geometry of the magnetic source and the wavelength of the target magnetic anomaly are relatively small, thus the Earth's magnetolithospheric region and the observation surface are treated as flat. However, the modern CHAMP and Swarm satellite constellations have provided unprecedentedly high-precision and nearly global-coverage magnetic measurement data. This has enabled quantitative geophysical and geological interpretations of regional and global lithospheric magnetic anomalies to be predominantly based on medium- and long-wavelength data. In this context, the magnetic sources and anomalies are of a larger scale; therefore, when performing actual forward modeling of the magnetic effects of the magnetic sources in spherical coordinates (SCS), the curvature of the Earth and the observation surface must be considered.
[0003] In SCS (Structured Convolutional Modeling), spherical prisms are the most logical choice as fundamental elements for discretizing complex sources. Furthermore, spherical prisms have been applied to property inversion, interface inversion, and lithosphere magnetic field modeling. However, there is no analytical formula for the magnetic effects of each prism. Therefore, several approximate solutions have been developed in SCS, such as magnetic dipoles, rectangular prisms, Taylor series expansions (TSE), and GLQ numerical integration, with GLQ being the most mature. GLQ outperforms 3D / 2D TSE and Newton-Cotes quadrature methods in terms of accuracy and computation time, making it more commonly used. Studies show that as long as the node spacing is less than the distance to the observation point, the GLQ solution remains essentially invariant for different numbers of Gaussian nodes. However, when the source geometry is large or the distance between the source and the observation point is short, large Gaussian nodes are required, resulting in a very heavy computational burden. GLQ integration with appropriate subdivision and reasonably high Gaussian nodes is currently the best forward modeling method.
[0004] The accuracy of the GLQ numerical integral forward modeling method is related to two factors: the Gaussian nodes and the ratio of observation distance to prism size (DSR). Larger Gaussian nodes result in higher accuracy, but also longer computation time. A higher DSR value also improves accuracy. Using fine subdivision to divide the volume into more blocks to increase the DSR value and improve accuracy is widely used in gravity forward modeling calculations, including 3D Adaptive Subdivision (3DAD) and 2D Adaptive Subdivision (2DAD). 2DAD only performs fine subdivision in two horizontal dimensions. Both 2DAD and 3DAD techniques employ recursive and "stacked" subdivision methods. Recursive methods require a manually defined minimum side length to prevent over-subdivision, while "stacked" methods may suffer from stack overflow. Both methods use fixed Gaussian nodes; however, fixed Gaussian nodes are not always the optimal choice. Summary of the Invention
[0005] In view of this, the purpose of the present invention is to provide a high-precision forward modeling method and device for the spatial domain of magnetic field effects of spherical prisms, which can solve the technical problem that recursive and "stacked" fine subdivision methods may produce excessive subdivision or stack overflow.
[0006] This invention provides a high-precision forward modeling method for the spatial domain of magnetic field effects on a spherical prism, comprising:
[0007] S1: Obtain the initial spherical prism. Given the computation time of a single spherical prism with different Gaussian nodes in the initial spherical prism, calculate the total number of subdivided spherical prisms with different Gaussian nodes in the initial spherical prism.
[0008] S2: Multiply the total number of subdivided spherical prisms of each Gaussian node by the calculation time of the corresponding single spherical prism to obtain the forward modeling prediction time of each Gaussian node, and select the Gaussian node with the shortest forward modeling prediction time as the node to perform forward modeling calculation.
[0009] S3: If the total number of subdivided spherical prisms of the node to be performed in forward modeling is greater than 1, then the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks. Forward modeling is performed on each spherical prism sub-block and accumulated to obtain the forward modeling simulation result; otherwise, proceed to step S4.
[0010] S4: Perform forward modeling calculations on the initial spherical prism to obtain the forward modeling simulation results.
[0011] Preferably, the Gaussian nodes are selected as i = 1-1-1 or 2-2-2, where i represents the number of the Gaussian node.
[0012] Preferably, step S1 specifically includes:
[0013] S11: Given an initial spherical prism with Gaussian node i, the computation time for a single spherical prism is t. i , where i represents the number of the Gaussian node;
[0014] S12: Obtain the number of subdivision blocks of Gaussian node i in each direction in the initial spherical prism. Where u is the direction number;
[0015] S13: Multiply the number of subdivided blocks of Gaussian node i in each direction to obtain the total number of subdivided spherical prisms of Gaussian node i, Num. i .
[0016] Preferably, the expression for the total number of subdivided spherical prisms is:
[0017]
[0018]
[0019]
[0020] Where u represents direction. r indicates the direction pointing towards the Earth's center. λ and λ represent the latitude and longitude directions respectively, floor() represents an integer less than the calculated result, d represents the minimum Euclidean distance between the geometric center of the spherical prism and all observation points, and D R L represents the ratio of the observation distance to the size of the spherical prism under the required calculation accuracy. u This represents the side length of the original spherical prism in the direction u.
[0021] The preferred expression for the forward modeling prediction time is:
[0022] Te i =t i ×Num i
[0023] Where i represents the Gaussian node number, Te i Let t represent the forward prediction time of Gaussian node i. i Num represents the computation time for a single spherical prism at Gaussian node i. i This represents the total number of subdivided spherical prisms at Gaussian node i.
[0024] Preferably, in step S3, the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks, specifically as follows:
[0025] After selecting the Gaussian node i of the initial spherical prism, in the direction r, according to the number of subdivision blocks... Divide into equal parts; in the direction Above, according to the number of subdivided blocks Divide into equal parts; along direction λ, according to the number of subdivided blocks. Divide into equal parts; after dividing into three directions, multiple spherical prism sub-blocks are obtained.
[0026] A storage medium storing instructions and data for implementing the high-precision forward modeling method for the spatial domain of the magnetic field effect of a spherical prism.
[0027] A high-precision forward modeling device for the spatial domain of magnetic field effects of a spherical prism includes: a processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement the high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism.
[0028] The present invention has the following beneficial effects:
[0029] The Gaussian nodes (1-1-1 or 2-2-2) are adaptively selected based on the number of subdivided blocks, instead of fixed Gaussian nodes, thereby maximizing computational efficiency while meeting accuracy requirements. After subdividing the initial spherical prism only once based on the total number of subdivided spherical prisms, the average DSR of all subdivided sub-blocks can be less than the DSR required to meet accuracy requirements, greatly reducing the associated computation time and balancing the forward modeling error of each observation point, keeping it below the threshold. Attached Figure Description
[0030] Figure 1 This is a flowchart of a method according to an embodiment of the present invention;
[0031] Figure 2 This is a schematic diagram of a spherical prism;
[0032] Figure 3 A 3D schematic diagram of multiple spherical prism sub-blocks;
[0033] Figure 4 A schematic diagram illustrating magnetization discontinuity and polar singularity;
[0034] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0035] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0036] Reference Figure 1 This application provides a high-precision forward modeling method for the spatial domain of magnetic field effects of spherical prisms. It adopts the GLQ numerical integration forward modeling method and uses a new method based on equal interval subdivision of the longest side length. After only one subdivision, the average DSR of all subdivided blocks can be less than the DSR required to meet the accuracy.
[0037] include:
[0038] S1: Obtain the initial spherical prism. Given the computation time of a single spherical prism with different Gaussian nodes in the initial spherical prism, calculate the total number of subdivided spherical prisms with different Gaussian nodes in the initial spherical prism.
[0039] S2: Multiply the total number of subdivided spherical prisms of each Gaussian node by the calculation time of the corresponding single spherical prism to obtain the forward modeling prediction time of each Gaussian node, and select the Gaussian node with the shortest forward modeling prediction time as the node to perform forward modeling calculation.
[0040] S3: If the total number of subdivided spherical prisms of the node to be performed in forward modeling is greater than 1, then the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks. Forward modeling is performed on each spherical prism sub-block and accumulated to obtain the forward modeling simulation result; otherwise, proceed to step S4.
[0041] S4: Perform forward modeling calculations on the initial spherical prism to obtain the forward modeling simulation results.
[0042] Specifically, the high-precision forward modeling method for the magnetic field effect of spherical prisms in this application is based on the fundamental principles of magnetic effect forward modeling and Gauss-Legend orthogonal integrals.
[0043] Forward modeling of magnetic effects:
[0044] like Figure 2 As shown, the tesseroid prism consists of six faces: two concentric spherical faces (r1, r2) at the top and bottom, two meridian faces (λ1, λ2) at the left and right, and two latitudinal sector faces at the front and back. In the local northward reference frame (x) on the sphere P ,y P ,z P The magnetic potential V and magnetic vector B at observation point P in the diagram are... i (i = x, y, z) and magnetic gradient tensor B ij The forward modeling formula for Tesseroid bodies (j = x, y, z) is as follows:
[0045]
[0046]
[0047]
[0048] In the formula, μ0 is the permeability in vacuum, and f V , and Kernel functions for the magnetic potential, magnetic vector, and magnetic gradient tensors, respectively:
[0049]
[0050]
[0051]
[0052] in, δ ij δ ik and δ jk (k = x, y, z) represents the Kronecker function; for example, if i = j, then it is δ. ij =1, otherwise δ ij =0. l represents the Gaussian node from the field source. To the observation point The Cartesian distance is as follows:
[0053]
[0054]
[0055]
[0056] x z =r-r'cosψ (6)
[0057] Wherein, the spherical angular distance ψ between the position vectors P(r) and Q(r') is determined by... The magnetization vector M is obtained. p In the local reference frame (x) P ,y P ,z P This is represented in ().
[0058] In geomagnetism, the magnetization vector M Q This usually indicates that the integral point is calculated. The local northward reference frame centered on (x) Q ,y Q ,z Q Therefore, before using equation (2) to calculate the magnetic effect, M Q It should be in a local northward reference frame centered on the observation (x P ,y P ,z P Converted to M P The transformation can be achieved through M. P =RM Q The rotation matrix R is calculated as follows:
[0059]
[0060] Where β = λ' - λ.
[0061] Gauss-Legend orthogonal integral:
[0062] According to the GLQ method, the integral in equation (1) can be written as
[0063]
[0064] in, Let n be the Gauss-Legendal nodal coefficients. r th、 With n λ Gaussian nodes of th The function; L, M, and N are the maximum Gaussian node numbers in the radial, latitudinal, and meridional directions, respectively; before practical application, it is necessary to calculate the number of nodes in the interval [-1, +1]. Transformed into node positions on the actual integration interval
[0065]
[0066]
[0067]
[0068] The GLQ method ensures that the distance between nodes and observations is always greater than zero. Therefore, there are no singularities even if the observation is located on the upper surface of the spherical prism. This shows that the GLQ method is theoretically not limited by the height of the observation.
[0069] Furthermore, the Gaussian nodes are selected as i = 1-1-1 or 2-2-2, where i represents the number of the Gaussian node.
[0070] Specifically, Gaussian nodes (1-1-1 or 2-2-2) are adaptively selected based on the number of subdivided blocks, rather than fixed Gaussian nodes, thereby maximizing computational efficiency while meeting accuracy requirements.
[0071] Furthermore, dividing the Tesseroid volume into more blocks using a finer subdivision method can increase the DSR value and improve forward modeling accuracy. DSR is the distance-to-size ratio, calculated using the following formula:
[0072]
[0073] Where, distance d is defined as the minimum Euclidean distance between the geometric center of the spherical prism and all observation points, and the side length L is calculated using the following formula:
[0074] L r =r2-r1 (11a)
[0075]
[0076]
[0077] When the following formula is satisfied
[0078]
[0079] The initial Tesseroid body does not need to be subdivided, where D R It is the required DSR corresponding to a certain level of accuracy.
[0080] D R The larger the value, the higher the accuracy of the forward modeling calculation, which also means more subdivisions of the blocks. When the DSR of a single block is greater than or equal to D... R When the required computational accuracy is achieved, then when the average DSR is greater than D R At this point, the forward modeling error of the subdivided Tesseroid bodies will reach the accuracy requirement. This condition is used as the termination criterion adopted in this invention, rather than requiring that the DSR of each subdivided Tesseroid body be greater than or equal to D. R .
[0081] As can be seen from equation (12), the length of the longest side determines whether further subdivision is needed. Therefore, d / D R It can be used as a basis for the number of subdivisions in each direction. In this way, the amount of calculation is simplified, the infinite subdivision that may be caused by further subdivision is avoided, and the required calculation accuracy is achieved.
[0082] Step S1 is as follows:
[0083] S11: Given an initial spherical prism with Gaussian node i, the computation time for a single spherical prism is t. i , where i represents the number of the Gaussian node;
[0084] S12: Obtain the number of subdivision blocks of Gaussian node i in each direction in the initial spherical prism. Where u is the direction number;
[0085] S13: Multiply the number of subdivided blocks of Gaussian node i in each direction to obtain the total number of subdivided spherical prisms of Gaussian node i, Num. i .
[0086] Furthermore, the expression for the total number of subdivided spherical prisms is:
[0087] side = d / D R (13a)
[0088]
[0089]
[0090] Where u represents direction. r indicates the direction pointing towards the Earth's center. λ and λ represent the latitude and longitude directions respectively, floor() represents an integer less than the calculated result, d represents the minimum Euclidean distance between the geometric center of the spherical prism and all observation points, and D R L represents the ratio of the observation distance to the size of the spherical prism under the required calculation accuracy. u This represents the side length of the original spherical prism in the direction u.
[0091] Furthermore, with the same Gaussian nodes, the forward modeling time should be proportional to the number of subdivided blocks. Under this assumption, the forward modeling prediction time for different Gaussian nodes can be estimated using the following formula;
[0092] The expression for the forward modeling prediction time is:
[0093] Te i =t i ×Num i (14)
[0094] Where i represents the Gaussian node number, Te i Let t represent the forward prediction time of Gaussian node i. i Num represents the computation time for a single spherical prism at Gaussian node i. i This represents the total number of subdivided spherical prisms at Gaussian node i.
[0095] Furthermore, in step S3, the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks, specifically as follows:
[0096] After selecting the Gaussian node i of the initial spherical prism, in the direction r, according to the number of subdivision blocks... Divide into equal parts; in the direction Above, according to the number of subdivided blocks Divide into equal parts; along direction λ, according to the number of subdivided blocks. Divide into equal parts; after dividing into three directions, multiple spherical prism sub-blocks are obtained.
[0097] Specifically, as can be seen from the formula, the average DSR of all subdivided blocks must be greater than or equal to D. R According to equation (13), for latitude ( Divide the direction (direction), longitude (λ direction), and geocenter (r direction) into equal parts to obtain the following: Figure 3 A 3D schematic diagram of multiple spherical prism sub-blocks is shown.
[0098] Furthermore, in order to obtain D through model experiments R A method for estimating overall numerical accuracy, namely the l2 norm relative error, was adopted:
[0099]
[0100] Where v represents the approximate theoretical value. This represents the forward modeling result of the GLQ method.
[0101] Simulated data demonstration
[0102] To verify the effectiveness of the technology, theoretical models were designed as shown in Table 1. Magnetic inclination I and magnetic declination D were set to 90° and 0°, respectively, and the Earth's average radius of 6371.2 km was used as the Earth's surface. Models A and B have side length L. r , L λ The values are 100m, 111.197m, 111.97m and 200m, 111.196m, 89.959m, respectively. A recursive subdivision method was used for comparison. This method first subdivides the prism into 2×2×2 blocks in three directions, with the first subdivision numbering being 2, 2, 2 and 2, 1, 1 respectively. The next subdivision uses a bisection method to divide the prism into 2×2×2 blocks. When the average DSR of all blocks is greater than D... R When the subdivision stops, the same stopping criteria as in this invention are applied.
[0103] Table 1 Model Parameters
[0104]
[0105] The number of subdivided blocks, the average DSR of all blocks, and the relative error were compared between the two methods, as shown in Table 2. B was calculated under different DSR conditions. z The components, different DSRs represent different observation heights. Given a calculation error of 0.01% and using a 1-1-1 Gaussian node, in this case, D... R The value is 5. It can be seen that the number of blocks in this invention is always less than or equal to the number of blocks in the recursive subdivision, sometimes even only 1 / 3, which means that the computational efficiency is significantly improved by 68.75%. The smaller the DSR, the lower the observation height, and the more obvious the difference. Although the number of subdivided blocks in this invention is much smaller, the average DSR of all subdivided blocks is still greater than or equal to D. R Furthermore, the relative error is less than the given error. This result demonstrates that the equidistant subdivision method based on the longest side length in this invention significantly reduces computation time while maintaining accuracy. Especially in practical applications, the computational efficiency will be further improved as the number of prisms and computation points increases.
[0106] Table 2 Comparison of two adaptive refinement methods
[0107]
[0108] Furthermore, Runcorn's theorem was used for verification. Runcorn's theorem states that a spherical shell with constant thickness and magnetic susceptibility, magnetized by any internal induced field, will not generate an external magnetic field. Here, it was used to test the efficiency of the proposed algorithm for forward modeling the magnetic effect of a spherical prism. A uniform spherical shell with a radius of 6371.2 km, a thickness of 26.5 km, and a magnetic susceptibility of 0.01 SI was used.
[0109] To demonstrate the computational efficiency of automatically selecting the optimal Gaussian nodes (1 or 2), two examples shown in Table 3 were calculated. The spherical shell was divided into spherical prisms of 0.5°×0.5° and 1°×1°, with computational heights set to 200km and 400km, respectively. The observation data space (latitude × longitude) was 1°×1° for both, with the same computational error (0.01%). Based on the subdivision strategy of this invention, B was calculated using three methods: fixed 1, fixed 2, and mixed Gaussian nodes. z Components. As latitude increases, the side length of the spherical prism in the longitude direction will gradually decrease, and the number of subdivided blocks will also be different. However, the hybrid method can always select the optimal Gaussian node with high computational efficiency according to equation (15). In general, Gaussian node 1 is used for high-latitude regions, while Gaussian node 2 is used for spherical prisms in other regions.
[0110] Due to magnetization discontinuity and polar singularity (e.g.) Figure 4 As shown in Table 3, the calculated model field differs slightly from the theoretical zero value. Table 3 shows that the computational accuracy of the three methods is consistent, with the maximum value and root mean square error on the same order of magnitude. A comparison reveals that the optimal Gaussian node method has the shortest computation time, reducing time by 8.56% and 24.07% for 0.5°×0.5°, and by 10.53% and 22.49% for 1°×1°. Based on the above analysis, it can be concluded that the selection of the optimal Gaussian node is not fixed, but rather determined based on the number of subdivided blocks, as in this invention.
[0111] Table 3 B under different Gaussian nodes z Component calculation results
[0112]
[0113] A storage medium storing instructions and data for implementing the high-precision forward modeling method for the spatial domain of the magnetic field effect of a spherical prism.
[0114] A high-precision forward modeling device for the spatial domain of magnetic field effects of a spherical prism includes: a processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement the high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism.
[0115] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.
[0116] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. In the unit claims listing several devices, several of these devices may be embodied by the same hardware item. The use of the terms first, second, and third, etc., does not indicate any order and can be interpreted as identifiers.
[0117] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A high-precision forward modeling method for the spatial domain of magnetic field effects on a spherical prism, characterized in that, include: S1: Obtain the initial spherical prism. Given the computation time of a single spherical prism with different Gaussian nodes in the initial spherical prism, calculate the total number of subdivided spherical prisms with different Gaussian nodes in the initial spherical prism. S2: Multiply the total number of subdivided spherical prisms of each Gaussian node by the calculation time of the corresponding single spherical prism to obtain the forward modeling prediction time of each Gaussian node, and select the Gaussian node with the shortest forward modeling prediction time as the node to perform forward modeling calculation. S3: If the total number of subdivided spherical prisms of the node to be performed in forward modeling is greater than 1, then the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks. Forward modeling is performed on each spherical prism sub-block and accumulated to obtain the forward modeling simulation result; otherwise, proceed to step S4. S4: Perform forward modeling calculations on the initial spherical prism to obtain the forward modeling simulation results.
2. The high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism according to claim 1, characterized in that, The Gaussian nodes are selected as i = 1-1-1 or 2-2-2, where i represents the number of the Gaussian node.
3. The high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism according to claim 1, characterized in that, Step S1 is as follows: S11: Given an initial spherical prism with Gaussian node i, the computation time for a single spherical prism is t. i , where i represents the number of the Gaussian node; S12: Obtain the number of subdivision blocks of Gaussian node i in each direction in the initial spherical prism. Where u is the direction number; S13: Multiply the number of subdivided blocks of Gaussian node i in each direction to obtain the total number of subdivided spherical prisms of Gaussian node i, Num. i .
4. The high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism according to claim 3, characterized in that, The expression for the total number of subdivided spherical prisms is: side=d / D R Where u represents direction. r indicates the direction pointing towards the Earth's center. λ and λ represent the latitude and longitude directions respectively, floor() represents an integer less than the calculated result, d represents the minimum Euclidean distance between the geometric center of the spherical prism and all observation points, and D R L represents the ratio of the observation distance to the size of the spherical prism under the required calculation accuracy. u This represents the side length of the original spherical prism in the direction u.
5. The high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism according to claim 1, characterized in that, The expression for the forward modeling prediction time is: The i =t i ×Number i Where i represents the Gaussian node number, Te i Let t represent the forward prediction time of Gaussian node i. i Num represents the computation time for a single spherical prism at Gaussian node i. i This represents the total number of subdivided spherical prisms at Gaussian node i.
6. The high-precision forward modeling method for the spatial domain of magnetic field effects of a spherical prism according to claim 4, characterized in that, In step S3, the initial spherical prism is subdivided to obtain multiple spherical prism sub-blocks, specifically as follows: After selecting the Gaussian node i of the initial spherical prism, in the direction r, according to the number of subdivision blocks... Divide into equal parts; in the direction Above, according to the number of subdivided blocks Divide into equal parts; along direction λ, according to the number of subdivided blocks. Divide into equal parts; after dividing into three directions, multiple spherical prism sub-blocks are obtained.
7. A storage medium, characterized in that: The storage medium stores instructions and data to implement the high-precision forward modeling method for the spatial domain of the magnetic field effect of a spherical prism as described in any one of claims 1 to 6.
8. A high-precision forward modeling device for the spatial domain of magnetic field effects of a spherical prism, characterized in that: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement the high-precision forward modeling method for the spatial domain of the magnetic field effect of a spherical prism as described in any one of claims 1 to 6.
Citation Information
Patent Citations
Gravity field forward modeling method and three-dimensional inversion method in spherical coordinate system based on 3D-GLQ
CN110045432A
System and method for evaluating petroleum reservoir using forward modeling
US20080255816A1
