Rock-soil body parameter three-dimensional random field generation method, device, equipment and medium

Abstract

The invention provides a rock-soil mass parameter three-dimensional random field generation method, device and equipment and a medium, and relates to the technical field of geotechnical engineering.The method comprises the steps that statistical information is acquired, and statistical characteristics are determined based on the statistical information; performing grid division to obtain a plurality of three-dimensional grid units, and constructing a target matrix based on the space coordinates and the statistical characteristics; obtaining a preset inversion model structure comprising an undetermined coefficient matrix; determining an undetermined coefficient matrix based on a preset inversion model structure and the target matrix, and obtaining an inversion model based on the undetermined coefficient matrix; and generating each independent random vector, and inputting each independent random vector into the inversion model to obtain a target three-dimensional random field. The cross-correlation part of the Gaussian random field generated by the inversion model and the preorder Gaussian random field share an independent random vector, and the new independent random vector introduced by the self-correlation part is shared with the subsequently constructed Gaussian random field, so that random field coupling modeling of multiple rock-soil body parameters is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geotechnical engineering, in particular to a method, device, equipment and medium for generating a three-dimensional random field of geotechnological parameters. BACKGROUND

[0002] In the field of geotechnical engineering, accurate construction of a random field of geotechnological parameters is crucial for engineering stability analysis, disaster prediction and other work. Various parameters of geotechnological bodies have a decisive influence on the safety and reliability of engineering, and accurate description of the spatial variability and mutual relationship of geotechnological parameters can provide key basis for engineering design and construction.

[0003] However, the current method for constructing a random field of geotechnological parameters can only reflect the distribution characteristics of a single parameter, and it is difficult to reflect the coupling relationship between multiple parameters, which has obvious limitations. SUMMARY

[0004] The problem solved by the present application is how to realize random field coupling modeling of multiple geotechnological parameters.

[0005] To solve the above problems, the present application provides a method for generating a three-dimensional random field of geotechnological parameters, comprising: obtaining statistical information of multiple geotechnological parameters, and determining statistical characteristics based on the statistical information; wherein the statistical characteristics include autocorrelation distances corresponding to each of the geotechnological parameters, and mutual correlation coefficients between each of the geotechnological parameters; dividing a preset simulation region into a plurality of three-dimensional grid cells, and constructing a target matrix based on the spatial coordinates of each of the three-dimensional grid cells and the statistical characteristics; wherein the target matrix includes autocorrelation matrices of each of the geotechnological parameters, and mutual correlation matrices between each two of the geotechnological parameters; obtaining a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a mutual correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent Gaussian random fields corresponding to each of the geotechnological parameters; a mutual correlation part of one of the Gaussian random fields is determined based on each of the mutual correlation undetermined coefficient matrices corresponding to one of the geotechnological parameters and each independent random vector used to generate a previous Gaussian random field; an autocorrelation part of one of the Gaussian random fields is determined based on the autocorrelation undetermined coefficient matrix corresponding to one of the geotechnological parameters and one newly added independent random vector; determining the undetermined coefficient matrix based on the preset inversion model structure and the target matrix, and obtaining an inversion model based on the undetermined coefficient matrix; wherein the undetermined coefficient matrix is used to make the inversion model satisfy a preset constraint condition; the preset constraint condition comprises that a self-covariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding self-correlation matrix, and a cross-covariance matrix between each two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; generating each independent random vector in the inversion model, and inputting each independent random vector into the inversion model to obtain a target three-dimensional random field corresponding to each geotechnical parameter.

[0006] Optionally, the method further comprises: determining a spatial distance between each three-dimensional grid unit based on the spatial coordinates, and determining a self-correlation coefficient between each three-dimensional grid unit and other three-dimensional grid units based on the spatial distance and the self-correlation distance corresponding to the geotechnical parameter, to obtain a self-correlation matrix corresponding to the geotechnical parameter; determining the cross-correlation matrix according to a product of the self-correlation matrix and the corresponding cross-correlation coefficient.

[0007] Optionally, the preset inversion model structure satisfies: ; wherein n represents the nth geotechnical parameter; Xn represents the Gaussian random field corresponding to the nth geotechnical parameter; L represents the undetermined coefficient matrix; represents the independent random vector.

[0008] Optionally, the preset constraint condition satisfies: ; wherein, represents the cross-correlation matrix corresponding to the first geotechnical parameter and the nth geotechnical parameter, which is equal to the cross-covariance matrix between the corresponding Gaussian random fields; represents the self-correlation matrix corresponding to the nth geotechnical parameter, which is equal to the self-covariance matrix of the corresponding Gaussian random field; represents the undetermined coefficient corresponding to the three-dimensional grid unit; ne represents the number of three-dimensional grid units.

[0009] Optionally, the method further comprises: decompose the autocorrelation matrix of the geotechnical parameter corresponding to the first Gaussian random field represented by the preset inversion model structure to obtain the autocorrelation undetermined coefficient matrix of the geotechnical parameter; According to the representation order of each Gaussian random field in the preset inversion model structure, sequentially perform solving operation on the geotechnical parameters corresponding to the remaining Gaussian random fields, and the solving operation comprises determining the undetermined coefficient matrix of the geotechnical parameter according to the target matrix and the determined autocorrelation undetermined coefficient matrix of the geotechnical parameter.

[0010] Optionally, the determining the undetermined coefficient matrix of the geotechnical parameter according to the target matrix and the determined autocorrelation undetermined coefficient matrix of the geotechnical parameter comprises: determining the cross-correlation undetermined coefficient matrix of the geotechnical parameter according to the cross-correlation matrix and the determined autocorrelation undetermined coefficient matrix of the geotechnical parameter; subtracting the covariance term corresponding to the cross-correlation undetermined coefficient matrix of the geotechnical parameter from the autocorrelation matrix of the geotechnical parameter to obtain a residual covariance matrix, and decomposing the residual covariance matrix to obtain the autocorrelation undetermined coefficient matrix of the geotechnical parameter; wherein one covariance term comprises the product of one cross-correlation undetermined coefficient matrix and the transpose of the cross-correlation undetermined coefficient matrix.

[0011] Optionally, the inputting each independent random vector into the inversion model to obtain the target three-dimensional random field corresponding to each geotechnical parameter comprises: inputting each independent random vector into the inversion model to obtain a three-dimensional random field corresponding to each geotechnical parameter, and performing equal probability transformation on each three-dimensional random field to obtain each target three-dimensional random field.

[0012] In the present application, by acquiring statistical information of multiple geotechnical parameters, determining statistical characteristics such as self-correlation distance corresponding to each geotechnical parameter and cross-correlation coefficient between geotechnical parameters, it is beneficial to provide a basis for joint modeling of three-dimensional random fields of multiple geotechnical parameters. Further, the present application performs three-dimensional grid division on a preset simulation region, based on spatial coordinates and statistical characteristics of each three-dimensional grid unit, constructs a target matrix containing self-correlation matrix of each geotechnical parameter and cross-correlation matrix between each two geotechnical parameters, and comprehensively represents the self-correlation and cross-correlation characteristics of multiple geotechnical parameters in three-dimensional space. On this basis, a preset inversion model structure containing self-correlation undetermined coefficient matrix and cross-correlation undetermined coefficient matrix (i.e. undetermined coefficient matrix) is obtained, which provides a reliable model framework for generating Gaussian random fields that match the coupling characteristics of multiple geotechnical parameters. In the present application, the preset inversion model structure is used to represent the Gaussian random fields corresponding to each geotechnical parameter in turn, the cross-correlation part of a Gaussian random field is determined based on each cross-correlation undetermined coefficient matrix corresponding to one geotechnical parameter and each independent random vector used to generate the previous Gaussian random field; the self-correlation part of a Gaussian random field is determined based on the self-correlation undetermined coefficient matrix corresponding to one geotechnical parameter and one newly added independent random vector. Correspondingly, in the actual application of the inversion model, the Gaussian random fields of each geotechnical parameter are constructed in turn based on the inversion model, the cross-correlation part of the Gaussian random field shares the independent random vector with the previous Gaussian random field, and the new independent random vector introduced by the self-correlation part of the Gaussian random field will be shared with the Gaussian random field constructed subsequently, thereby coupling the spatial dependence relationship between multiple geotechnical parameters. After obtaining the preset inversion model structure, the present application can determine the undetermined coefficient matrix that can make the inversion model satisfy the preset constraint condition based on the preset inversion model structure and the target matrix. The self-covariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding self-correlation matrix, and the cross-covariance matrix between each two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix, which is the preset constraint condition, combined with the preset inversion model structure and the target matrix, the undetermined coefficient matrix can be inversely solved, thereby obtaining the inversion model. It is beneficial to ensure that the statistical characteristics corresponding to the Gaussian random field generated by the inversion model can match the target matrix, thereby truly reflecting the spatial distribution characteristics of each geotechnical parameter and the coupling effect between multiple geotechnical parameters in space. In this way, after generating each independent random vector in the inversion model, the present application inputs each independent random vector into the inversion model, thereby obtaining the target three-dimensional random field corresponding to each geotechnical parameter, thereby realizing the random field coupling modeling of multiple geotechnical parameters.

[0013] The present application also provides a geotechnical parameter three-dimensional random field generation device, comprising: a statistical module configured to obtain statistical information of a plurality of geotechnical parameters, and determine statistical characteristics based on the statistical information, wherein the statistical characteristics comprise autocorrelation distances corresponding to the geotechnical parameters and cross-correlation coefficients between the geotechnical parameters; a division module configured to divide a preset simulation region into a plurality of three-dimensional grid cells, and construct a target matrix based on the spatial coordinates of the three-dimensional grid cells and the statistical characteristics, wherein the target matrix comprises autocorrelation matrices of the geotechnical parameters and cross-correlation matrices between the geotechnical parameters; an acquisition module configured to acquire a preset inversion model structure comprising an undetermined coefficient matrix, wherein the undetermined coefficient matrix comprises autocorrelation undetermined coefficient matrices and cross-correlation undetermined coefficient matrices, and the preset inversion model structure is configured to sequentially represent Gaussian random fields corresponding to the geotechnical parameters, a cross-correlation part of one of the Gaussian random fields is determined based on the cross-correlation undetermined coefficient matrices corresponding to one of the geotechnical parameters and independent random vectors used to generate a previous Gaussian random field, and an autocorrelation part of one of the Gaussian random fields is determined based on the autocorrelation undetermined coefficient matrix corresponding to one of the geotechnical parameters and a newly added independent random vector; an inversion module configured to determine the undetermined coefficient matrix based on the preset inversion model structure and the target matrix, and obtain an inversion model based on the undetermined coefficient matrix, wherein the undetermined coefficient matrix is configured to make the inversion model satisfy a preset constraint condition, and the preset constraint condition comprises that a self-covariance matrix of the Gaussian random fields generated by the inversion model is equal to the autocorrelation matrix corresponding to the geotechnical parameters, and a cross-covariance matrix between every two of the Gaussian random fields generated by the inversion model is equal to the cross-correlation matrix corresponding to the geotechnical parameters; a generation module configured to generate the independent random vectors in the inversion model, and input the independent random vectors into the inversion model to obtain target three-dimensional random fields corresponding to the geotechnical parameters.

[0014] The geotechnical parameter three-dimensional random field generation device provided by the application has basically the same advantages as the geotechnical parameter three-dimensional random field generation method of the prior art, and thus will not be described here.

[0015] The application further provides an electronic device comprising a memory and a processor. The memory is configured to store a computer program. The processor is configured to implement the geotechnical parameter three-dimensional random field generation method as described above when executing the computer program.

[0016] The electronic device provided by the present application has basically the same advantages as the rock-soil parameter three-dimensional random field generation method, which will not be repeated here.

[0017] The present application also provides a computer readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the rock-soil parameter three-dimensional random field generation method is realized.

[0018] The computer readable storage medium provided by the present application has basically the same advantages as the rock-soil parameter three-dimensional random field generation method, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The flowchart of the rock-soil parameter three-dimensional random field generation method of the embodiment of the present application is shown in the figure. Figure 2 The schematic diagram of the target three-dimensional random field generated by the embodiment of the present application is shown in the figure. Figure 3 The mathematical statistics diagram of the target three-dimensional random field of the embodiment of the present application is shown in the figure. Figure 4 The structural schematic diagram of the rock-soil parameter three-dimensional random field generation device of the embodiment of the present application is shown in the figure. Figure 5 The structural schematic diagram of the electronic device of the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to 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 realized in various forms, and should not be interpreted as being limited to the embodiments described herein, on the contrary, these embodiments are provided to make the present application more thorough and complete. It should be understood that the drawings and embodiments of the present application are only for illustrative purposes, and are not intended to limit the scope of protection of the present application.

[0021] It should be understood that each step described in the method embodiment of the present application can be executed in different order and / or in parallel. In addition, the method embodiment 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.

[0022] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based 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 "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0023] 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".

[0024] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for generating a three-dimensional random field of soil and rock parameters, which includes the following steps: S1: Obtain statistical information on multiple soil and rock parameters, and determine statistical characteristics based on the statistical information; among which, the statistical characteristics include the autocorrelation distance corresponding to each soil and rock parameter, and the cross-correlation coefficient between each soil and rock parameter.

[0025] Specifically, the soil and rock parameters referred to in this embodiment may include physical parameters such as soil and rock cohesion and internal friction angle, which can be set according to actual needs. The statistical information referred to in this embodiment can represent the data set obtained after testing soil and rock samples. The statistical features referred to in this embodiment represent key indicators that characterize the properties of soil and rock parameters based on statistical data, which may include the autocorrelation distance corresponding to each soil and rock parameter, and the cross-correlation coefficient between every two soil and rock parameters.

[0026] In one embodiment, after obtaining statistical data, the spatial distribution characteristics of soil and rock parameters can be fitted based on theoretical correlation functions (such as exponential or Gaussian correlation functions), and the autocorrelation distance corresponding to each soil and rock parameter can be determined using the least squares method. The cross-correlation coefficient between different soil and rock parameters can be determined by using statistical data to determine the covariance and standard deviation between the parameters, and then determining the cross-correlation coefficient based on the covariance and standard deviation. For example, in this embodiment, the cross-correlation coefficient satisfies: ; in, This represents the cross-correlation coefficient between soil / rock mass parameter a and soil / rock mass parameter b. This represents the covariance between soil / rock mass parameter a and soil / rock mass parameter b. This represents the standard deviation corresponding to parameter 'a' in the soil and rock mass. This represents the standard deviation corresponding to the soil and rock mass parameter b.

[0027] S2: Divide the preset simulation area into grids to obtain multiple three-dimensional grid units, and construct a target matrix based on the spatial coordinates and statistical characteristics of each three-dimensional grid unit; wherein, the target matrix includes the autocorrelation matrix of each soil and rock mass parameter, and the cross-correlation matrix between every two soil and rock mass parameters.

[0028] Specifically, in this embodiment, the preset simulation area refers to the target engineering space range for which three-dimensional random field modeling is required, which can be determined by engineering design drawings or on-site surveys. The three-dimensional mesh unit in this embodiment refers to the smallest spatial unit after discretizing the continuous simulation area, used to carry the random field value of soil and rock parameters at a single location. Its size and number can be preset according to the simulation accuracy. For example, assuming the preset simulation area is a 20m×20m×15m three-dimensional region, it can be discretized into 48,000 (i.e., 40×40×30) three-dimensional mesh units with a size of 0.5m×0.5m×0.5m. The spatial coordinates in this embodiment represent the unique coordinate value of the centroid of each three-dimensional mesh unit in the preset coordinate system. The target matrix in this embodiment includes the autocorrelation matrix of each soil and rock parameter, and the cross-correlation matrix between every two soil and rock parameters. An autocorrelation matrix describes the spatial autocorrelation characteristics of a single soil and rock parameter, and a cross-correlation matrix describes the spatial cross-correlation characteristics between two soil and rock parameters.

[0029] In one embodiment, the preset simulation area can be divided into regular grids (cubes, hexahedrons) or irregular grids (tetrahedrons, adaptive grids) based on its geometry and geological complexity, discretizing the continuous spatial domain into several three-dimensional grid units. These three-dimensional grid units can then be numbered, and the spatial coordinates of their centroids can be obtained to ensure the unique location of each unit in three-dimensional space. Furthermore, the distances between each three-dimensional grid unit and other three-dimensional grid units can be determined, and a symmetric positive definite autocorrelation matrix for each soil and rock mass parameter can be constructed based on the autocorrelation distance. Finally, a cross-correlation matrix can be obtained by multiplying the cross-correlation coefficients between the soil and rock mass parameters by the autocorrelation matrix.

[0030] S3: Obtain a preset inversion model structure including the undetermined coefficient matrix; wherein, the undetermined coefficient matrix includes the autocorrelation undetermined coefficient matrix and the cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to represent the Gaussian random fields corresponding to each soil and rock mass parameter in sequence; the cross-correlation part of a Gaussian random field is determined based on the cross-correlation undetermined coefficient matrices corresponding to a soil and rock mass parameter and the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to a soil and rock mass parameter and a newly added independent random vector.

[0031] Specifically, the undetermined coefficient matrix referred to in this embodiment includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix. The autocorrelation undetermined coefficient matrix characterizes the autocorrelation properties of the Gaussian random field corresponding to a single soil / rock mass parameter, while the cross-correlation undetermined coefficient matrix characterizes the cross-correlation properties between the Gaussian random fields corresponding to any two soil / rock mass parameters, and is a key parameter of the inversion model. In this embodiment, the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each soil / rock mass parameter; that is, the Gaussian random fields corresponding to each soil / rock mass parameter are constructed sequentially. Therefore, in this embodiment, the cross-correlation undetermined coefficient matrices corresponding to a soil / rock mass parameter include the cross-correlation undetermined coefficient matrices between that soil / rock mass parameter and the soil / rock mass parameters corresponding to the preceding Gaussian random fields.

[0032] In one embodiment, the cross-correlation component of a Gaussian random field is determined based on the undetermined cross-correlation coefficient matrices corresponding to a soil / rock mass parameter and the independent random vectors used to generate the previous Gaussian random field; the autocorrelation component of a Gaussian random field is determined based on the undetermined autocorrelation coefficient matrix corresponding to a soil / rock mass parameter and a newly added independent random vector. For example, suppose there are three soil / rock mass parameters, denoted as parameter a, parameter b, and parameter c. The structure of the inversion model is used to represent the Gaussian random fields corresponding to the three soil / rock mass parameters in sequence. For parameter a, since there are no preceding soil / rock mass parameters, the Gaussian random field corresponding to parameter a only includes the autocorrelation component, which can be represented based on the undetermined autocorrelation coefficient matrix corresponding to parameter a and the first independent random vector (denoted as vector 1 for ease of understanding and description). Further, the undetermined cross-correlation coefficient matrix corresponding to parameter b includes the undetermined cross-correlation coefficient matrices corresponding to parameter b and parameter a. Accordingly, the cross-correlation component of the Gaussian random field corresponding to parameter b is represented based on the cross-correlation undetermined coefficient matrix and the independent random vectors used to generate the previous Gaussian random field (i.e., vector 1). The autocorrelation component of the Gaussian random field corresponding to parameter b is represented based on the autocorrelation undetermined coefficient matrix corresponding to parameter b and the newly added independent random vector (denoted as vector 2 for ease of understanding and description). Based on this, the cross-correlation undetermined coefficient matrix corresponding to parameter c includes the cross-correlation undetermined coefficient matrices corresponding to parameters b and c (denoted as matrix bc for ease of understanding and description), and the cross-correlation undetermined coefficient matrices corresponding to parameters a and c (denoted as matrix ac for ease of understanding and description). Accordingly, the cross-correlation component of the Gaussian random field corresponding to parameter c can be represented based on matrix bc, matrix ac, and the independent random vectors used to generate the previous Gaussian random field (i.e., vector 1 and vector 2). The autocorrelation component of the Gaussian random field corresponding to parameter c is represented based on the autocorrelation undetermined coefficient matrix corresponding to parameter c and the newly added independent random vector (denoted as vector 3 for ease of understanding and description).

[0033] S4: Based on the preset inversion model structure and target matrix, determine the undetermined coefficient matrix, and obtain the inversion model based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to make the inversion model satisfy the preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between every two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix.

[0034] Specifically, the autocovariance matrix referred to in this embodiment describes the covariance relationship of a Gaussian random field corresponding to a soil / rock mass parameter at different spatial points (i.e., at different three-dimensional grid cells), reflecting the autocorrelation of the random field in space. The cross-covariance matrix referred to in this embodiment is used to describe the covariance relationship of two Gaussian random fields corresponding to two soil / rock mass parameters at different spatial points (i.e., at different three-dimensional grid cells), reflecting the mutual correlation characteristics between the two random fields.

[0035] In one embodiment, a preset inversion model structure describes the Gaussian random fields corresponding to each soil and rock mass parameter. After determining the undetermined coefficient matrix, the actual inversion model can be obtained. In practical application of the inversion model, the objective is that the statistical properties of the Gaussian random fields generated by the inversion model conform to the target matrix determined based on the statistical characteristics. Therefore, preset constraints can be determined based on this: the autocovariance matrix of the Gaussian random fields generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between any two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix. Given the preset inversion model structure and the target matrix, a solution expression is constructed based on the preset constraints, thereby obtaining the undetermined coefficient matrix using matrix operations. The interface then reverse-engineers the undetermined coefficient matrix that ensures the inversion model conforms to the preset constraints.

[0036] S5: Generate each independent random vector in the inversion model, and input each independent random vector into the inversion model to obtain the target three-dimensional random field corresponding to each soil and rock mass parameter.

[0037] Specifically, in this embodiment, the independent random vectors refer to independent random vectors that conform to a standard Gaussian distribution, which can be generated using methods such as Latin hypercube sampling. The generated independent random vectors are input into the inversion model, and according to the structure of the inversion model and the determined matrix of undetermined coefficients, the target three-dimensional random field corresponding to each soil and rock mass parameter can be obtained.

[0038] In this embodiment, by acquiring statistical information of multiple soil and rock parameters, the statistical characteristics such as the autocorrelation distance and cross-correlation coefficients between each soil and rock parameter are determined, which is beneficial for providing a basis for the joint modeling of three-dimensional random fields with multiple soil and rock parameters. Furthermore, this embodiment divides the preset simulation region into three-dimensional meshes. Based on the spatial coordinates and statistical characteristics of each three-dimensional mesh unit, a target matrix is ​​constructed containing the autocorrelation matrix of each soil and rock parameter and the cross-correlation matrix between every two soil and rock parameters, comprehensively characterizing the autocorrelation and cross-correlation characteristics of multiple soil and rock parameters in three-dimensional space. Based on this, a preset inversion model structure containing the autocorrelation undetermined coefficient matrix and the cross-correlation undetermined coefficient matrix (i.e., the undetermined coefficient matrix) is obtained, providing a reliable model framework for generating a Gaussian random field that fits the coupling characteristics of multiple soil and rock parameters. In this embodiment, the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each soil and rock mass parameter. The cross-correlation part of a Gaussian random field is determined based on the cross-correlation undetermined coefficient matrix corresponding to each soil and rock mass parameter and the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to each soil and rock mass parameter and a newly added independent random vector. Accordingly, in the actual application of the inversion model, the Gaussian random fields of each soil and rock mass parameter are constructed sequentially based on the inversion model. The cross-correlation part of the Gaussian random field shares independent random vectors with the preceding Gaussian random field, and the new independent random vectors introduced by the autocorrelation part of the Gaussian random field are shared with the subsequently constructed Gaussian random fields, thereby coupling the spatial dependencies between multiple soil and rock mass parameters. After obtaining the preset inversion model structure, this embodiment can determine the undetermined coefficient matrix that enables the inversion model to satisfy the preset constraints based on the preset inversion model structure and the target matrix. This embodiment uses the pre-defined constraint that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between any two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix. Combined with the pre-defined inversion model structure and the target matrix, the undetermined coefficient matrix can be solved in reverse, thus obtaining the inversion model. This helps ensure that the statistical characteristics of the Gaussian random field generated by the inversion model match the target matrix, thereby truly reflecting the spatial distribution characteristics of each soil and rock parameter and the spatial coupling influence between multiple soil and rock parameters. Thus, after generating each independent random vector in the inversion model, this embodiment inputs each independent random vector into the inversion model to obtain the target three-dimensional random field corresponding to each soil and rock parameter, thereby realizing the random field coupling modeling of multiple soil and rock parameters.

[0039] Optionally, based on the obtained spatial coordinates and statistical characteristics of each 3D mesh cell, including: The spatial distance between each three-dimensional grid cell is determined based on the spatial coordinates. Based on the spatial distance and the autocorrelation distance corresponding to the soil and rock mass parameters, the autocorrelation coefficient between each three-dimensional grid cell and other three-dimensional grid cells is determined, and the autocorrelation matrix corresponding to the soil and rock mass parameters is obtained. The cross-correlation matrix is ​​determined by multiplying the autocorrelation matrix and the corresponding cross-correlation coefficient.

[0040] Specifically, in this embodiment, the spatial distance refers to the actual distance between the centroids of two three-dimensional mesh elements in space, which can be determined by their corresponding spatial coordinates. The autocorrelation coefficient in this embodiment represents the degree of correlation between random variables of the same soil / rock mass parameter at two different spatial locations (i.e., three-dimensional mesh elements), and its magnitude is determined by the relative relationship between the spatial distance and the autocorrelation distance of the parameter. The autocorrelation matrix in this embodiment is a symmetric positive definite matrix composed of the pairwise autocorrelation coefficients of all three-dimensional mesh elements, with dimensions ne×ne (ne represents the total number of three-dimensional mesh elements).

[0041] In one embodiment, the autocorrelation coefficient between two three-dimensional mesh elements satisfies: ; Where ρ represents the autocorrelation coefficient; x, y, and z represent the x-axis, y-axis, and z-axis of the spatial coordinate system in which the preset simulation region is located; Represents the spatial distance along the x-axis. Represents the spatial distance along the y-axis. Represents the spatial distance along the z-axis; Represents the autocorrelation distance along the x-axis. Represents the autocorrelation distance along the y-axis. This represents the autocorrelation distance along the z-axis.

[0042] After determining the autocorrelation coefficients of the soil and rock parameters at the beginning of every two 3D grid cells, a symmetric positive definite matrix of dimension ne×ne can be constructed to obtain the autocorrelation matrix: ; in, This represents the autocorrelation matrix corresponding to the first soil / rock mass parameter; This represents the autocorrelation coefficient between the first 3D mesh element and the nth 3D mesh element.

[0043] Based on this, the cross-correlation matrix between the two soil and rock parameters can be obtained by multiplying the autocorrelation matrix by the corresponding cross-correlation coefficient. For example, in this embodiment, the cross-correlation matrix between the first and second soil and rock parameters satisfies: ; in, This represents the cross-correlation matrix between the first and second soil / rock mass parameters. This represents the cross-correlation coefficient between the first and second soil / rock mass parameters.

[0044] In this embodiment, the spatial distance between three-dimensional grids is determined by spatial coordinates, and the autocorrelation coefficient is determined by combining the autocorrelation distance of the soil and rock parameters, thus obtaining the autocorrelation matrix. This is beneficial for accurately quantifying the spatial correlation of the same soil and rock parameter between different three-dimensional grids. Each element in the autocorrelation matrix represents the autocorrelation coefficient between two corresponding three-dimensional grid units, comprehensively reflecting the spatial dependence of the same soil and rock parameter within the entire preset simulation area. Based on this, this embodiment determines the cross-correlation matrix between soil and rock parameters by multiplying the autocorrelation matrix and the corresponding cross-correlation coefficient. This incorporates the influence of location information on the cross-correlation of multiple parameters, so that the cross-correlation matrix not only considers the inherent relationship between different soil and rock parameters but also incorporates their spatial distribution characteristics. This is beneficial for reflecting the complex coupling relationship between multiple soil and rock parameters, thereby improving the realism and reliability of the subsequently generated three-dimensional random field.

[0045] Optionally, the preset inversion model structure satisfies: ; Where n represents the nth soil / rock mass parameter; Xn represents the Gaussian random field corresponding to the nth soil / rock mass parameter; and L represents the matrix of undetermined coefficients. This represents an independent random vector.

[0046] In one embodiment, a Gaussian random field with the first soil mass parameters is used. Taking the establishment of Gaussian random fields as an example, its Gaussian random field can be constructed as follows: ; in, This represents the Gaussian random field corresponding to the first soil and rock mass parameter, k represents the generated k-th Gaussian random field, and G represents the type of random field as Gaussian random field; This represents the autocorrelation undetermined coefficient matrix corresponding to the first soil and rock mass parameter, which is an unknown matrix. This represents the independent random vector introduced when constructing the Gaussian random field corresponding to the first soil and rock mass parameters. It follows a standard normal distribution and can be obtained through random sampling. After determining the autocorrelation matrix corresponding to the soil and rock mass parameters, the spatial randomness of the subsequent random fields can be determined. Since it conforms to a real symmetric positive definite matrix, a Choreski decomposition is performed to determine its overall structure. ; in, This represents the autocovariance matrix of the Gaussian random field corresponding to the first soil and rock mass parameters at each three-dimensional grid cell location; This represents the autocorrelation matrix of undetermined coefficients corresponding to the first soil / rock mass parameter. express The transpose of .

[0047] Based on this, for random fields of multiple soil and rock parameters, in order to fully express the autocorrelation and cross-correlation of soil and rock parameters, sharing can be used. The cross-correlation undetermined coefficient matrix is ​​used to establish the cross-correlation between soil and rock parameters, and a new independent random vector is introduced. The autocorrelation is expressed by the autocorrelation undetermined coefficient matrix corresponding to the soil and rock mass parameters. In this embodiment, the Gaussian random field corresponding to the second soil and rock mass parameter can be constructed as follows: ; in, This represents the Gaussian random field corresponding to the second soil and rock mass parameter; This represents the undetermined coefficient matrix of the cross-correlation between the first and second soil parameters, and is an unknown matrix. This represents the autocorrelation undetermined coefficient matrix corresponding to the second soil and rock mass parameter, which is an unknown matrix.

[0048] Therefore, the preset inversion model structure used in this embodiment to sequentially express the Gaussian random field corresponding to n soil and rock parameters can be constructed as follows: ; Where n represents the nth soil and rock mass parameter; Xn represents the Gaussian random field corresponding to the nth soil and rock mass parameter; Represents an independent random vector. L represents the newly added independent random vector when constructing the Gaussian random field of the nth soil and rock mass parameter; L represents the undetermined coefficient matrix (including the autocorrelation undetermined coefficient matrix and the cross-correlation undetermined coefficient matrix). This represents the undetermined cross-correlation coefficient matrix between the nth soil / rock parameter and the first soil / rock parameter. This represents the autocorrelation undetermined coefficient matrix corresponding to the nth soil / rock parameter.

[0049] Optionally, the preset constraints satisfy: ; in, The cross-correlation matrix between the first soil / rock mass parameter and the nth soil / rock mass parameter is equal to the cross-covariance matrix between the corresponding Gaussian random fields. The autocorrelation matrix corresponding to the nth soil parameter is equal to the autocovariance matrix of the corresponding Gaussian random field. ne represents the undetermined coefficients corresponding to the three-dimensional mesh element; ne represents the number of three-dimensional mesh elements.

[0050] Specifically, in practical applications of the inversion model, it is expected that the autocovariance matrix of the Gaussian random field of the soil and rock parameters output by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between any two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix. These can be used as preset constraints. In this embodiment, the autocorrelation matrix of the first soil and rock parameter satisfies: ; in, The autocorrelation matrix corresponding to the first soil and rock mass parameter is equal to the autocovariance matrix of the corresponding Gaussian random field. This represents the Gaussian random field corresponding to the first soil / rock mass parameter; T represents the transpose; E represents the mathematical expectation operation.

[0051] Based on this, the cross-correlation matrix of the second soil and rock mass parameters satisfies: ; in, The cross-correlation matrix between the first and second soil parameters is equal to the cross-covariance matrix between the corresponding Gaussian random fields. This represents the Gaussian random field corresponding to the second soil and rock mass parameter; T represents the transpose.

[0052] Furthermore, the autocorrelation matrix corresponding to the second soil and rock mass parameters satisfies: ; in, The autocorrelation matrix corresponding to the second soil / rock mass parameter is equal to the autocovariance matrix of the corresponding Gaussian random field. Based on the above derivation, the preset constraints that the preset inversion model structure corresponding to the two soil / rock mass parameters needs to satisfy are: ; Based on this, the preset constraints that the inversion model structure corresponding to n soil and rock parameters needs to satisfy are as follows: ; in, The cross-correlation matrix between the first soil / rock mass parameter and the nth soil / rock mass parameter is equal to the cross-covariance matrix between the corresponding Gaussian random fields. The autocorrelation matrix corresponding to the nth soil and rock parameter is equal to the autocovariance matrix of its corresponding Gaussian random field. ne represents the undetermined coefficients corresponding to the three-dimensional mesh element; ne represents the number of three-dimensional mesh elements.

[0053] Optionally, based on a preset inversion model structure and target matrix, the undetermined coefficient matrix is ​​determined, including: The autocorrelation matrix of the soil and rock parameters corresponding to the first Gaussian random field represented by the preset inversion model structure is decomposed to obtain the autocorrelation undetermined coefficient matrix corresponding to the soil and rock parameters. According to the representation order of each Gaussian random field in the preset inversion model structure, the solution operation is performed on the soil and rock parameters corresponding to the remaining Gaussian random fields in sequence. The solution operation includes determining the undetermined coefficient matrix corresponding to the soil and rock parameter based on the target matrix corresponding to the soil and rock parameter and the determined undetermined coefficient matrix of each parameter.

[0054] Specifically, based on the preset constraints, it is known that the autocovariance matrix of the Gaussian random field corresponding to the first soil parameter is equal to the autocorrelation matrix corresponding to that soil parameter. Therefore, the relationship between the autocorrelation matrix and the autocorrelation undetermined coefficient matrix satisfies: ; By decomposing the autocorrelation matrix, the autocorrelation coefficient matrix corresponding to the soil and rock parameters can be determined: ; Based on this, since the Gaussian random fields corresponding to each soil and rock mass parameter in the preset inversion model structure are constructed sequentially, when actually solving each undetermined coefficient matrix, it is necessary to perform the solution operation on the soil and rock mass parameters corresponding to the remaining Gaussian random fields in sequence according to the representation order of each Gaussian random field in the preset inversion model structure, so as to solve the target matrix corresponding to the soil and rock mass parameters and the determined related undetermined coefficient matrices.

[0055] Optionally, based on the target matrix corresponding to the soil and rock mass parameters and the already determined respective related undetermined coefficient matrices, the undetermined coefficient matrix corresponding to the soil and rock mass parameters is determined, including: Based on the cross-correlation matrices corresponding to the soil and rock parameters, and the determined cross-correlation undetermined coefficient matrices, determine the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters. Subtract the covariance terms corresponding to each cross-correlation undetermined coefficient matrix corresponding to the soil and rock parameters from the autocorrelation matrix corresponding to the soil and rock parameters to obtain the residual covariance matrix. Then decompose the residual covariance matrix to obtain the autocorrelation undetermined coefficient matrix corresponding to the soil and rock parameters. Each covariance term includes the product of a cross-correlation undetermined coefficient matrix and the corresponding transpose of the cross-correlation undetermined coefficient matrix.

[0056] Specifically, taking the preset inversion model structure corresponding to two soil and rock mass parameters as an example, after determining the autocorrelation undetermined coefficient matrix corresponding to the first soil and rock mass parameter, it can be substituted into the preset constraints, and based on the cross-correlation matrix between the first and second soil and rock mass parameters, the cross-correlation undetermined coefficient matrix corresponding to the second soil and rock mass parameter can be obtained: ; because: ; We can obtain: ; therefore: ; Therefore, based on the cross-correlation matrix corresponding to the second soil and rock mass parameter and the determined autocorrelation undetermined coefficient matrix, the cross-correlation undetermined coefficient matrix corresponding to the soil and rock mass parameter can be determined.

[0057] Based on this, since the autocovariance of the Gaussian random field corresponding to the soil and rock parameters in the preset constraints includes covariance terms contributed by the cross-correlation undetermined coefficient matrix and covariance terms contributed by the autocorrelation undetermined coefficients, and the autocovariance matrix of the soil and rock parameters is equal to the autocorrelation matrix corresponding to the soil and rock parameters, after determining the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters, subtracting the covariance terms corresponding to the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters from the autocovariance matrix (i.e., the autocorrelation matrix) corresponding to the soil and rock parameters yields the covariance terms contributed by the autocorrelation undetermined coefficients of the soil and rock parameters, thus obtaining the remaining covariance matrix. Taking the preset inversion model structure corresponding to two soil and rock parameters as an example, based on the preset constraints, it can be seen that after obtaining the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters, the remaining covariance matrix satisfies: ; Performing a Cholliski decomposition on it yields the autocorrelation matrix of undetermined coefficients corresponding to the second soil and rock parameters, which satisfies the following: ; Thus, by sequentially performing the solution operation on the soil and rock parameters corresponding to the remaining Gaussian random fields, the autocorrelation undetermined coefficient matrix and the cross-correlation undetermined coefficient matrix corresponding to each soil and rock parameter can be obtained in sequence, thereby determining the undetermined coefficient matrix of the constrained inversion structure and obtaining the inversion model.

[0058] In this embodiment, after determining the autocorrelation undetermined coefficient matrix of the soil and rock parameters corresponding to the first Gaussian random field, the cross-correlation undetermined coefficient matrix of the subsequent soil and rock parameters is determined by combining it with the cross-correlation matrix between the first Gaussian random field and the subsequent soil and rock parameters. This is beneficial for establishing the cross-dependency relationship between different soil and rock parameters. Based on this, according to the representation order of each Gaussian random field in the preset inversion model structure, the solution operation is sequentially performed on the soil and rock parameters corresponding to the remaining Gaussian random fields. This is beneficial for fully utilizing the characteristics of the preset inversion model structure and the determined undetermined coefficient matrix. This embodiment determines the undetermined coefficient matrix of each parameter in the preset inversion model structure sequentially through sequential decomposition and recursive solution. This is beneficial for ensuring that the Gaussian random fields with multiple soil and rock parameters generated by the subsequent inversion model strictly meet the preset constraints while following the preset inversion model structure.

[0059] Optionally, each independent random vector is input into the inversion model to obtain the target three-dimensional random field corresponding to each soil and rock mass parameter, including: Each independent random vector is input into the inversion model to obtain the three-dimensional random field corresponding to each soil and rock mass parameter. Then, each three-dimensional random field is transformed with equal probability to obtain the three-dimensional random field of each target.

[0060] In this embodiment, after inputting each independent random vector into the inversion model to obtain the three-dimensional random fields corresponding to each soil and rock mass parameter, each three-dimensional random field can be transformed with equal probability to adjust the probability distribution of the generated three-dimensional random field to conform to the actual probability distribution of the soil and rock mass parameters, thereby obtaining the target three-dimensional random field and achieving effective characterization of spatial variability. For example, assuming that the soil and rock mass parameters in this embodiment include cohesion and friction angle, after obtaining their corresponding three-dimensional random fields, in order to satisfy the non-negativity of the soil and rock mass parameters, they can be transformed with equal probability to convert them into a log-normal distribution. Taking soil and rock mass parameters including cohesion and friction angle as an example, a schematic diagram of the target three-dimensional random field generated in this embodiment is shown below. Figure 2 As shown, Figure 2 In the diagram, 2a represents the target three-dimensional random field corresponding to cohesion, and 2b represents the target three-dimensional random field corresponding to friction angle. Cohesion represents cohesion, and friction angle represents friction angle. Profile 1 represents profile 1, and Profile 2 represents profile 2. Distance represents distance, and Depth represents depth.

[0061] Optionally, the statistical features also include the mean and variance corresponding to each soil and rock mass parameter. After obtaining the target three-dimensional random field corresponding to each soil and rock mass parameter, the current mean and standard deviation of the target three-dimensional random field can also be determined and compared with the mean and standard deviation corresponding to each soil and rock mass parameter determined based on statistical data to verify the reliability of the three-dimensional random field. Taking soil and rock mass parameters including cohesion and friction angle as an example, the mathematical statistics diagram corresponding to the target three-dimensional random field constructed in this embodiment is as follows: Figure 3 As shown, Figure 3 In the diagram, 3a represents the mathematical statistics of the three-dimensional random field corresponding to cohesion, and 3b represents the mathematical statistics of the three-dimensional random field corresponding to the friction angle. μ1 represents the mean, μ2 represents the current mean; σ1 represents the standard deviation, and σ2 represents the current standard deviation. Based on Figure 3 The mathematical statistics show that the current mean values ​​of the two soil and rock mass parameters simulated in the three-dimensional random field simulation are basically consistent with the mean values ​​of the two soil and rock mass parameters determined based on statistical data. Similarly, the current standard deviations of the two soil and rock mass parameters simulated in the three-dimensional random field simulation are also basically consistent with the current standard deviations of the two soil and rock mass parameters determined based on statistical data.

[0062] like Figure 4 As shown, an embodiment of the present invention provides a three-dimensional random field generation device 200 for soil and rock parameters, comprising: The statistics module 410 is used to acquire statistical information of multiple soil and rock parameters and determine statistical features based on the statistical information; wherein, the statistical features include the autocorrelation distance corresponding to each soil and rock parameter and the cross-correlation coefficient between each soil and rock parameter; The partitioning module 420 is used to partition a preset simulation area into multiple three-dimensional grid units, and to construct a target matrix based on the spatial coordinates of each three-dimensional grid unit and the statistical characteristics; wherein, the target matrix includes the autocorrelation matrix of each soil and rock mass parameter, and the cross-correlation matrix between every two soil and rock mass parameters. The acquisition module 430 is used to acquire a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each of the soil and rock parameters; the cross-correlation part of a Gaussian random field is determined based on the cross-correlation undetermined coefficient matrices corresponding to each of the soil and rock parameters and the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to each of the soil and rock parameters and a newly added independent random vector; The inversion module 440 is used to determine the undetermined coefficient matrix based on the preset inversion model structure and the target matrix, and to obtain the inversion model based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to make the inversion model satisfy preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between any two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; The generation module 450 is used to generate each of the independent random vectors in the inversion model, and input each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters.

[0063] The three-dimensional random field generation device for soil and rock parameters provided in this embodiment has basically the same technical effect as the three-dimensional random field generation method for soil and rock parameters, and will not be described in detail here.

[0064] like Figure 5 As shown, an electronic device 500 provided in this embodiment of the invention includes a memory 510 and a processor 520; the memory 510 is used to store a computer program; the processor 520 is used to implement the three-dimensional random field generation method for soil and rock parameters as described above when the computer program is executed.

[0065] Alternatively, an electronic device 500 includes a memory 510 and a processor 520 coupled to the memory 510; the memory 510 is configured to store a computer program; and the processor 520 is configured to perform the following operations when the computer program is executed: Statistical information of multiple soil and rock mass parameters is obtained, and statistical features are determined based on the statistical information; wherein, the statistical features include the autocorrelation distance corresponding to each soil and rock mass parameter, and the cross-correlation coefficient between each soil and rock mass parameter; The preset simulation area is divided into grids to obtain multiple three-dimensional grid cells. Based on the obtained spatial coordinates of each three-dimensional grid cell and the statistical characteristics, a target matrix is ​​constructed. The target matrix includes the autocorrelation matrix of each soil and rock mass parameter and the cross-correlation matrix between every two soil and rock mass parameters. Obtain a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each of the soil and rock parameters; the cross-correlation part of a Gaussian random field is determined based on each of the cross-correlation undetermined coefficient matrices corresponding to a soil and rock parameter and each of the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to a soil and rock parameter and a newly added independent random vector; Based on the preset inversion model structure and the target matrix, the undetermined coefficient matrix is ​​determined, and the inversion model is obtained based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to make the inversion model satisfy preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between every two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; Generate each of the independent random vectors in the inversion model, and input each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters.

[0066] The electronic device provided in this embodiment and the three-dimensional random field generation method for soil and rock parameters can produce basically the same technical effects, and will not be described again here.

[0067] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the three-dimensional random field generation method for soil and rock parameters as described above.

[0068] Alternatively, a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the following operations: Statistical information of multiple soil and rock mass parameters is obtained, and statistical features are determined based on the statistical information; wherein, the statistical features include the autocorrelation distance corresponding to each soil and rock mass parameter, and the cross-correlation coefficient between each soil and rock mass parameter; The preset simulation area is divided into grids to obtain multiple three-dimensional grid cells. Based on the obtained spatial coordinates of each three-dimensional grid cell and the statistical characteristics, a target matrix is ​​constructed. The target matrix includes the autocorrelation matrix of each soil and rock mass parameter and the cross-correlation matrix between every two soil and rock mass parameters. Obtain a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each of the soil and rock parameters; the cross-correlation part of a Gaussian random field is determined based on each of the cross-correlation undetermined coefficient matrices corresponding to a soil and rock parameter and each of the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to a soil and rock parameter and a newly added independent random vector; Based on the preset inversion model structure and the target matrix, the undetermined coefficient matrix is ​​determined, and the inversion model is obtained based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to make the inversion model satisfy preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between every two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; Generate each of the independent random vectors in the inversion model, and input each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters.

[0069] The computer-readable storage medium provided in this embodiment and the three-dimensional random field generation method for soil and rock parameters can produce basically the same technical effects, and will not be described again here.

[0070] Electronic device 500, which can serve as a server or client of the present invention, is described below as an example of a hardware device applicable to various aspects of the present invention. Electronic device 500 is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 500 can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0071] Electronic device 500 includes a computing unit that can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) or a computer program loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The computing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0072] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. In this application, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention according to actual needs. Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units can be implemented in hardware or as software functional units.

[0073] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for generating a three-dimensional random field of soil and rock parameters, characterized in that, include: Statistical information of multiple soil and rock mass parameters is obtained, and statistical features are determined based on the statistical information; wherein, the statistical features include the autocorrelation distance corresponding to each soil and rock mass parameter, and the cross-correlation coefficient between each soil and rock mass parameter; The preset simulation area is divided into grids to obtain multiple three-dimensional grid cells. Based on the obtained spatial coordinates of each three-dimensional grid cell and the statistical characteristics, a target matrix is ​​constructed. The target matrix includes the autocorrelation matrix of each soil and rock mass parameter and the cross-correlation matrix between every two soil and rock mass parameters. Obtain a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each of the soil and rock parameters; the cross-correlation part of a Gaussian random field is determined based on each of the cross-correlation undetermined coefficient matrices corresponding to a soil and rock parameter and each of the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to a soil and rock parameter and a newly added independent random vector; Based on the preset inversion model structure and the target matrix, the undetermined coefficient matrix is ​​determined, and the inversion model is obtained based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to make the inversion model satisfy preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between every two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; Generate each of the independent random vectors in the inversion model, and input each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters.

2. The method for generating three-dimensional random fields of soil and rock parameters according to claim 1, characterized in that, The acquisition of the spatial coordinates of each of the three-dimensional mesh units and the statistical features include: Based on the spatial coordinates, the spatial distance between each of the three-dimensional mesh units is determined, and based on the spatial distance and the autocorrelation distance corresponding to the soil and rock parameters, the autocorrelation coefficient between each of the three-dimensional mesh units and other three-dimensional mesh units is determined, thereby obtaining the autocorrelation matrix corresponding to the soil and rock parameters; The cross-correlation matrix is ​​determined by multiplying the autocorrelation matrix and the corresponding cross-correlation coefficient.

3. The method for generating three-dimensional random fields of soil and rock parameters according to claim 1, characterized in that, The preset inversion model structure satisfies: ； Wherein, n represents the nth soil / rock mass parameter; Xn represents the Gaussian random field corresponding to the nth soil / rock mass parameter; and L represents the undetermined coefficient matrix. This represents the independent random vector.

4. The method for generating three-dimensional random fields of soil and rock parameters according to claim 1, characterized in that, The preset constraint conditions are satisfied as follows: ； in, The cross-correlation matrix between the first soil / rock parameter and the nth soil / rock parameter is equal to the cross-covariance matrix between the corresponding Gaussian random fields. The autocorrelation matrix corresponding to the nth soil and rock parameter is equal to the autocovariance matrix of the corresponding Gaussian random field; ne represents the undetermined coefficient corresponding to the three-dimensional mesh unit; ne represents the number of the three-dimensional mesh units.

5. The method for generating a three-dimensional random field of soil and rock parameters according to claim 1, characterized in that, The step of determining the undetermined coefficient matrix based on the preset inversion model structure and the target matrix includes: The autocorrelation matrix of the soil and rock mass parameters corresponding to the first Gaussian random field represented by the preset inversion model structure is decomposed to obtain the autocorrelation undetermined coefficient matrix corresponding to the soil and rock mass parameters; According to the representation order of each Gaussian random field in the preset inversion model structure, the remaining soil and rock parameters corresponding to each Gaussian random field are solved sequentially. The solution operation includes determining the undetermined coefficient matrix corresponding to the soil and rock parameter based on the target matrix corresponding to the soil and rock parameter and the determined autocorrelation undetermined coefficient matrices.

6. The method for generating three-dimensional random fields of soil and rock parameters according to claim 4, characterized in that, The step of determining the undetermined coefficient matrix corresponding to the soil and rock mass parameter based on the target matrix corresponding to the soil and rock mass parameter and the determined autocorrelation undetermined coefficient matrices includes: Based on the cross-correlation matrices corresponding to the soil and rock parameters, and the determined autocorrelation undetermined coefficient matrices, determine the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters. Subtract the covariance terms corresponding to each of the cross-correlation undetermined coefficient matrices corresponding to the soil and rock parameters from the autocorrelation matrix corresponding to the soil and rock parameters to obtain the residual covariance matrix. Then, decompose the residual covariance matrix to obtain the autocorrelation undetermined coefficient matrix corresponding to the soil and rock parameters. Each covariance term includes the product of a cross-correlation undetermined coefficient matrix and its corresponding transpose.

7. The method for generating a three-dimensional random field of soil and rock parameters according to claim 1, characterized in that, The step of inputting each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters includes: Each independent random vector is input into the inversion model to obtain a three-dimensional random field corresponding to each soil and rock mass parameter. Each three-dimensional random field is then transformed with equal probability to obtain a target three-dimensional random field.

8. A three-dimensional random field generation device for soil and rock parameters, characterized in that, include: The statistical module is used to acquire statistical information of multiple soil and rock parameters and determine statistical features based on the statistical information; wherein, the statistical features include the autocorrelation distance corresponding to each soil and rock parameter and the cross-correlation coefficient between each soil and rock parameter; The partitioning module is used to partition a preset simulation area into multiple three-dimensional grid cells, and to construct a target matrix based on the spatial coordinates of each three-dimensional grid cell and the statistical characteristics; wherein, the target matrix includes the autocorrelation matrix of each soil and rock mass parameter, and the cross-correlation matrix between every two soil and rock mass parameters; An acquisition module is used to acquire a preset inversion model structure including an undetermined coefficient matrix; wherein the undetermined coefficient matrix includes an autocorrelation undetermined coefficient matrix and a cross-correlation undetermined coefficient matrix; the preset inversion model structure is used to sequentially represent the Gaussian random fields corresponding to each of the soil and rock parameters; the cross-correlation part of a Gaussian random field is determined based on the cross-correlation undetermined coefficient matrices corresponding to each of the soil and rock parameters and the independent random vectors used to generate the previous Gaussian random field; the autocorrelation part of a Gaussian random field is determined based on the autocorrelation undetermined coefficient matrix corresponding to each of the soil and rock parameters and a newly added independent random vector; An inversion module is used to determine the undetermined coefficient matrix based on the preset inversion model structure and the target matrix, and to obtain an inversion model based on the undetermined coefficient matrix; wherein, the undetermined coefficient matrix is ​​used to ensure that the inversion model satisfies preset constraints; the preset constraints include that the autocovariance matrix of the Gaussian random field generated by the inversion model is equal to the corresponding autocorrelation matrix, and the cross-covariance matrix between any two Gaussian random fields generated by the inversion model is equal to the corresponding cross-correlation matrix; The generation module is used to generate each of the independent random vectors in the inversion model, and input each of the independent random vectors into the inversion model to obtain the target three-dimensional random field corresponding to each of the soil and rock parameters.

9. An electronic device, characterized in that, 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 generating three-dimensional random fields of soil and rock parameters as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method for generating three-dimensional random fields of soil and rock parameters as described in any one of claims 1 to 7.

Images