Method and system for analyzing spatial variability of foundation rheological parameters of covering layer

The method and system for analyzing spatial variability of overburden foundation parameters in rockfill dams using Markov chain theory and finite element models improve the accuracy and reliability of predicting long-term deformation by optimizing mesh size and considering spatial variability, enhancing safety and stability.

CN119089744APending Publication Date: 2024-12-06XIAN UNIV OF TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411202930.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Current technologies lack accurate prediction of spatial variability of overburden lithology parameters and rheological analysis methods for overburden rockfill dams, which affects the safety and stability of concrete face rockfill dams due to unpredictable deformation and potential cracking.

Method used

A method using the Markov chain theory to predict geological lithology distribution based on drilling information, combined with a three-dimensional finite element model and mesh optimization, and the correlation between rheological and material parameters, to analyze spatial variability and optimize mesh size for improved deformation analysis.

Benefits of technology

Enhances the accuracy and reliability of predicting long-term deformation in rockfill dams by considering spatial variability, providing a theoretical basis for construction and improving safety by reducing inaccurate parameter predictions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089744A_ABST
    Figure CN119089744A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of covering layer foundation rheological parameter analysis, in particular to a covering layer foundation rheological parameter spatial variability analysis method and system.The method comprises the steps that a rock-fill dam three-dimensional finite element model is established based on covering layer rock-fill dam stratum data, and the optimal network size is determined through a grid size optimization algorithm; according to the optimal grid size, area division is conducted on the covering layer rock-fill dam; according to actual measurement data of the drill holes, rock stratum characteristics in the area units where the drill holes are located are predicted through the Markov chain algorithm, and corresponding material parameters are obtained according to the rock stratum characteristics of all the area units; establishing a correlation between the rheological parameter and each material parameter, and obtaining a related standard normal distribution random field of the rheological parameter of the rock-fill dam foundation material; and inputting the rheological parameters in each area unit into the three-dimensional finite element model of the overburden rockfill dam, and carrying out rheological analysis to obtain a displacement deformation result corresponding to each area unit. The method can predict the lithologic parameter distribution of the covering layer more accurately.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of rheological parameter analysis of overburden foundation, and in particular to a method and system for analyzing spatial variability of rheological parameters of overburden foundation. Background Art

[0002] Concrete face rockfill dam built on the overburden has the advantages of simplifying construction diversion, shortening construction period and saving investment. It has been widely used in the anti-seepage system of engineering dams, and the stress and deformation control of the anti-seepage system is the key to the project. The anti-seepage body of the face rockfill dam built on the overburden is deformed due to the compression deformation of the foundation. The deformation of the anti-seepage body of the face rockfill dam mainly depends on the deformation of the rockfill body. If the deformation of the rockfill body is too large, cracks will appear on the face plate, thus affecting its anti-seepage performance and even endangering the stability of the dam body. Rheological deformation, as an important component of the deformation of the face rockfill dam, is the main reason for the long-term deformation of the dam and the failure of the anti-seepage body. Due to the effects of sedimentation, historical load conditions, physical weathering process and natural transportation, the physical and mechanical parameters of natural soil show spatial variability, and the spatial variability of parameters is an inherent property of soil. The riverbed overburden is complex in type and loose in structure, and its physical and mechanical properties show large inhomogeneity, and the material parameters are random. The spatial variability of the material parameters of the overburden has an important influence on the long-term deformation of the foundation and the rockfill dam, which may affect the safety of the dam and the anti-seepage structure.

[0003] In terms of the influence of parameter spatial variability on the properties of rockfill dams, the current research at home and abroad mainly focuses on the influence of the spatial variability of the mechanical parameters of the dam body or foundation on the slope stability and static and dynamic response. The current technology lacks an accurate prediction of the distribution of overburden lithology parameters and a reasonable and feasible rheological analysis method for overburden rockfill dams that considers parameter spatial variability. The development of related technology is crucial to evaluate the influence of the overburden foundation on the mechanical properties of the rockfill dam and thus control the safety of the dam structure. Summary of the invention

[0004] The present invention provides a method and system for analyzing the spatial variability of rheological parameters of overburden foundation. The Markov chain theory is used to predict the geological lithology distribution in the engineering area through drilling information, revealing the influence of the spatial variability of overburden materials on the long-term deformation of the rockfill dam, providing a theoretical basis and technical support for the construction of the rockfill dam, and solving the problem of inaccurate prediction of the distribution of overburden lithology parameters.

[0005] The objective of the present invention is achieved by the following technical solutions:

[0006] A method for analyzing spatial variability of rheological parameters of overburden foundation, comprising:

[0007] Obtaining the stratigraphic information of the overburden rockfill dam and the borehole measurement data;

[0008] A three-dimensional finite element model of the rockfill dam was established based on the stratigraphic data of the overburden rockfill dam, and the optimal network size was determined using a mesh size optimization algorithm. The overburden rockfill dam was divided into regions based on the optimal mesh size.

[0009] According to the measured data of the borehole, the Markov chain algorithm is used to predict the rock formation characteristics in the regional unit where the borehole is located, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit;

[0010] The correlation between rheological parameters and material parameters is established, and the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials is obtained;

[0011] The rheological parameters in each regional unit are input into the three-dimensional finite element model of the overburden rockfill dam, and rheological analysis is performed to obtain the displacement deformation results corresponding to each regional unit.

[0012] As a further improvement of the present invention, the Markov chain algorithm is used to predict the rock formation characteristics in the regional unit where the borehole is located based on the measured data of the borehole, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit, specifically including:

[0013] Using the measured data of boreholes, the Markov chain algorithm is used to estimate the regional unit of each borehole, and the horizontal state transfer count matrix and the vertical state transfer count matrix are obtained.

[0014] A transfer probability matrix in the horizontal direction is obtained based on the state transfer count matrix in the horizontal direction, and a corresponding transfer probability matrix is ​​obtained based on the state transfer count matrix in the vertical direction;

[0015] According to the horizontal and vertical transfer probability matrices, the occurrence probability of the soil state in a single borehole is obtained;

[0016] The soil state of each regional unit is obtained by coupling the data of each single borehole, and the rock formation characteristic information is predicted based on the soil state. The material parameters are obtained based on the rock formation characteristic information, and the material parameters include density, elastic modulus and three rheological parameters.

[0017] As a further improvement of the present invention, the method of obtaining a horizontal transition probability matrix based on a horizontal state transition count matrix and obtaining a corresponding state transition probability matrix based on a vertical state transition count matrix specifically includes:

[0018] The state transition probability matrix is ​​obtained by using the elements in the horizontal state transition count matrix and the sum of each element. The calculation expression is:

[0019]

[0020] In the formula, is the horizontal transition probability matrix, is the element in the horizontal state transition count matrix, Represented by the first state S l Transition to the second state S k The number of transfers, is the sum of the elements in the horizontal state transition count matrix;

[0021] The state transition probability matrix is ​​obtained by using the elements in the vertical state transition count matrix and the sum of each element. The calculation expression is:

[0022]

[0023] In the formula, is the vertical state transition probability matrix, is the element in the vertical state transition count matrix, Represented by the first state S l Transition to the second state S k The number of transfers, is the sum of the elements in the vertical state transition count matrix.

[0024] As a further improvement of the present invention, after obtaining the corresponding material parameters according to the rock formation characteristics of each regional unit, it also includes the step of optimizing the material parameters using a simulation model, and the specific steps are:

[0025] The linear elastic model and the DP plastic model are used to simulate the elastic-plastic properties of the granular material in the rockfill dam to obtain the plastic yield surface function, wherein the granular material is obtained based on the rock formation property information;

[0026] The time-hardening creep law is used to simulate the rheological characteristics of the rockfill dam to obtain the rheological parameters. Based on the characteristic values ​​of each material parameter, the probability distribution corresponding to each material parameter is obtained. The rheological parameter grouping interval is adjusted according to the probability distribution to obtain the optimal probability distribution type. The characteristic values ​​include mean, standard deviation, and coefficient of variation.

[0027] As a further improvement of the present invention, the establishment of the correlation between the rheological parameters and the material parameters to obtain the relevant standard normal distribution random field of the rheological parameters of the rockfill dam foundation material specifically includes:

[0028] The correlation distance of each regional unit is calculated based on the node coordinates of the geological unit of the overburden layer, and the autocorrelation coefficient matrix corresponding to each material parameter is obtained based on the correlation distance, and the correlation distance is used to describe the correlation between soil bodies in space;

[0029] The autocorrelation coefficient matrix and the corresponding position coordinates are decomposed by Cholesky to obtain the autocorrelation standard normal random field of each base position;

[0030] Constructing a mutual correlation coefficient matrix and a cross-correlation matrix between the parameters, and combining the autocorrelation matrix of each parameter to form a correlated normal distribution random field of the parameter, wherein the autocorrelation matrix is ​​obtained by Cholesky decomposition of the autocorrelation standard normal random field of each material parameter;

[0031] The correlation coefficient matrix is ​​subjected to Cholesky decomposition to obtain a lower triangular matrix, and a standard normal random field of rheological parameter cross-correlation is constructed based on the lower triangular matrix and the cross-correlation matrix.

[0032] As a further improvement of the present invention, the relevant distance is:

[0033]

[0034] Where D(Δz) is the correlation function, Δz is the sampling interval, h is the sample interval (a multiple of the sampling interval), Γ 2 (h) is the variance reduction function.

[0035] As a further improvement of the present invention, constructing a rheological parameter mutual correlation standard normal random field according to the lower triangular matrix and the mutual correlation matrix includes: the rheological parameter mutual correlation standard normal random field is:

[0036]

[0037] In the formula, is the cross-correlation standard normal random field of rheological parameters, B i is the parameter of the o-dimensional standard normal distribution random variable, L is the lower triangular matrix obtained by decomposing the autocorrelation coefficient, L C T is the transpose of the lower triangular matrix obtained by decomposing the cross-correlation coefficient matrix, i = ρ, E, A, n, m.

[0038] As a further improvement of the present invention, the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit, and also includes using Latin hypercube sampling method to extract random samples from the normal distribution of different materials in the rockfill dam to obtain a material random parameter library.

[0039] As a further improvement of the present invention, the rheological parameters in each regional unit are input into the three-dimensional finite element model of the overburden rockfill dam, and the rheological parameters are corrected. The specific steps are:

[0040] Establish a Kriging model and train it using the material random parameter library;

[0041] The trained Kriging model is used for random sampling, and the digital eigenvalues ​​of the correction parameters are estimated according to the central limit theorem. The rheological parameters are corrected using the digital eigenvalues ​​of the correction parameters, and rheological calculations are performed based on the corrected rheological parameters.

[0042] In a second aspect, the present invention further provides a system for analyzing the spatial variability of rheological parameters of an overburden foundation, which is used to implement a method for analyzing the spatial variability of rheological parameters of an overburden foundation, comprising:

[0043] Data acquisition module, used to obtain the stratigraphic data of the overburden rockfill dam and the measured data of the borehole;

[0044] The regional division module is used to establish a three-dimensional finite element model of the rockfill dam based on the stratigraphic data of the overburden rockfill dam, determine the optimal network size using the grid size optimization algorithm, and divide the overburden rockfill dam into regions according to the optimal grid size;

[0045] The rock formation property prediction module is used to predict the rock formation properties in the regional unit where the borehole is located based on the measured data of the borehole and use the Markov chain algorithm to obtain the corresponding material parameters according to the rock formation properties of each regional unit;

[0046] A data processing module is used to establish the correlation between rheological parameters and various material parameters, and obtain the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials;

[0047] The rheological analysis module is used to input the rheological parameters in each regional unit into the three-dimensional finite element model of the overburden rockfill dam, perform rheological analysis, and obtain the displacement deformation results corresponding to each regional unit.

[0048] The beneficial effects of the present invention are as follows: a method for analyzing the spatial variability of rheological parameters of the overburden foundation of the present invention meshes the overburden through a finite element model, and improves the accuracy and reliability of the model analysis through the application of a mesh size optimization algorithm and a Markov chain algorithm. The optimization of the mesh size should comprehensively consider the calculation accuracy and calculation efficiency to ensure that accurate results are obtained within a reasonable time. With respect to the drilling data information, the influence of the spatial variability of the parameters on the long-term rheology of the overburden rockfill dam is considered. The rheology of the rockfill dam is affected by the spatial variability of the parameters of the overburden foundation material, and the spatial variability of the overburden parameters directly affects the rheological properties of the foundation and part of the dam body. The Markov chain theory is used to predict the geological lithology distribution in the engineering area through drilling information, revealing the influence of the spatial variability of the overburden material on the long-term deformation of the rockfill dam, providing a theoretical basis and technical support for the construction of the rockfill dam, and solving the problem of inaccurate prediction of the current distribution of overburden lithology parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 It is the predicted rock formation distribution information map of the present invention;

[0051] Figure 2-1 It is a normal distribution result diagram of various parameters of the overburden foundation of the present invention;

[0052] Figure 2-2 It is a normal distribution result diagram of various parameters of the overburden foundation of the present invention;

[0053] Figure 2-3 It is a normal distribution result diagram of various parameters of the overburden foundation of the present invention;

[0054] Figure 2-4 It is a normal distribution result diagram of various parameters of the overburden foundation of the present invention;

[0055] Figure 2-5 It is a normal distribution result diagram of various parameters of the overburden foundation of the present invention;

[0056] Figure 3 is a flow chart of the finite element mesh size optimization algorithm of the present invention;

[0057] Figure 4 is a flow chart of the random field implementation of the present invention taking into account autocorrelation and cross-correlation;

[0058] Figure 5 It is a distribution diagram of the random field under the influence of the spatial variability of the overburden foundation parameters of the present invention;

[0059] Figure 6 It is a comparison diagram between the numerical value of the foundation settlement time curve and the measured results when the spatial variability of the foundation is taken into account in the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose and technical solution of the present invention clearer and easier to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0061] The idea of ​​the present invention is to provide a method and system for analyzing the spatial variability of rheological parameters of overburden foundation, which mainly includes:

[0062] Obtaining the stratigraphic information of the overburden rockfill dam and the borehole measurement data;

[0063] A three-dimensional finite element model of the rockfill dam was established based on the stratigraphic data of the overburden rockfill dam, and the optimal network size was determined using a mesh size optimization algorithm. The overburden rockfill dam was divided into regions based on the optimal mesh size.

[0064] According to the measured data of the borehole, the Markov chain algorithm is used to predict the rock formation characteristics in the regional unit where the borehole is located, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit;

[0065] The correlation between rheological parameters and material parameters is established, and the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials is obtained;

[0066] The rheological parameters in each regional unit are input into the three-dimensional finite element model of the overburden rockfill dam, and rheological analysis is performed to obtain the displacement deformation results corresponding to each regional unit.

[0067] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments, wherein the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0068] Example 1

[0069] like Figure 1-Figure 6 The method for analyzing the spatial variability of rheological parameters of an overburden foundation in the embodiment shown mainly includes the following specific implementation steps.

[0070] S1: Obtain the stratigraphic data and borehole measured data of the overburden rockfill dam, including the drilling depth, soil mechanical parameters, and stratigraphic distribution information.

[0071] S2: According to the stratigraphic data of the overburden rockfill dam and the measured data of the borehole in the actual project, a three-dimensional finite element model of the overburden rockfill dam is established, and the finite element mesh is optimized by the established mesh optimization algorithm to reduce the influence of the mesh size on the random field simulation results. In this embodiment, the geological area is divided according to the mesh divided by the finite element software, and the overburden rockfill dam is converted into multiple geological units through the three-dimensional finite element model. On this basis, the relevant distances between the overburden foundation units are calculated. In this embodiment, the three-dimensional finite element model of the overburden rockfill dam is established by using ABAQUS software. The relevant distances between the foundation units need to be calculated using the unit centroid coordinates, and the three-dimensional finite element model of the overburden rockfill dam established by ABAQUS software cannot directly extract the centroid coordinates of each unit. Therefore, in this embodiment, the centroid coordinates are calculated by the coordinates of each vertex of the unit. The specific steps of this method are: respectively extracting the vertex node numbers and node coordinates of the three-dimensional finite element space units in the three-dimensional finite element model of the overburden rockfill dam, processing the vertex node numbers and node coordinates of each unit to obtain the corresponding node coordinates of each unit, and then calculating the centroid coordinates of each unit.

[0072] S3: Based on the measured data of the borehole, the Markov chain algorithm is used to predict the rock formation characteristics within the regional unit where the borehole is located, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit.

[0073] According to the field drilling data, the Markov chain theory is used to convert the borehole rock formation into a state, and each state corresponds to one or more rock formation characteristics. Then the state transition probability matrix is ​​calculated according to the changes in the previous and subsequent drilling data. The trained Markov chain model is used to predict the drilling in the project area and determine the rock formation information.

[0074] The boreholes are scattered in the rectangular area of ​​the cover layer. To establish a two-dimensional Markov chain model, all the boreholes in this embodiment are projected onto a straight line. Figure 1 As shown, in this embodiment, the boreholes are dispersed in a rectangular area of ​​70 m×40 m.

[0075] Specifically, after the overburden rockfill dam is divided into various geological units, a linear fit is performed on the original boreholes on the horizontal plane, the borehole points are projected onto the straight line obtained by the linear fitting, and the projected points are used instead of the actual boreholes to perform coupled Markov chain simulation.

[0076] The Markov chain is used to estimate the vertical transfer probability matrix (VTPM) of the borehole data, where the boreholes used in the estimation process are all the boreholes obtained, and the sampling distance in the vertical direction is the minimum geological unit thickness revealed by the borehole. All vertical transfer probability matrices are diagonally dominant, that is, the value of the diagonal element is greater than the sum of the values ​​of the non-diagonal elements. The numerical direction transfer probability matrix table is calculated based on the vertical soil state transfer count matrix table obtained by the borehole calculation. The specific implementation steps are:

[0077] First, the borehole is divided into small segments of equal thickness using a certain sampling distance, and the state transition count matrix T is calculated. v , the state transition probability is calculated according to the state transition count matrix. The calculation expression of the state transition probability is:

[0078]

[0079] In the formula, is the vertical state transition probability matrix, T v is the state transition count matrix, the elements of the count matrix In the vertical direction, the state S r Transition to state S k The number of transitions and the transition probability For a certain state S r To state S k The number of transitions from state Sr The ratio of the number of all state transitions that start.

[0080] The estimation of the horizontal transfer probability matrix is ​​similar to that of the vertical direction. The horizontal transfer probability matrix estimation only uses the left and right boundary boreholes as conditional information. When using coupled Markov chains to simulate stratum variation and perform probability analysis, all boreholes should be used as conditional information (i.e., the soil state of the area between two adjacent boreholes is simulated in sequence). This simulation scheme can more effectively reduce the uncertainty of the soil state than the simulation scheme that only uses the two boundary boreholes as conditional information.

[0081] The specific operation steps for estimating the horizontal transfer probability matrix are as follows: the number of soil state transfers in the horizontal direction is based on the Walther phase sequence law (Walther phase law), and the geological map is divided into small segments of equal length in the horizontal direction using a certain sampling distance, and the state transfer count matrix T is calculated. h , according to the state transition count matrix T h The transfer probability matrix in the horizontal direction is calculated, and the calculation expression of the transfer probability matrix is:

[0082]

[0083] In the formula, is the horizontal transition probability matrix, the elements of the counting matrix In the horizontal direction, the state S l Transition to state S k The number of transitions and the transition probability For a certain state S l To state S k The number of transitions from state S r The ratio of the number of all state transitions that start.

[0084] Calculate the probability of the observed borehole soil state. The probability of the observed borehole soil state is expressed by the product of the probability of each unit state in the observed borehole, that is:

[0085]

[0086] Where P l is the probability of observing the occurrence of drilling soil, P(Z i,j =S kj ) is the soil state of unit (i, j) is S kj The probability of Nz is the number of units in the vertical direction.

[0087] Find the K value corresponding to the maximum probability of occurrence among the Nk observed borehole soils, that is, Kmax, and perform variation simulation. At this time, the vertical transfer probability matrix of the formation variation simulation is the estimated result in formula (1), and the horizontal transfer probability matrix is ​​obtained by substituting Kmax into formula (4) and calculating it using formula (2). The conditional information here is provided by all the borehole data, that is, the soil state between adjacent boreholes is simulated in turn.

[0088]

[0089] If the states of the leftmost and rightmost columns of cells and the top row of cells are known, the cells on the left and right sides can be revealed by drilling, and the top cell can be directly observed, these known cells can be used to conditionally simulate the states of unknown cells row by row from left to right.

[0090] The conditional Latin hypercube sampling method is used to perform normal distribution random sampling on different materials to generate a material random parameter library.

[0091] The linear elastic model and Drucker-Prager plastic model (DP model) are used to simulate the elastic-plastic properties of granular materials. The plastic yield surface function is:

[0092] F=t-ptanβ-d=0 (5)

[0093] Where F is the yield surface stress, d is the cohesion, which is related to the input hardening parameter, β is the inclination of the yield surface in the pt stress space, which is related to the friction angle, and t is the measurement parameter of the deviatoric stress.

[0094] The time-hardening creep law is used to simulate the rheological characteristics of the rockfill, and the corresponding calculation expression is:

[0095]

[0096] In the formula, is the equivalent creep strain rate, A, n, and m are all fitted values ​​of creep experiments, is the equivalent creep stress.

[0097] Among them, the characteristic values ​​of material parameters include mean, standard deviation, coefficient of variation, probability distribution type and related functions. Among them, the mean, standard deviation and coefficient of variation are obtained by statistical analysis of the rheological parameters in formula (6) obtained by actual measurement. Taking parameter density ρ as an example, the corresponding mean, standard deviation and coefficient of variation are:

[0098]

[0099] Where: μ ρ is the mean, σ ρis the standard deviation, Cov is the covariance, ρ i represents the dry density obtained in each experiment, and l represents the number of data obtained in the experiment.

[0100] The other parameters are processed as above. The probability distribution type of each parameter is obtained by drawing a parameter probability distribution histogram, fitting the probability distribution histogram through a probability distribution function, and adjusting the parameter grouping interval to obtain the optimal probability distribution type. Since the rheological parameter m has a value range of -1 to 0, it is not suitable for lognormal distribution, so normal distribution is selected as the probability distribution type of rheological parameters. The distribution diagrams of each parameter are shown in Figure 3 As shown, the average values ​​of the parameters density ρ, elastic modulus E, rheological parameters A, n, and m are 2.15 g / cm 3 , 69.37Mpa, 0.0000071, 1.2525, -0.9817, and the standard deviation is 0.2g / cm 3 , 3.27Mpa, 0.043, 0.0000018, 0.08894, 0.0052. The coefficient of variation is 0.2.

[0101] S4: Establish the correlation between rheological parameters and various material parameters, and obtain the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials.

[0102] The correlation distance between each unit is calculated based on the extracted node coordinates of the overburden foundation unit, and the autocorrelation random field of density is obtained. The correlation between rheological parameters and density is established, and the correlation standard normal distribution random field of rockfill dam foundation material parameters is obtained.

[0103] In this embodiment, the correlation distance δ is a concept of spatial range, which reflects and describes the correlation between soil bodies in space. The soil bodies within this range are considered to be completely correlated. If the soil bodies exceed this range, they are considered to be basically no longer correlated.

[0104] First, define the conceptual formula of the correlation distance δ, namely:

[0105]

[0106] Where D(Δz) is the correlation function, Δz is the sampling interval, h is the sample interval (usually a multiple of the sampling interval), Γ 2 (h) is the variance reduction function. When the sample spacing is large enough, the correlation distance can be approximated, that is:

[0107]

[0108] In the three-dimensional finite element mesh size of this embodiment, the longest mesh side is 25.54m, the highest mesh side is 15.247m, and the widest mesh side is 13.65m. By calculating the horizontal relevant distance of the overburden layer, the vertical relevant distance is 5m, and the axial relevant distance is 10m.

[0109] The correlation function in the statistical characteristics of each parameter uses a three-dimensional Gaussian function with good stability to describe the autocorrelation relationship of each parameter. The corresponding expression is:

[0110]

[0111] In the formula, τ x , τ y , τ z Represents X i Dot and X j The relative distance between points in three directions; δ x , δ y , δ z , represents the relevant distances in the three directions of x, y, and z.

[0112] Based on formula (12), the autocorrelation coefficient matrix at any position is obtained, that is:

[0113] D n×n =[D 11 ,D 12 ,...,D 1i ,...,D 1n ,] (13)

[0114] Where D n×n is the autocorrelation coefficient matrix, D 1n is the autocorrelation coefficient, and n is the corresponding number of the geological unit.

[0115] Through Cholesky decomposition (Cholesky decomposition method) into the lower triangular matrix L, the density (ρ) related standard normal random field of each position of the foundation is established:

[0116]

[0117] Where L is a lower triangular matrix, is the related standard normal random field, X={(x i ,y i ,z i ), i = 1, ..., n are the coordinates of the position of the covering layer in the foundation. Z is an o-dimensional independent standard normal distribution random variable of dry density.

[0118] Considering the mutual correlation of parameters, the correlation between two parameters can be described by the mutual correlation function and the mutual correlation coefficient function. When the random fields of two parameters selected from the density (ρ) and elastic modulus (E) and the rheological parameters A, m, n are ρ(t) and m(t), the mutual correlation function is defined as:

[0119]

[0120] Where: R ρm , R mρ are the cross-correlation functions of parameters ρ(t) and m(t), and parameters m(t) and ρ(t). The corresponding cross-correlation coefficient function is:

[0121]

[0122] Where: γ Am and γ mA are the mutual correlation coefficient functions of ρ and m, μ ρ , μ m is the mean of the parameter random fields ρ(t) and m(t), σ ρ , σ m is the standard deviation of the parameter random fields ρ(t) and m(t).

[0123] The mutual correlation coefficient matrix and cross-correlation matrix between each parameter are constructed, and the correlated normal distribution random field of each parameter is formed by combining the autocorrelation matrix of each parameter.

[0124] In this embodiment, the autocorrelation and cross-correlation matrices with parameters are taken as follows:

[0125]

[0126] In the formula, B ρρ is the autocorrelation matrix of the density, C ρρ is the autocorrelation coefficient matrix, B mm is the autocorrelation matrix of m, C mm is the autocorrelation coefficient matrix of parameter m, B mρ is the cross-correlation matrix between the density parameter and the parameter m, C mρ is the correlation coefficient matrix between the density parameter and the parameter m, for, for,

[0127] Since the autocorrelation matrix of the density parameter is a real symmetric matrix, its autocorrelation coefficient matrix C ρρ It can be obtained by Cholesky decomposition, that is:

[0128] C ρρ =Choleky(B ρρ ) (18)

[0129] Similarly, we get:

[0130]

[0131] According to the cross-correlation matrix B mρ Find the correlation coefficient matrix of ρ and m:

[0132]

[0133] According to the above formula, the correlation coefficients between the parameters are obtained in the same way, and then the correlation coefficient matrix is ​​obtained:

[0134]

[0135] Performing Cholesky decomposition on the correlation coefficient matrix C0, we can get the following formula:

[0136]

[0137] Among them, L C is the lower triangular matrix of the mutual correlation coefficient, L C T is the matrix transpose, then the cross-correlation standard normal random field of the rheological parameters of the rockfill dam foundation material is as follows:

[0138]

[0139] In the formula, (i=ρ,E,A,n,m) is the parameter description of the standard normal random field considering the correlation between material parameters and B i (i=ρ,E,A,n,m) is the parameter of an o-dimensional independent standard normal distribution random variable, generated using Latin hypercube sampling. L is the lower triangular matrix obtained by decomposing the autocorrelation coefficient. L C T is the transpose of the lower triangular matrix obtained by decomposing the cross-correlation coefficient matrix, where the implementation process of the parameter random field is as follows Figure 4 shown.

[0140] S5: inputting the rheological parameters in each regional unit into the three-dimensional finite element model of the overburden rockfill dam, performing rheological analysis, and obtaining the displacement deformation results corresponding to each regional unit.

[0141] The random parameter library is used as training samples and test samples to construct the Kriging model. The optimal Kriging proxy model for general training uses the MT-DREAM algorithm to carry out random sampling, estimates the digital eigenvalues ​​of the correction parameters based on the central limit theorem, and obtains the correction parameters.

[0142] The Kriging model in this embodiment consists of two parts: a deterministic regression model and a random process. The structural response can be expressed as a function expression of the structural design parameter x using the Kriging model:

[0143]

[0144] In the formula, is the Kriging model estimate of the structural response; f(β,x) is a polynomial function used to simulate the overall trend of the sample points; β is the polynomial regression coefficient matrix; z(x) is a random process with a mean of 0 and a variance of σ 2 The normal distribution of is used to simulate the local approximation of sample points.

[0145] Based on the standard DREAM algorithm, improvements such as history vector differential sampling, snooker update and multiple attempts Metropolis sampling are introduced to better explore the posterior distribution of high-dimensional parameters to be corrected. It uses MTM (Ⅱ) and DREAM algorithms as the main building blocks.

[0146] This embodiment determines the dimension d of the parameter to be corrected, initializes and stores the historical state matrix Z and the initial population X of the Markov chain, sets the maximum number of iterations T and the number of candidate sample points m t . Calculate the posterior probability density value under the current state and start the evolution of N parallel chains, usually N=3.

[0147] Establish multiple candidate samples xtl,i (l = 1, 2, ..., mt). The calculation expression of the candidate samples is:

[0148]

[0149] Determine whether to accept the candidate sample component according to the crossover probability CR (l=1,2,…,mt), as shown in formula (25):

[0150]

[0151] calculate (l=1,2,…mt-1) posterior probability value And order Calculate the posterior probability value of the new candidate sample point, and calculate the acceptance probability according to formula (26) u is randomly generated in the uniform distribution U(0,1), if Then accept new candidate samples

[0152]

[0153] After K population iterations, X is added to Z, and M = M + N. The convergence diagnosis of the Markov chain is performed based on the potential scale reduction factor R. Then, the digital eigenvalues ​​of the modified parameters are estimated based on the central limit theorem to obtain the modified parameter characteristics.

[0154] The unit number of each partition material is extracted separately, and the obtained material parameters are imported into the finite element numerical analysis software through secondary development of the ABAQUS platform to perform rheological calculations and output the displacement deformation of each unit.

[0155] The rheological deformation of the rockfill dam is calculated by analyzing the spatial variability of the rheological parameters of the overburden foundation based on the drilling data information, and the final stress deformation is obtained. In order to verify the rationality and accuracy of the rheological analysis of the rockfill dam considering the spatial variability of the overburden foundation parameters and the lithology parameter prediction method of the project area, the Miaojiaba face rockfill dam project is used as a research example to carry out numerical simulation calculations. The stress deformation calculation results are compared with the actual monitoring data, and the comparison results are shown in Figure 6 .

[0156] from Figure 6 It can be seen that the results calculated by the present invention are basically consistent with the actual engineering results, and the results in the later stage of water storage are closer to the actual engineering. In the early stage of water storage, as the water load increases, the simulation calculation results deviate. In summary, it can be seen that the rheological deformation of the face rockfill dam considering the spatial variability of the overburden foundation parameters is more consistent with the actual engineering than the rheological deformation without considering the spatial variability of the overburden foundation parameters.

[0157] Based on the drilling data information, the present invention considers the influence of parameter spatial variability on the long-term rheology of the overburden rockfill dam. The rheology of the rockfill dam is affected by the spatial variability of the parameters of the overburden foundation material, and the spatial variability of the overburden parameters directly affects the rheological characteristics of the foundation and part of the dam body. Based on the numerical simulation of the long-term rheological deformation of the rockfill dam under the action of the spatial variability of the overburden foundation parameters, the long-term rheological law of the rockfill dam under the action of the spatial variability of the overburden material is revealed by predicting the geological lithology distribution in the engineering area. It provides a theoretical basis and technical support for the construction of rockfill dams.

[0158] Example 2

[0159] A system for analyzing the spatial variability of rheological parameters of an overburden foundation in this embodiment is used to implement a method for analyzing the spatial variability of rheological parameters of an overburden foundation in embodiment 1. The system includes:

[0160] Data acquisition module, used to obtain the stratigraphic data of the overburden rockfill dam and the measured data of the borehole;

[0161] The regional division module is used to establish a three-dimensional finite element model of the rockfill dam based on the stratigraphic data of the overburden rockfill dam, determine the optimal network size using the grid size optimization algorithm, and divide the overburden rockfill dam into regions according to the optimal grid size;

[0162] The rock formation property prediction module is used to predict the rock formation properties in the regional unit where the borehole is located based on the measured data of the borehole and use the Markov chain algorithm to obtain the corresponding material parameters according to the rock formation properties of each regional unit;

[0163] A data processing module is used to establish the correlation between rheological parameters and various material parameters, and obtain the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials;

[0164] The rheological analysis module is used to input the rheological parameters in each regional unit into the three-dimensional finite element model of the overburden rockfill dam, perform rheological analysis, and obtain the displacement deformation results corresponding to each regional unit.

[0165] The implementation principles of each module have been described in detail in the method for analyzing the spatial variability of rheological parameters of an overburden foundation in Example 1, and will not be repeated here.

[0166] It should be understood that the device and method disclosed in the present invention can be implemented in other ways. For example, the device embodiments described above are only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0167] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0168] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A method for analyzing the spatial variability of rheological parameters of overburden foundation, characterized in that: include: Obtaining the stratigraphic information of the overburden rockfill dam and the borehole measurement data; A three-dimensional finite element model of the rockfill dam was established based on the stratigraphic data of the overburden rockfill dam, and the optimal network size was determined using a mesh size optimization algorithm. The overburden rockfill dam was divided into regions based on the optimal mesh size. According to the measured data of the borehole, the Markov chain algorithm is used to predict the rock formation characteristics in the regional unit where the borehole is located, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit; The correlation between rheological parameters and material parameters is established, and the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials is obtained; The rheological parameters in each regional unit are input into the three-dimensional finite element model of the overburden rockfill dam, and rheological analysis is performed to obtain the displacement deformation results corresponding to each regional unit.

2. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 1, characterized in that: The Markov chain algorithm is used to predict the rock formation characteristics in the regional unit where the borehole is located based on the measured data of the borehole, and the corresponding material parameters are obtained according to the rock formation characteristics of each regional unit, specifically including: Using the measured data of boreholes, the Markov chain algorithm is used to estimate the regional unit of each borehole, and the horizontal state transfer count matrix and the vertical state transfer count matrix are obtained. A transfer probability matrix in the horizontal direction is obtained based on the state transfer count matrix in the horizontal direction, and a corresponding transfer probability matrix is ​​obtained based on the state transfer count matrix in the vertical direction; According to the horizontal and vertical transfer probability matrices, the occurrence probability of the soil state in a single borehole is obtained; The soil state of each regional unit is obtained by coupling the data of each single borehole, and the rock formation characteristic information is predicted based on the soil state. The material parameters are obtained based on the rock formation characteristic information, and the material parameters include density, elastic modulus and three rheological parameters.

3. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 2, characterized in that: The method of obtaining a horizontal transition probability matrix based on a horizontal state transition count matrix and obtaining a corresponding state transition probability matrix based on a vertical state transition count matrix specifically includes: The state transition probability matrix is ​​obtained by using the elements in the horizontal state transition count matrix and the sum of each element. The calculation expression is: In the formula, is the horizontal transition probability matrix, is the element in the horizontal state transition count matrix, Represented by the first state S l Transition to the second state S k The number of transfers, is the sum of the elements in the horizontal state transition count matrix; The state transition probability matrix is ​​obtained by using the elements in the vertical state transition count matrix and the sum of each element. The calculation expression is: In the formula, is the vertical state transition probability matrix, is the element in the vertical state transition count matrix, Represented by the first state S l Transition to the second state S k The number of transfers, is the sum of the elements in the vertical state transition count matrix.

4. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 1, characterized in that: After obtaining the corresponding material parameters according to the rock formation characteristics of each regional unit, the method further includes the step of optimizing the material parameters by using a simulation model. The specific steps are as follows: The linear elastic model and the DP plastic model are used to simulate the elastic-plastic properties of the granular material in the rockfill dam to obtain the plastic yield surface function, wherein the granular material is obtained based on the rock formation property information; The time-hardening creep law is used to simulate the rheological characteristics of the rockfill dam to obtain the rheological parameters. Based on the characteristic values ​​of each material parameter, the probability distribution corresponding to each material parameter is obtained. The rheological parameter grouping interval is adjusted according to the probability distribution to obtain the optimal probability distribution type. The characteristic values ​​include mean, standard deviation, and coefficient of variation.

5. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 1, characterized in that: The establishment of the correlation between the rheological parameters and the material parameters to obtain the relevant standard normal distribution random field of the rheological parameters of the rockfill dam foundation material specifically includes: The correlation distance of each regional unit is calculated based on the node coordinates of the geological unit of the overburden layer, and the autocorrelation coefficient matrix corresponding to each material parameter is obtained based on the correlation distance, and the correlation distance is used to describe the correlation between soil bodies in space; The autocorrelation coefficient matrix and the corresponding position coordinates are decomposed by Cholesky to obtain the autocorrelation standard normal random field of each base position; Constructing a mutual correlation coefficient matrix and a cross-correlation matrix between the parameters, and combining the autocorrelation matrix of each parameter to form a correlated normal distribution random field of the parameter, wherein the autocorrelation matrix is ​​obtained by Cholesky decomposition of the autocorrelation standard normal random field of each material parameter; The correlation coefficient matrix is ​​subjected to Cholesky decomposition to obtain a lower triangular matrix, and a standard normal random field of rheological parameter cross-correlation is constructed based on the lower triangular matrix and the cross-correlation matrix.

6. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 5, characterized in that: The relevant distance is: Where D(Δz) is the correlation function, Δz is the sampling interval, h is the sample interval (a multiple of the sampling interval), Γ 2 (h) is the variance reduction function.

7. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 5, characterized in that: The cross-correlation standard normal random field of rheological parameters constructed according to the lower triangular matrix and the cross-correlation matrix includes: the cross-correlation standard normal random field of rheological parameters is: In the formula, is the cross-correlation standard normal random field of rheological parameters, B i is the parameter of the o-dimensional standard normal distribution random variable, L is the lower triangular matrix obtained by decomposing the autocorrelation coefficient, L C T is the transpose of the lower triangular matrix obtained by decomposing the cross-correlation coefficient matrix, i = ρ, E, A, n, m.

8. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 1, characterized in that: The method of obtaining corresponding material parameters according to the rock formation characteristics of each regional unit also includes extracting random samples from the normal distribution of different materials in the rockfill dam using the Latin hypercube sampling method to obtain a material random parameter library.

9. The method for analyzing spatial variability of rheological parameters of overburden foundation according to claim 8, characterized in that: Inputting the rheological parameters in each regional unit into the three-dimensional finite element model of the overburden rockfill dam also includes correcting the rheological parameters. The specific steps are: Establish a Kriging model and train it using the material random parameter library; The trained Kriging model is used for random sampling, and the digital eigenvalues ​​of the correction parameters are estimated according to the central limit theorem. The rheological parameters are corrected using the digital eigenvalues ​​of the correction parameters, and rheological calculations are performed based on the corrected rheological parameters.

10. A system for analyzing the spatial variability of rheological parameters of an overburden foundation, used to implement the method for analyzing the spatial variability of rheological parameters of an overburden foundation according to any one of claims 1 to 9, characterized in that: include: Data acquisition module, used to obtain the stratigraphic data of the overburden rockfill dam and the measured data of the borehole; The regional division module is used to establish a three-dimensional finite element model of the rockfill dam based on the stratigraphic data of the overburden rockfill dam, determine the optimal network size using the grid size optimization algorithm, and divide the overburden rockfill dam into regions according to the optimal grid size; The rock formation property prediction module is used to predict the rock formation properties in the regional unit where the borehole is located based on the measured data of the borehole and use the Markov chain algorithm to obtain the corresponding material parameters according to the rock formation properties of each regional unit; A data processing module is used to establish the correlation between rheological parameters and various material parameters, and obtain the relevant standard normal distribution random field of rheological parameters of rockfill dam foundation materials; The rheological analysis module is used to input the rheological parameters in each regional unit into the three-dimensional finite element model of the overburden rockfill dam, perform rheological analysis, and obtain the displacement deformation results corresponding to each regional unit.

Citation Information

Cited By

  • Simulation method of three-dimensional irregular complex heterostructure body in rock-soil stratum

    CN119783229A