A method and device for modeling structural plane occurrence, electronic equipment and storage medium

Abstract

The application provides a structural plane occurrence modeling method and device, electronic equipment and storage medium, and relates to the technical field of digital data processing, which comprises the following steps: obtaining structural plane occurrence information of a research area, determining corresponding random field characteristics according to the structural plane occurrence information, dividing a preset modeling area into a plurality of local subdomains, generating a local random field in each local subdomain, splicing the local random fields in the plurality of local subdomains into a global random field, performing geological constraint correction on the global random field, and generating a three-dimensional structural plane model based on the corrected global random field; the application can efficiently generate large-scale three-dimensional structural plane random fields, can take into account local differences and global statistical consistency, and can ensure the rationality of the model geology.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital data processing, in particular to a modeling method and device for structural plane occurrence, electronic equipment and storage medium. BACKGROUND

[0002] In geotechnical engineering, structural plane occurrence (dip direction, dip angle, etc.) is an important parameter for evaluating rock mass stability, block analysis, and tunnel design. However, in actual engineering, the measurement data of structural planes is usually limited, and the spatial distribution has significant randomness and variability. Traditional structural plane modeling methods mainly rely on point measurement data interpolation or regular grid generation, which has the following shortcomings:

[0003] (1) Limited data expansion capability: when the measured structural plane data is sparse, it is difficult to effectively generate a representative three-dimensional structural plane distribution, and the spatial variability of the rock mass cannot be fully reflected;

[0004] (2) Insufficient consideration of randomness: existing methods are mostly deterministic modeling, lacking quantification of structural plane occurrence uncertainty and spatial correlation, making it difficult to meet the needs of reliability analysis and probability evaluation;

[0005] (3) Low computational efficiency: in large engineering areas or high-resolution models, traditional random field generation methods have large computational load and high time consumption, making it difficult to generate random field samples of thousands, hundreds of thousands, or larger scales:

[0006] (4) Difficulty in balancing local and global consistency: when using subdomains or local blocks to generate random fields, it is still a technical difficulty to ensure the continuity between subblocks and the consistency of global statistical characteristics.

[0007] Therefore, there is an urgent need for an efficient modeling method that can efficiently generate large-scale three-dimensional structural plane random fields, while balancing local differences and global statistical consistency, and ensuring the geological reasonableness of the model. SUMMARY

[0008] The problem solved by the present application is to provide an efficient and stable structural plane occurrence modeling method.

[0009] To solve the above problems, the present application provides a modeling method and device for structural plane occurrence, electronic equipment and storage medium.

[0010] In a first aspect, the present application provides a modeling method for structural plane occurrence, comprising:

[0011] Obtaining structural plane occurrence information of a study area, and determining corresponding random field characteristics according to the structural plane occurrence information; the random field characteristics include statistical characteristics of the structural plane occurrence information and spatial correlation characteristics of the structural plane occurrence information;

[0012] divide a preset modeling region into a plurality of local sub-domains based on the spatial correlation feature of the structural plane occurrence information;

[0013] generate a local random field in each of the local sub-domains based on the random field feature of the structural plane occurrence information;

[0014] splice the local random fields in the plurality of local sub-domains into a global random field;

[0015] correct the global random field based on a geological constraint condition, and generate a three-dimensional structural plane model based on the corrected global random field.

[0016] Optionally, the structural plane occurrence information includes a set of dip direction data points and a set of dip angle data points of the structural plane.

[0017] Optionally, the statistical feature of the structural plane occurrence information includes a mean, a variance, a standard deviation, a coefficient of variation, and a probability distribution form of the structural plane occurrence information; and / or,

[0018] The spatial correlation feature of the structural plane occurrence information includes a spatial correlation function and a spatial correlation length of the structural plane occurrence information.

[0019] Optionally, the dividing of the preset modeling region into a plurality of local sub-domains based on the spatial correlation feature of the structural plane occurrence information includes:

[0020] determining a width of an overlapping region between two adjacent local sub-domains according to the spatial correlation length of the structural plane occurrence information;

[0021] determining the size of the local sub-domains;

[0022] dividing the preset modeling region into a plurality of local sub-domains that are continuous in space and partially overlapped according to the width of the overlapping region and the size of the local sub-domains.

[0023] Optionally, the generating of the local random field in each of the local sub-domains based on the random field feature of the structural plane occurrence information includes:

[0024] constructing a covariance matrix describing the correlation between any two sample points in the local sub-domain according to the spatial correlation feature in each of the local sub-domains;

[0025] decomposing the covariance matrix by a matrix decomposition method to obtain a lower triangular matrix;

[0026] generating a random vector subject to a standard normal distribution according to the statistical feature of the structural plane occurrence information;

[0027] generate a local random field in each of the local sub-domains based on the lower triangular matrix and the random vector subject to the standard normal distribution.

[0028] Optionally, the splicing of the local random fields in the plurality of local sub-domains into a global random field comprises:

[0029] For each sample point in the space in the overlapping region of the two adjacent local sub-domains, a weight coefficient of each sample point relative to the adjacent local sub-domain is calculated;

[0030] Based on the weight coefficient of each sample point relative to the adjacent local sub-domain, a local random field value of the corresponding adjacent local sub-domain is weighted and fused to generate a global random field value of each sample point in the space in the overlapping region;

[0031] In the non-overlapping region of the two adjacent local sub-domains, a local random field value of the local sub-domain is used as a global random field value of each sample point in the space in the non-overlapping region;

[0032] According to the global random field value of each sample point in the space in the overlapping region and the global random field value of each sample point in the space in the non-overlapping region, a global random field is obtained.

[0033] Optionally, the generating of the three-dimensional structural surface model based on the corrected global random field comprises:

[0034] A preset structural surface geometric equation is used to construct the three-dimensional structural surface model based on the corrected global random field.

[0035] In a second aspect, the application provides a modeling device for structural surface occurrence, comprising:

[0036] A random field feature determination module is configured to acquire structural surface occurrence information of a research region and determine corresponding random field features according to the structural surface occurrence information; the random field features include statistical features of the structural surface occurrence information and spatial correlation features of the structural surface occurrence information.

[0037] A local sub-domain division module is configured to divide a preset modeling region into a plurality of local sub-domains based on the spatial correlation features of the structural surface occurrence information.

[0038] A local random field generation module is configured to generate a local random field in each of the local sub-domains based on the random field features of the structural surface occurrence information.

[0039] A global random field generation module is configured to splice the local random fields in the plurality of local sub-domains into a global random field.

[0040] The three-dimensional structural surface model generation module is configured to correct the global random field based on a geological constraint condition, and generate a three-dimensional structural surface model based on the corrected global random field.

[0041] In a third aspect, the present application provides an electronic device comprising a memory and a processor;

[0042] The memory is configured to store a computer program.

[0043] The processor is configured to implement the modeling method of structural surface occurrence when executing the computer program.

[0044] In a fourth aspect, the present application provides a computer readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the modeling method of structural surface occurrence is implemented.

[0045] The modeling method of structural surface occurrence, the device, the electronic device and the storage medium have the following beneficial effects: the structural surface occurrence information of a research area is obtained, and the corresponding random field characteristics are determined according to the structural surface occurrence information, the random field characteristics include statistical characteristics of the structural surface occurrence information and spatial correlation characteristics of the structural surface occurrence information, and accurate statistical rules and spatial structure constraints are provided for subsequent modeling. Based on the spatial correlation characteristics of the structural surface occurrence information, a preset modeling area is divided into a plurality of local sub-domains, the computational complexity and memory occupation are reduced from the global scale to the sub-domain scale, and the modeling efficiency is improved. Based on the random field characteristics of the structural surface occurrence information, a local random field is generated in each local sub-domain, and a parameter field that meets the preset statistical and spatial characteristics is independently generated in each local sub-domain, so that the local spatial structure rationality is ensured. The local random fields in the plurality of local sub-domains are spliced into a global random field, the numerical jump at the boundary of the local sub-domains is effectively eliminated, and the spatial continuity and statistical consistency of the final global random field are ensured. The global random field is corrected based on a geological constraint condition, and a three-dimensional structural surface model is generated based on the corrected global random field, the parameters that exceed the geological constraint condition are constrained and corrected, the geological rationality of each structural surface occurrence parameter is ensured, and finally the three-dimensional structural surface model is efficiently and stably output. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 A flowchart of a modeling method of structural surface occurrence according to an embodiment of the present application;

[0047] Figure 2 A flowchart of dividing a modeling area according to an embodiment;

[0048] Figure 3 A flowchart of generating a local random field according to an embodiment;

[0049] Figure 4 A flow chart of local random field splicing to global random field for an embodiment;

[0050] Figure 5 A polar coordinate plot of a revised structural plane inclination angle random field sample for an embodiment;

[0051] Figure 6 A spatial correlation test plot of a structural plane occurrence information random field sample for an embodiment;

[0052] Figure 7 A visualization schematic of a three-dimensional structural plane model obtained by modeling for an embodiment;

[0053] Figure 8 A structural schematic of a structural plane occurrence modeling device for an embodiment of the application;

[0054] Figure 9 A structural schematic of an electronic device for an embodiment of the application. DETAILED DESCRIPTION

[0055] In order to make the above objectives, characteristics and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with the accompanying drawings. Although some embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms, and should not be interpreted as being limited to the embodiments set forth herein, but rather, these embodiments are provided in order to more thoroughly and completely understand the present application. It should be understood that the drawings and embodiments of the present application are for exemplary purposes only, and are not intended to limit the scope of protection of the present application.

[0056] It should be understood that each of the steps recited in the method embodiments of the present application can be executed in different orders, and / or in parallel. In addition, the method embodiments can include additional steps and / or omit the execution of the steps shown. The scope of the present application is not limited in this respect.

[0057] The term "comprising" and variations thereof as used herein are open-ended, that is, "comprising but not limited to"; the term "based on" is "based at least in part on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optional" means "optional in at least some embodiments". Related definitions are given throughout the description. It should be noted that the concepts mentioned in the present application using "first", "second", etc. are only used to distinguish different devices, modules or units, and are not intended to limit the order or interdependence of the functions performed by these devices, modules or units.

[0058] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0059] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0060] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for modeling the orientation of structural surfaces, comprising the following steps:

[0061] Step S110: Obtain the structural surface attitude information of the study area, and determine the corresponding random field characteristics based on the structural surface attitude information.

[0062] Specifically, the study area refers to the spatial extent of structural planes within an actual geological body, used to provide accurate geostatistical characteristics. It represents the "physical region" of the structural planes in three-dimensional space. For example, the surrounding rock extent around an underground cavern, slope, or tunnel. The attitude information of the structural planes in the study area refers to the set of dip data points and the set of dip angle data points. Each data point in the dip data point set includes its spatial coordinates and its corresponding dip value, where the dip value is the direction of the structural plane's inclination on the horizontal plane. Each data point in the dip angle data point set includes its spatial coordinates and its corresponding dip angle value, where the dip angle is the angle between the structural plane and the horizontal plane.

[0063] Specifically, the random field characteristics include: statistical characteristics of structural surface attitude information and spatial correlation characteristics of structural surface attitude information. The statistical characteristics of structural surface attitude information include: the mean, variance, standard deviation, coefficient of variation, and probability distribution form of the structural surface attitude information. These statistical characteristics can be obtained by statistically analyzing the dip and angle data points separately. The probability distribution form can be a log-normal distribution, a Von Mises distribution, or a Fisher distribution. The spatial correlation characteristics of structural surface attitude information include: the spatial correlation function and the spatial correlation length. The spatial correlation function is a mathematical model that quantitatively describes how the statistical similarity between the dip and angle of a structural surface at different locations in space decreases with increasing distance. The spatial correlation length is a key scale parameter in the correlation function, quantitatively characterizing the average distance at which the dip and angle of a structural surface maintain a significant correlation in space.

[0064] Step S120: Based on the spatial correlation characteristics of the structural surface attitude information, the preset modeling region is divided into multiple local subdomains.

[0065] In some embodiments, as shown in FIG. 1, based on the spatial correlation characteristics of the structural plane occurrence information, the preset modeling region is divided into a plurality of local sub-domains, including the following steps: Figure 2

[0066] Step S210: determining the width of the overlapping region between two adjacent local sub-domains according to the spatial correlation length of the structural plane occurrence information.

[0067] Specifically, since the spatial correlation length L is the characteristic scale of the structural plane occurrence information, the width of the overlapping region must be wide enough to cover the correlation decay process on this scale. In some embodiments, the width of the overlapping region can be 1 to 3 times the spatial correlation length L. θ θ Since the overlapping region is a key buffer area to eliminate the numerical mutation of the block boundary and ensure the spatial continuity of the global random field, a reasonable width of the overlapping region lays a foundation for the smooth splicing between the subsequent local sub-domains.

[0068]

[0069] Step S220: determining the size of the local sub-domain.

[0070] Specifically, the upper limit of the sample capacity of the local sub-domain is set to Nmax, and the sample size of each local sub-domain is dynamically determined according to the total sample number N, and the rule is as follows: M N

[0071] (1)

[0072] wherein, L is the size of the local sub-domain, and Nmax can be 1000. M

[0073] When the total sample number N is small, all samples are generated at one time; when the total sample number N is large, the local sub-domains are generated in groups of at most 1000 samples, and then spliced, so as to effectively reduce the memory occupation on the premise of ensuring the generation accuracy.

[0074] Step S230: dividing the preset modeling region into a plurality of local sub-domains which are spatially continuous and partially overlapped according to the width of the overlapping region and the size of the local sub-domain.

[0075] In this optional embodiment, by using the above block method of the modeling region, the calculation amount of the large-scale matrix operation in the traditional global random field generation can be significantly reduced, and the random field parameter inversion of the order of ten million can be easily realized.

[0076] Step S130: generating a local random field in each local sub-domain based on the random field characteristics of the structural plane occurrence information. ​​​​​​​

[0077] In some embodiments, as shown in FIG. 3, based on the random field characteristics of the structural plane occurrence information, a local random field is generated in each local subdomain, including: Figure 3

[0078] Step S310: In each local subdomain, a covariance matrix describing the correlation between any two sample points in the space is constructed according to the spatial correlation characteristics.

[0079] Specifically, for the structural plane inclination, dip angle and other parameter samples in each local subdomain, based on the spatial correlation function ρ ( h ):

[0080] ; (2)

[0081] Solving the spatial correlation function between two sample points in the above local subdomain, the covariance matrix C is constructed by traversing all samples in the subdomain, where is the spatial correlation length, h is the distance between two sample points in the local subdomain.

[0082] Specifically, the form of the covariance matrix C is:

[0083] ;

[0084] In the formula, ρ is the correlation function between two sample points, ρ 12 represents the correlation function between sample point 1 and sample point 2, n e represents the number of sample points.

[0085] By calculating the correlation function ρ for all sample points in the local subdomain, the covariance matrix C can be obtained, and the covariance matrix C is constructed.

[0086] Step S320: The matrix decomposition method is used to decompose the covariance matrix C to obtain a lower triangular matrix L .

[0087] Step S330: According to the statistical characteristics of the structural plane occurrence information, a random vector subject to standard normal distribution is generated.

[0088] ​Specifically, according to the mean, variance, standard deviation, coefficient of variation and probability distribution form in the statistical characteristics, a random vector that follows a standard normal distribution is generated, so that the samples in the local random field generated in each subsequent local subdomain satisfy the statistical characteristics of the structural surface attitude information.

[0089] Step S340: Generate a local random field in each local subdomain based on the lower triangular matrix and random vectors that follow a standard normal distribution.

[0090] Specifically, according to Generate a local random field, where, L It is a lower triangular matrix. For a random vector that follows a standard normal distribution, It is a local random field.

[0091] In this optional embodiment, by constructing and decomposing a covariance matrix that conforms to geological spatial correlation, abstract statistical features are efficiently transformed into random field samples with precise spatial structure. At the same time, it provides independent units for block-based parallel computing, laying the foundation for large-scale and efficient generation.

[0092] Step S140: Concatenate the local random fields in multiple local subdomains into a global random field.

[0093] In some embodiments, such as Figure 4 As shown, concatenating local random fields from multiple local subdomains into a global random field includes the following steps:

[0094] Step S410: For each sample point in the space within the overlapping region of two adjacent local subdomains, calculate the weight coefficient of each sample point relative to the adjacent local subdomain.

[0095] Specifically, the weight coefficients are calculated using a Gaussian weighting function, expressed as follows:

[0096] (3)

[0097] in, For sample points in the overlapping region, For the first The center coordinates of a local subdomain For the transition band smoothing parameter, it can be taken as 0.3 to 0.5 times the spatial correlation length. θ , For sample points For relative to the first Weight coefficients of each local subdomain For the first The center coordinates of the local subdomain, the th The local subdomain and the first The local subdomains are adjacent and partially overlap.

[0098] Step S420: Based on the weight coefficient of each sample point relative to the adjacent local subdomain, the local random field value of the corresponding adjacent local subdomain is weighted and fused to generate the global random field value of each sample point in the space in the overlapping region.

[0099] The local random field values of the adjacent local subdomains are weighted and fused to form the global random field value:

[0100] (4)

[0101] In the above formula, n is the number of local subdomains, is the local random field value of the sample point of the i-th local subdomain, is the global random field value of the sample point .

[0102] By the above method, the spatial difference of the subdomain transition zone can be reduced, and the generated structural surface occurrence information random field sample has global statistical feature consistency and local difference.

[0103] The above Z(x) is a standard normal random field, that is, a random field with a mean of 0, a standard deviation of 1 and a normal distribution, and through an equal probability mathematical transformation, a random field with a mean of , a variance of , a coefficient of variation COV and a target probability distribution form can be obtained.

[0104] Step S430: In the non-overlapping region of the adjacent two local subdomains, the local random field value of the local subdomain is used as the global random field value of each sample point in the space in the non-overlapping region. That is, for the non-overlapping region, the local random field value is used as the global random field value.

[0105] Step S440: According to the global random field value of each sample point in the space in the overlapping region and the global random field value of each sample point in the space in the non-overlapping region, a global random field is obtained.

[0106] In the above method, the spatial difference of the local subdomain transition zone can be reduced, and the generated structural surface occurrence information random field sample has global statistical feature consistency and local difference.

[0107] Step S150: Based on the geological constraint condition, the global random field is modified, and a three-dimensional structural surface model is generated based on the modified global random field.

[0108] ​Specifically, the generated global random field is geologically constrained and corrected based on geological constraints to ensure that the parameters meet physical and geological rationality. The geological rationality and statistical consistency of the global random field are ensured by setting physical reasonable interval constraints on the generated structural surface occurrence information (including dip direction and dip angle), and when the global random field value exceeds the geologically acceptable range, the value is adjusted by truncation correction or resampling method. For example, in geology, the interval of the dip direction of the structural surface is [0, 360), and the interval of the dip angle is [0, 90]. By limiting the generated global random field value within a fixed interval, the generated global random field can have actual physical meaning.

[0109] In some embodiments, a three-dimensional structural surface model is generated based on the corrected global random field, including:

[0110] A three-dimensional structural surface model is constructed based on the corrected global random field using a preset structural surface geometric equation.

[0111] Specifically, the preset structural surface geometric equation is:

[0112] (5)

[0113] The three-dimensional structural surface model is constructed, wherein, d is a preset parameter, a , b , c is a normal vector coefficient, which satisfies:

[0114] ( a , b , c ) = (sin sin , cos sin , cos ) (6)

[0115] In the formula, is a dip direction, is a dip angle.

[0116] The preset parameter d is determined by the normal vector coefficient and the coordinates of any one point (P0, P1, P2) on the structural surface, x 0, y 0, z 0, d = -( a × x 0+ b × y 0+ c × z 0).

[0117] A set of dip direction and dip angle of the structural plane and dip angle Then the parameters of the structural plane geometry equation (i.e. formula (5)) can be solved respectively by this formula (6) a , b , c Then the preset parameters x 0, y 0, z 0) of any point on the structural plane in space are determined d Thus a structural plane geometry equation can be completely determined, and a plane, i.e. a structural plane model, can be drawn in space according to this structural plane geometry equation. Thus, each set of dip direction and dip angle can uniquely correspond to a structural plane model. Finally, the generated multiple structural plane models are superimposed and displayed in a unified coordinate system to realize three-dimensional visualization of the structural plane in space, i.e. to obtain a three-dimensional structural plane model, thereby completing the efficient modeling and visualization of the structural plane based on the block random field.

[0118] Specifically, for the tunnel surrounding rock, the coordinates of any point x 0, y 0, z 0) can be the spatial position point coordinates of the structural plane in the modeling area, and are preferably the center of each local subdomain or equidistant sampling points along the tunnel axis direction to approximately represent the spatial position of the structural plane in the tunnel surrounding rock.

[0119] By iteratively generating a set of structural planes in the entire domain and outputting a three-dimensional structural plane model in a standard three-dimensional geometry format (such as OBJ or STL), efficient modeling and visualization of the spatial occurrence of the structural plane are realized.

[0120] In this embodiment, the structural plane occurrence information of the study area is obtained, and the corresponding random field characteristics are determined according to the structural plane occurrence information. The random field characteristics include statistical characteristics of the structural plane occurrence information and spatial correlation characteristics of the structural plane occurrence information, which provide accurate statistical rules and spatial structure constraints for subsequent modeling. Based on the spatial correlation characteristics of the structural plane occurrence information, the preset modeling area is divided into multiple local subdomains, and the computational complexity and memory occupation are reduced from the global scale to the subdomain scale, thereby improving the modeling efficiency. Based on the random field characteristics of the structural plane occurrence information, a local random field is generated in each local subdomain, and a parameter field that satisfies the preset statistical and spatial characteristics is independently generated in each local subdomain, thereby ensuring the rationality of the local spatial structure. The local random fields in multiple local subdomains are spliced into a global random field, which effectively eliminates the numerical discontinuity at the boundary of the local subdomain and ensures the spatial continuity and statistical consistency of the final global random field. Based on the geological constraint conditions, the global random field is modified, and a three-dimensional structural plane model is generated based on the modified global random field. By constraining and modifying the parameters that exceed the geological constraint conditions, the geological rationality of each structural plane occurrence parameter is ensured, and finally a three-dimensional structural plane model is efficiently and stably output.

[0121] The following provides an example based on the above structural plane occurrence modeling method.

[0122] Step S110: Obtain the structural plane occurrence information of the study area, and determine the corresponding random field characteristics according to the structural plane occurrence information.

[0123] According to the collected geological statistical information, 31 groups of structural planes are counted, and the average inclination of the structural planes is 193°, the average dip angle is 56.8°, and the standard deviations are 107.2° and 18.45°, respectively. According to the calculation, the coefficients of variation are 0.324 and 0.555, respectively. The spatial correlation function can be a single exponential type (such as the above formula (2)), and the spatial correlation length The spatial autocorrelation function fitting method is used to locate 1.8m.

[0124] Step S210: Determine the width of the overlapping region between the two adjacent local subdomains according to the spatial correlation length of the structural plane occurrence information.

[0125] In order to illustrate the effectiveness of the above block method of the modeling area in establishing a large-scale structural plane occurrence parameter random field, in an embodiment, 10000 structural plane parameters are generated, the size of the local subdomain is 1000, and then the number of local subdomains is 10000 / 1000=10, and the width of the overlapping region is 3 times the spatial correlation distance θ , that is, 5.4m, and the integer part is 5m.

[0126] Step S130: generating a local random field in each local sub-domain based on the random field characteristics of the structural plane occurrence information.

[0127] Solving the correlation coefficient of each sample point in the space within the random field of the 10 local sub-domains by using the above formula (2) ρ h ), and constructing a covariance matrix C , and then performing matrix decomposition on the covariance matrix C , and obtaining the local random field in the 10 local sub-domains by using .

[0128] It should be noted that the covariance matrix C and the matrix decomposition are not solved in the whole space, but in the local sub-domains after blocking. In this embodiment, when generating the structural plane occurrence information random field by using the matrix decomposition method, the traditional method needs to process a 10000x10000 covariance matrix, and the matrix decomposition calculation complexity is , and the memory occupation is about 763MB. Adding the lower triangular matrix L after decomposition, the memory occupation will reach about 1.5GB; while the block matrix decomposition method adopted in the embodiment of the present application processes the large matrix by decomposing it into multiple matrix sub-domains: the first block only needs to process a 1000x1000 matrix, and the matrix dimension of the non-first block is 1005x1005 after considering the overlapping area. Theoretical analysis shows that the calculation complexity of the block method is significantly reduced from of the traditional method to , with a reduction of 98.875%; in terms of memory occupation, the peak memory requirement of the block method is greatly reduced from 1.5GB of the traditional method to about 7.68MB, with a reduction of 98.99%. Such a performance improvement of several orders of magnitude makes it possible to efficiently generate large-scale structural plane occurrence random fields on ordinary computing devices, and successfully solves the calculation complexity disaster and memory bottleneck problem faced by the existing method.

[0129] Step S140: splicing the local random fields in the multiple local sub-domains into a global random field.

[0130] The specific implementation process of splicing the overlapping areas of the local sub-domains is as follows: first, a transition constraint area with a width of 3 times the spatial correlation length is set at the boundary of each sub-domain to ensure that the adjacent local sub-domains have sufficient overlapping areas for smooth transition. In the transition constraint area, a Gaussian-type weight function (such as the above formula (3)) is used to calculate the weight coefficients of each sample point in the space relative to the adjacent local sub-domains, wherein the transition band smoothing parameter is 2 times the spatial correlation length θ. In the specific calculation process, for any sample point x in the transition area, the weight values and relative to the center coordinates of the adjacent local sub-domains are calculated respectively.Then, the random field values of adjacent sub-domains are fused by weighting according to formula (4) and to generate the global random field value of each point in the overlapping area . In the non-overlapping area, the random field value of the local sub-domain is directly used.

[0131] Step S150: Based on the geological constraint conditions, the global random field is modified, and a three-dimensional structural surface model is generated based on the modified global random field.

[0132] Based on geological exploration data and engineering experience, the reasonable value range of each occurrence parameter is determined, wherein the dip angle is limited between 0°-90°, and the dip direction is limited within 0°-360°. For random field sample values exceeding the reasonable range, a boundary constraint correction method based on the truncation method is used for processing: when the generated dip angle value exceeds 90°, a uniform distribution resampling method is used for correction, the number of samples exceeding the range is counted, and the corresponding number of replacement values are uniformly generated within the reasonable value interval for replacement; when the dip direction value exceeds the range of 0°-360°, it is mapped to the [0, 360) interval through a modulo operation. Through the above multi-level constraint correction, it is ensured that the generated random field samples meet the requirements of physical and geological laws, and the rationality of their statistical characteristics is maintained, so that the final result not only meets the mathematical statistical requirements, but also has practical geological significance. For example, Figure 5 and Figure 6 wherein, Figure 5 Fig. 4 shows the polar coordinate diagram of the modified structural surface dip direction and dip angle random field samples, Figure 6 Fig. 5 shows the spatial correlation test diagram of the structural surface occurrence information random field samples, and it can be seen that the generated samples are in good consistency with the preset conditions in terms of randomness and correlation.

[0133] The global random field modified based on the geological constraint conditions is taken as input, and the normal vector coefficient a of each structural surface is calculated according to formula (6) b , c . On this basis, according to the preset structural surface spatial geometric equation shown in formula (5), the value of the preset parameter d is determined in combination with the structural surface center point coordinates, so as to completely define the position and occurrence of each structural surface in the three-dimensional space, i.e. to obtain the three-dimensional structural surface model. Through this method, the modified global random field is converted into a three-dimensional structural surface model with clear spatial position and geometric shape, realizing effective mapping from statistical parameters to three-dimensional geometric entities. In the visualization link, the generated structural surface is rendered in three dimensions based on the normal vector coloring technology, and through transparency adjustment and spatial distribution optimization, the spatial distribution law and occurrence change characteristics of the three-dimensional structural surface model are clearly displayed. For example,Figure 7 As shown, Figure 7 The visualization of the three-dimensional structural surface model obtained by modeling is shown, and the output format of the three-dimensional structural surface model can be OBJ format.

[0134] As Figure 8 As shown, the structural surface occurrence modeling device 800 provided by the embodiment of the application comprises:

[0135] The random field feature determination module 810 is configured to acquire structural surface occurrence information of a study area, and determine corresponding random field features according to the structural surface occurrence information; the random field features comprise statistical features of the structural surface occurrence information and spatial correlation features of the structural surface occurrence information.

[0136] The local subdomain division module 820 is configured to divide a preset modeling region into a plurality of local subdomains based on the spatial correlation features of the structural surface occurrence information.

[0137] The local random field generation module 830 is configured to generate a local random field in each of the local subdomains based on the random field features of the structural surface occurrence information.

[0138] The global random field generation module 840 is configured to splice the local random fields in the plurality of local subdomains into a global random field.

[0139] The three-dimensional structural surface model generation module 850 is configured to correct the global random field based on a geological constraint condition, and generate a three-dimensional structural surface model based on the global random field after the correction.

[0140] Optionally, the structural surface occurrence information comprises a set of dip data points and a set of dip angle data points of the structural surface.

[0141] Optionally, the statistical features of the structural surface occurrence information comprise a mean value, a variance, a standard deviation, a coefficient of variation and a probability distribution form of the structural surface occurrence information; and / or,

[0142] The spatial correlation features of the structural surface occurrence information comprise a spatial correlation function and a spatial correlation length of the structural surface occurrence information.

[0143] Optionally, the division of the preset modeling region into the plurality of local subdomains based on the spatial correlation features of the structural surface occurrence information comprises:

[0144] Determining a width of an overlapping region between two adjacent local subdomains according to the spatial correlation length of the structural surface occurrence information;

[0145] Determining the size of the local subdomain;

[0146] According to the overlap region width and the size of the local subdomain, a preset modeling region is divided into a plurality of spatially continuous and partially overlapping local subdomains.

[0147] Optionally, the random field feature based on the structural plane occurrence information comprises:

[0148] In each local subdomain, a covariance matrix describing the correlation between any two sample points in the space in the local subdomain is constructed according to the spatial correlation feature;

[0149] The covariance matrix is decomposed by using a matrix decomposition method to obtain a lower triangular matrix;

[0150] According to the statistical feature of the structural plane occurrence information, a random vector subject to a standard normal distribution is generated;

[0151] Based on the lower triangular matrix and the random vector subject to the standard normal distribution, a local random field is generated in each local subdomain.

[0152] Optionally, the local random fields in the plurality of local subdomains are spliced into a global random field, comprising:

[0153] For each sample point in the space in the overlap region of the two adjacent local subdomains, a weight coefficient of each sample point relative to the adjacent local subdomain is calculated;

[0154] Based on the weight coefficient of each sample point relative to the adjacent local subdomain, the local random field value of the corresponding adjacent local subdomain is weighted and fused to generate a global random field value of each sample point in the space in the overlap region;

[0155] In the non-overlapping region of the two adjacent local subdomains, the local random field value of the local subdomain is used as the global random field value of each sample point in the space in the non-overlapping region;

[0156] According to the global random field value of each sample point in the space in the overlap region and the global random field value of each sample point in the space in the non-overlapping region, a global random field is obtained.

[0157] Optionally, the three-dimensional structural plane model is generated based on the corrected global random field, comprising:

[0158] A preset structural plane geometric equation is used to construct the three-dimensional structural plane model based on the corrected global random field.

[0159] As Figure 9As shown, an electronic device 900 provided by an embodiment of the present application includes a memory 910 and a processor 920; the memory 910 is configured to store a computer program; the processor 920 is configured to implement the modeling method of structural plane occurrence when executing the computer program.

[0160] Alternatively, an electronic device 900 includes a memory 910 and a processor 920 coupled to the memory 910; the memory 910 is configured to store a computer program; the processor 920 is configured to execute the following operations when executing the computer program:

[0161] obtain structural plane occurrence information of a study area, and determine corresponding random field characteristics according to the structural plane occurrence information; the random field characteristics include statistical characteristics of the structural plane occurrence information and spatial correlation characteristics of the structural plane occurrence information;

[0162] divide a preset modeling area into a plurality of local subdomains based on the spatial correlation characteristics of the structural plane occurrence information;

[0163] generate a local random field in each of the local subdomains based on the random field characteristics of the structural plane occurrence information;

[0164] splice the local random fields in the plurality of local subdomains into a global random field;

[0165] correct the global random field based on a geological constraint condition, and generate a three-dimensional structural plane model based on the corrected global random field.

[0166] A computer readable storage medium provided by an embodiment of the present application has a computer program stored thereon, and when the computer program is executed by a processor, the modeling method of structural plane occurrence as described above is implemented.

[0167] Alternatively, a non-volatile computer readable storage medium has a computer program stored thereon, and when the computer program is executed by a processor, the processor executes the following operations:

[0168] obtain structural plane occurrence information of a study area, and determine corresponding random field characteristics according to the structural plane occurrence information; the random field characteristics include statistical characteristics of the structural plane occurrence information and spatial correlation characteristics of the structural plane occurrence information;

[0169] divide a preset modeling area into a plurality of local subdomains based on the spatial correlation characteristics of the structural plane occurrence information;

[0170] generate a local random field in each of the local subdomains based on the random field characteristics of the structural plane occurrence information;

[0171] stitching a local random field in a plurality of the local subdomains into a global random field;

[0172] correcting the global random field based on a geological constraint condition, and generating a three-dimensional structural surface model based on the corrected global random field.

[0173] An electronic device 900, which can be a server or a client of the present application, will now be described, which is an example of a hardware device that can be applied to aspects of the present application. The electronic device 900 is intended to represent various forms of digital electronic computer devices such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device 900 can also represent various forms of mobile devices such as personal digital assistants, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0174] The electronic device 900 includes a computing unit that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) or a computer program loaded into a random access memory (RAM) from a storage unit. Various programs and data required for device operation can also be stored in the RAM. The computing unit, the ROM, and the RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0175] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM), or a random access memory (RAM), etc. In this application, the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments of the present application. In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0176] Although the present application has been disclosed with reference to the above embodiments, the scope of the present application is not limited to the above. Various changes and modifications can be made to the present application without departing from the spirit and scope thereof, and such changes and modifications are intended to fall within the scope of the present application.

Claims

1. A method of modeling structural plane attitudes, characterized by, The method comprises the following steps: obtaining the structural plane occurrence information of a study area, and determining corresponding random field characteristics according to the structural plane occurrence information; the random field characteristics comprise statistical characteristics of the structural plane occurrence information and spatial correlation characteristics of the structural plane occurrence information; the structural plane occurrence information comprises a set of dip data points and a set of dip angle data points of the structural plane; the statistical characteristics of the structural plane occurrence information comprise mean, variance, standard deviation, coefficient of variation and probability distribution form of the structural plane occurrence information; and / or the spatial correlation characteristics of the structural plane occurrence information comprise spatial correlation function and spatial correlation length of the structural plane occurrence information; dividing a preset modeling area into a plurality of local subdomains based on the spatial correlation characteristics of the structural plane occurrence information; generating a local random field in each local subdomain based on the random field characteristics of the structural plane occurrence information; splicing the local random fields in the plurality of local subdomains into a global random field; correcting the global random field based on a geological constraint condition, and generating a three-dimensional structural plane model based on the corrected global random field; the step of generating a local random field in each local subdomain based on the random field characteristics of the structural plane occurrence information comprises: in each local subdomain, constructing a covariance matrix describing the correlation between any two sample points in the space in the local subdomain according to the spatial correlation characteristics; decomposing the covariance matrix by using a matrix decomposition method to obtain a lower triangular matrix; generating a random vector subject to a standard normal distribution according to the statistical characteristics of the structural plane occurrence information; generating a local random field in each local subdomain based on the lower triangular matrix and the random vector subject to a standard normal distribution; the step of splicing the local random fields in the plurality of local subdomains into a global random field comprises: for each sample point in the space in the overlapping region of two adjacent local subdomains, calculating a weight coefficient of each sample point relative to the adjacent local subdomain; based on the weight coefficient of each sample point relative to the adjacent local subdomain, weighting and fusing the local random field values of the corresponding adjacent local subdomains to generate a global random field value of each sample point in the space in the overlapping region; in the non-overlapping region of two adjacent local subdomains, using the local random field value of the local subdomain as the global random field value of each sample point in the space in the non-overlapping region; obtaining the global random field according to the global random field value of each sample point in the space in the overlapping region and the global random field value of each sample point in the space in the non-overlapping region.

2. The method for modeling structural plane behavior according to claim 1, wherein, the step of dividing a preset modeling area into a plurality of local subdomains based on the spatial correlation characteristics of the structural plane occurrence information comprises: determining the width of the overlapping region between two adjacent local subdomains according to the spatial correlation length of the structural plane occurrence information; determining the size of the local subdomain; dividing the preset modeling area into a plurality of local subdomains which are continuous in space and partially overlapping according to the width of the overlapping region and the size of the local subdomain.

3. The method for modeling structural plane behavior according to claim 1, wherein, the step of generating a three-dimensional structural plane model based on the corrected global random field comprises: The three-dimensional structural surface model is constructed using a preset structural surface geometric equation and based on the modified global random field.

4. An apparatus for modeling structural plane attitudes, characterized by include: The random field feature determination module is used to acquire the structural surface attitude information of the study area and determine the corresponding random field features based on the structural surface attitude information. The random field features include: statistical features of the structural surface attitude information and spatial correlation features of the structural surface attitude information; the structural surface attitude information includes: the set of dip data points and the set of dip angle data points of the structural surface; the statistical features of the structural surface attitude information include: the mean, variance, standard deviation, coefficient of variation, and probability distribution form of the structural surface attitude information; and / or, the spatial correlation features of the structural surface attitude information include: the spatial correlation function and spatial correlation length of the structural surface attitude information; The local subdomain partitioning module is used to divide the preset modeling region into multiple local subdomains based on the spatial correlation characteristics of the structural surface attitude information. A local random field generation module is used to generate a local random field in each local subdomain based on the random field characteristics of the structural surface attitude information. A global random field generation module is used to concatenate local random fields in multiple local subdomains into a global random field; The three-dimensional structural surface model generation module is used to modify the global random field based on geological constraints, and generate a three-dimensional structural surface model based on the modified global random field. The random field feature based on the structural plane attitude information generates a local random field in each local subdomain, including: In each of the local subdomains, a covariance matrix describing the correlation between any two sample points in the space within the local subdomain is constructed based on the spatial correlation characteristics. The covariance matrix is ​​decomposed using matrix decomposition to obtain a lower triangular matrix; Based on the statistical characteristics of the structural surface attitude information, a random vector following a standard normal distribution is generated; Based on the lower triangular matrix and the random vectors following a standard normal distribution, a local random field is generated in each local subdomain; The step of concatenating local random fields from multiple local subdomains into a global random field includes: For each sample point in the space within the overlapping region of two adjacent local subdomains, calculate the weight coefficient of each sample point relative to the adjacent local subdomain; Based on the weight coefficient of each sample point relative to the adjacent local subdomain, the local random field values ​​of the corresponding adjacent local subdomains are weighted and fused to generate the global random field value of each sample point in the space of the overlapping region. In the non-overlapping region of two adjacent local subdomains, the local random field value of the local subdomain is used as the global random field value of each sample point in the space within the non-overlapping region. The global random field is obtained by using the global random field value of each sample point in the space within the overlapping region and the global random field value of each sample point in the space within the non-overlapping region.

5. An electronic device, comprising: Including memory and processor; The memory is used to store computer programs; The processor is configured to, when executing the computer program, implement the method for modeling the orientation of structural surfaces as described in any one of claims 1 to 3.

6. A computer readable storage medium characterized by, The storage medium has stored thereon a computer program which, when executed by a processor, implements the modeling method of structural plane attitudes according to any one of claims 1 to 3.

Images