Finite element model construction method considering rock mechanical parameter joint distribution

Abstract

The invention provides a finite element model construction method considering rock mechanical parameter joint distribution, which comprises the following steps: carrying out a compression test on a rock sample to obtain a rock mechanical parameter set; re-sampling the rock mechanical parameter set through a statistical re-sampling technology to obtain a random sample set; obtaining an optimal edge distribution function of the random sample set through a probability distribution function and KS test; constructing a joint distribution function according to the optimal edge distribution function and the Gaussian connection function; performing random extraction operation on the random sample set according to the joint distribution function and the random function to obtain an optimal edge distribution random variable; and constructing a finite element model through a finite element analysis technology and the optimal edge distribution random variable. According to the method, joint distribution of the rock mechanical parameters is fully considered, the problem that a traditional model is single or simplified when the rock mechanical parameters are considered is effectively solved, and the reliability and applicability of the model are improved.

Description

Technical Field

[0001] The invention relates to the field of rock mechanics and engineering technology, and in particular to a finite element model construction method considering the joint distribution of rock mechanics parameters. Background Art

[0002] In geological engineering and geotechnical engineering, rock physical and mechanical parameters such as elastic modulus, Poisson's ratio, internal friction angle, cohesion, etc. have a vital impact on the stability analysis of surrounding rocks. With the improvement of engineering requirements and technological development, such as tunnel excavation and mining, it is necessary to fully consider the joint distribution of rock mechanical parameters to provide strong theoretical support and methodological guarantee for the stability analysis of surrounding rocks, otherwise it may lead to unreasonable support design, increased construction risks and other problems.

[0003] However, the traditional finite element model construction method is often relatively simple or simplified when considering rock mechanical parameters. In actual situations, these parameters do not exist in isolation, but are interrelated and work together, so they cannot fully reflect the real complex characteristics of rock physical and mechanical parameters. Therefore, it is very necessary to design a finite element model construction method that considers the joint distribution of rock mechanical parameters. Summary of the invention

[0004] The purpose of the present invention is to provide a finite element model construction method considering the joint distribution of rock mechanical parameters, so as to realize the parametric modeling of the finite element model through the complex correlation between rock physical and mechanical parameters and Python language compilation, so as to provide a more accurate and reliable theoretical basis and method support for the stability analysis of surrounding rocks under complex geological conditions.

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

[0006] A finite element model construction method considering the joint distribution of rock mechanical parameters includes the following steps:

[0007] Compression tests are carried out on rock samples to obtain a set of rock mechanics parameters; the compression tests include: uniaxial compression tests and triaxial compression tests under different confining pressures; the rock mechanics parameters include: elastic modulus, Poisson's ratio, internal friction angle and cohesion;

[0008] The rock mechanics parameter set is resampled by statistical resampling technology to obtain a random sample set;

[0009] The optimal marginal distribution function of the random sample set is obtained through probability distribution function and KS test; probability distribution functions include: normal distribution, gamma distribution, half-normal distribution, lognormal distribution, Weibull distribution and extreme value type I distribution;

[0010] Construct a joint distribution function based on the optimal marginal distribution function and the Gaussian link function;

[0011] According to the joint distribution function and the random function, a random sample set is randomly drawn to obtain the optimal marginal distribution random variable;

[0012] The finite element model is constructed by finite element analysis technology and optimal marginal distribution random variables.

[0013] Optionally, a compression test is performed on the rock sample to obtain a rock mechanics parameter set, specifically including:

[0014] Conduct no less than n groups of uniaxial compression tests and triaxial compression tests under different confining pressures to obtain compression test data; where n≥9;

[0015] According to the compression test data, the Mohr circle is drawn on the rectangular coordinate system in the two-dimensional plane to obtain n Mohr envelopes;

[0016] Randomly select two different Mohr envelopes for data analysis and obtain C n 2 The rock mechanics parameters are grouped into a rock mechanics parameter set; n 2 ≥36.

[0017] Optionally, the rock mechanics parameter set is resampled by statistical resampling technology to obtain a random sample set, specifically including:

[0018] C n 2 Perform random sampling with replacement, and record the rock mechanical parameters of each random sampling to obtain a new sample set;

[0019] Repeat the steps of obtaining a new sample set N-2 times to obtain N-1 different new sample sets, and integrate the N-1 different new sample sets and the rock mechanics parameter set to obtain a random sample set; wherein N≥10000.

[0020] Optionally, the normal distribution is calculated as: Among them, f1(x) is the normal distribution function value, x is the input variable, is the mean of x, σ is the standard deviation of x;

[0021] The calculation formula for the gamma distribution is: Where f2(x) is the value of the gamma distribution function, k is the shape parameter, θ is the scale parameter, μ is the location parameter, Γ(k) is the gamma function, x>0;

[0022] The calculation formula for the half-normal distribution is: Where, f3(x) is the value of the half-normal distribution function, x ≥ μ;

[0023] The calculation formula for the lognormal distribution is: Where f4(x) is the value of the lognormal distribution function, and s is the shape parameter;

[0024] The calculation formula of Weibull distribution is: Where f5(x) is the value of the Weibull distribution function, x ≥ 0;

[0025] The calculation formula of extreme value type I distribution is: Among them, f6(x) is the value of the extreme value type I distribution function.

[0026] Optionally, the optimal marginal distribution function of the random sample set is obtained through the probability distribution function and the KS test, specifically including:

[0027] The KS statistics of normal distribution, gamma distribution, half-normal distribution, lognormal distribution, Weibull distribution and extreme value type I distribution were obtained through KS test;

[0028] Different KS statistics of the same rock mechanical parameters are compared, and the probability distribution function corresponding to the smallest KS statistic is selected as the optimal marginal distribution function.

[0029] Optionally, a joint distribution function is constructed according to the optimal marginal distribution function and the Gaussian link function, specifically including:

[0030] According to the optimal marginal distribution function, the uniform distribution variables corresponding to different rock mechanical parameters are obtained;

[0031] Based on the definition of Gaussian link function, the joint distribution function is constructed by the correlation coefficient matrix of the random sample set and the uniform distribution variable; the calculation formula of the joint distribution function is: C(u) = Φ R (Φ -1 (u1), Φ -1 (u2), Φ -1 (u3), Φ -1 (u4)); where C(u) is the joint distribution function, Φ is the cumulative distribution function of the standard normal distribution, Φ -1 is the inverse cumulative distribution function of the standard normal distribution, R is the correlation coefficient matrix, and u1, u2, u3 and u4 are all uniformly distributed variables.

[0032] Optionally, a random sampling operation is performed on a random sample set according to a joint distribution function and a random function to obtain an optimal marginal distribution random variable, specifically including:

[0033] Randomly extract uniformly distributed variables through random functions to obtain random numbers;

[0034] The random numbers are converted into correlated random numbers through the Gaussian link function;

[0035] The correlated random numbers are transformed into optimal marginal distribution random variables through the inverse cumulative function of the marginal distribution.

[0036] Optionally, a finite element model is constructed by using finite element analysis techniques and optimal marginal distribution random variables, specifically including:

[0037] Grouping the random sample set to obtain geometric partition units;

[0038] Meshing the geometrical division units according to the finite element analysis technique and setting the divided units as a separate set;

[0039] The parameters of individual sets are assigned through optimal marginal distribution random variables, and the finite element model is constructed through parametric modeling technology.

[0040] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: the present invention provides a finite element model construction method considering the joint distribution of rock mechanical parameters, the method comprising: performing a compression test on a rock sample to obtain a rock mechanical parameter set; resampling the rock mechanical parameter set by statistical resampling technology to obtain a random sample set; obtaining the optimal marginal distribution function of the random sample set by probability distribution function and KS test; constructing a joint distribution function based on the optimal marginal distribution function and Gaussian link function; performing a random sampling operation on the random sample set based on the joint distribution function and the random function to obtain an optimal marginal distribution random variable; constructing a finite element model by finite element analysis technology and the optimal marginal distribution random variable. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] 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 will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0042] Figure 1 A flow chart for constructing a finite element model of an embodiment of the present invention;

[0043] Figure 2 A Mohr circle curve diagram of an embodiment of the present invention;

[0044] Figure 3 Schematic diagram of a division unit according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] like Figure 1 As shown, an embodiment of the present invention provides a finite element model construction method considering the joint distribution of rock mechanics parameters, comprising the following steps:

[0048] Step 100: Perform a compression test on the rock sample to obtain a rock mechanics parameter set; the compression test includes: a uniaxial compression test and a triaxial compression test under different confining pressures; the rock mechanics parameters include: elastic modulus, Poisson's ratio, internal friction angle and cohesion;

[0049] Specifically, no less than n groups of uniaxial compression tests and triaxial compression tests under different confining pressures are performed to obtain compression test data; wherein n≥9; in this embodiment, n=9, and the compression test results are shown in Table 1:

[0050] Table 1 Comparison of compression test results

[0051]

[0052]

[0053] like Figure 2 As shown, nine Mohr circles are drawn on a rectangular coordinate system in a two-dimensional plane according to nine sets of compression test data to obtain the Mohr envelope;

[0054] Randomly select two different Mohr envelopes for data analysis and obtain C n 2 The rock mechanics parameters are grouped into a rock mechanics parameter set; n 2 ≥36. In this embodiment, 36 sets of analysis results of elastic modulus, Poisson's ratio, internal friction angle and cohesion are obtained, as shown in Table 2:

[0055] Table 2 Analysis results of rock mechanics parameter set

[0056]

[0057]

[0058] Step 200: resampling the rock mechanics parameter set by statistical resampling technology to obtain a random sample set;

[0059] Specifically, the rock mechanics parameter set is C n 2 Perform random sampling with replacement, and record the rock mechanical parameters of each random sampling to obtain a new sample set; repeat the step of obtaining a new sample set N-2 times to obtain N-1 different new sample sets, and integrate the data of the N-1 different new sample sets and the rock mechanical parameter set to obtain a random sample set; where N≥10000.

[0060] It should be noted that the statistical resampling technology of this embodiment uses the Bootstrap method to perform 36 random samplings with replacement, and re-records the rock mechanics parameter data sampled each time, thereby obtaining a new sample set with the same sample size as the rock mechanics parameter set.

[0061] In this embodiment, N is 10000, that is, the step of obtaining a new sample set is repeated 9998 times, and a total of 9999 new sample sets are obtained. All new sample sets and the initial rock mechanics parameter set are combined to obtain a random sample set.

[0062] Step 300: Obtaining the optimal marginal distribution function of the random sample set through probability distribution function and KS test; the probability distribution function includes: normal distribution, gamma distribution, half-normal distribution, lognormal distribution, Weibull distribution and extreme value type I distribution;

[0063] Specifically, the calculation formula for the normal distribution is: Among them, f1(x) is the value of the normal distribution function, x is the input variable, that is, the rock mechanics parameter, is the mean of x, and σ is the standard deviation of x; in this embodiment, the means of elastic modulus, Poisson's ratio, internal friction angle and cohesion are 56.39, 0.25, 52.44 and 41.54, respectively, and the standard deviations are 1.69, 0.026, 4.46 and 12.97, respectively.

[0064] The calculation formula for the gamma distribution is: Wherein, f2(x) is the value of the gamma distribution function, k is the shape parameter, θ is the scale parameter, μ is the location parameter, Γ(k) is the gamma function, and x>0; the values ​​of the shape parameter, location parameter and scale parameter of the elastic modulus, Poisson's ratio, internal friction angle and cohesion in the gamma distribution in this embodiment are shown in Table 3:

[0065] Table 3 Gamma distribution parameters

[0066] Characteristic parameters Elastic modulus Poisson's ratio Internal friction angle Cohesion Shape parameters 312.75 802.91 314.75 1.92 Positional parameters 26.27 -0.49 -27.31 22.15 Scale parameter 0.096 0.00091 0.25 10.08

[0067] The calculation formula for the half-normal distribution is: Wherein, f3(x) is the value of the half-normal distribution function, x≥μ; the values ​​of the position parameters and scale parameters of the elastic modulus, Poisson's ratio, internal friction angle and cohesion in the half-normal distribution in this embodiment are shown in Table 4:

[0068] Table 4 Half-normal distribution parameters

[0069]

[0070]

[0071] The calculation formula for the lognormal distribution is: Wherein, f4(x) is the lognormal distribution function value, and s is the shape parameter; the values ​​of the shape parameter, position parameter and scale parameter of the elastic modulus, Poisson's ratio, internal friction angle and cohesion in the lognormal distribution in this embodiment are shown in Table 5:

[0072] Table 5 Lognormal distribution parameters

[0073] Characteristic parameters Elastic modulus Poisson's ratio Internal friction angle Cohesion Shape parameters 1.47 2.17 1.70 0.55 Positional parameters -2044.04 -1190.34 -261355.89 16.89 Scale parameter 2048.08 1190.58 261408.33 21.34

[0074] The calculation formula of Weibull distribution is: Wherein, f5(x) is the value of the Weibull distribution function, x≥0; the values ​​of the shape parameter, position parameter and scale parameter of the elastic modulus, Poisson's ratio, internal friction angle and cohesion in the Weibull distribution in this embodiment are shown in Table 6:

[0075] Table 6 Weibull distribution parameters

[0076] Characteristic parameters Elastic modulus Poisson's ratio Internal friction angle Cohesion Shape parameters 6.14 3.65 6.05 1.44 Positional parameters 47.61 0.16 29.31 22.67 Scale parameter 9.46 0.093 24.92 20.75

[0077] The calculation formula of extreme value type I distribution is: Wherein, f6(x) is the value of the extreme value type I distribution function. The values ​​of the position parameters and shape parameters of the elastic modulus, Poisson's ratio, internal friction angle and cohesion in the extreme value type I distribution in this embodiment are shown in Table 7:

[0078] Table 7 Extreme value type I distribution parameters

[0079] Characteristic parameters Elastic modulus Poisson's ratio Internal friction angle Cohesion Positional parameters 55.98 0.24 51.35 35.01 Shape parameters 0.51 1.25 0.50 -0.10

[0080] Specifically, the optimal marginal distribution function of the random sample set is obtained through the probability distribution function and the KS test, including:

[0081] The KS statistics of normal distribution, gamma distribution, half-normal distribution, lognormal distribution, Weibull distribution and extreme value type I distribution are obtained through KS test respectively; the KS statistics of different probability distribution functions of this embodiment are shown in Table 8:

[0082] Table 8 KS statistics table

[0083]

[0084] Different KS statistics of the same rock mechanics parameter are compared, and the probability distribution function corresponding to the smallest KS statistic is selected as the optimal marginal distribution function. For example, if the KS statistic corresponding to the normal distribution is the smallest for the elastic modulus, then the normal distribution is determined as the optimal marginal distribution function of the elastic modulus. In this embodiment, the optimal marginal distribution of the elastic modulus is the Weibull distribution, the optimal marginal distribution of the Poisson's ratio is the Weibull distribution, the optimal marginal distribution of the internal friction angle is the normal distribution, and the optimal marginal distribution of the cohesion is the Weibull distribution.

[0085] Step 400: constructing a joint distribution function according to the optimal marginal distribution function and the Gaussian link function; the specific steps include:

[0086] According to the optimal marginal distribution function, the uniform distribution variables corresponding to different rock mechanical parameters are obtained;

[0087] Specifically, the uniform distribution variable calculation formula of the elastic modulus in this embodiment is: Among them, x i is a random variable.

[0088] The Poisson's ratio for a uniformly distributed variable is calculated as:

[0089] The uniform distribution variable calculation formula of the internal friction angle is:

[0090] The calculation formula for the uniform distribution variable of cohesion is:

[0091] Based on the definition of Gaussian link function, the joint distribution function is constructed through the correlation coefficient matrix of the random sample set and the uniformly distributed variables;

[0092] Specifically, the correlation coefficient matrix R in this embodiment is:

[0093]

[0094] The calculation formula of the joint distribution function is:

[0095] C(u)=Φ R (Φ -1 (u1), Φ -1 (u2), Φ -1 (u3), Φ -1 (u4));

[0096] Where C(u) is the joint distribution function, Φ is the cumulative distribution function of the standard normal distribution, Φ -1 is the inverse cumulative distribution function of the standard normal distribution, R is the correlation coefficient matrix, and u1, u2, u3 and u4 are all uniformly distributed variables after transformation by the optimal marginal distribution function.

[0097] Step 500: Perform a random sampling operation on a random sample set according to a joint distribution function and a random function to obtain an optimal marginal distribution random variable; the specific steps include:

[0098] Randomly extract uniformly distributed variables through random functions to obtain random numbers;

[0099] Specifically, the random numbers of elastic modulus, Poisson's ratio, internal friction angle and cohesion obtained by random drawing in this embodiment are 56.23 GPa, 0.23, 53.20° and 38.22 MPa, respectively, and the obtained random numbers are converted into random numbers u1, u2, u3 and u4 that obey uniform distribution, which are 0.4316, 0.4331, 0.5677 and 0.4969, respectively, in this embodiment.

[0100] The random numbers are converted into correlated random numbers through the Gaussian link function;

[0101] Specifically, u1, u2, u3 and u4 are converted into random numbers u1′, u2′, u3′ and u4′ with correlation through the Gaussian Copula function, which are 0.3194, 0.2575, 0.3895 and 0.3733 respectively in this embodiment.

[0102] The correlated random numbers are transformed into optimal marginal distribution random variables through the inverse cumulative function of the marginal distribution.

[0103] Specifically, the inverse cumulative function of the marginal distribution in this embodiment is solved by the Newton-Raphson iteration method. The optimal marginal distribution random variables U1′, U2′, U3′ and U4′ are 56.84 GPa, 0.25, 54.18° and 44.01 MPa respectively.

[0104] It should be noted that all the steps in step 500 are implemented through Python coding to generate a random program that can implement step 500, and the random program is tested and debugged to ensure that the program can accurately perform random sampling operations according to the given joint distribution function and provide a reliable data generation tool for subsequent numerical modeling.

[0105] Step 600: Construct a finite element model using finite element analysis technology and optimal marginal distribution random variables. The specific steps include:

[0106] Grouping the random sample set to obtain geometric partition units;

[0107] Meshing the geometrical division units according to the finite element analysis technique and setting the divided units as a separate set;

[0108] Specifically, based on the Mesh grid division in ABAQUS software, the divided units are set as a separate set through macro job recording and parametric modeling. Figure 3 As shown, in this embodiment, the number of divided units is 1512, and there are 1512 sets.

[0109] The parameters of individual sets are assigned through optimal marginal distribution random variables, and the finite element model is constructed through parametric modeling technology.

[0110] Specifically, based on the parametric modeling technology, the parameters of the random sample set are assigned according to the optimal marginal distribution random variables U1′, U2′, U3′ and U4′, so as to construct a finite element model.

[0111] The beneficial effects of the present invention are as follows:

[0112] 1) Based on Python programming language, the finite element model is constructed through statistical resampling technology, Gaussian link function and finite element analysis technology, fully considering the joint distribution of rock mechanical parameters;

[0113] 2) The present invention more truly reflects the complex characteristics of rock physical and mechanical parameters, effectively solves the problem of single or simplified traditional models when considering rock mechanical parameters, and improves the reliability and applicability of the model.

[0114] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0115] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A finite element model construction method considering the joint distribution of rock mechanical parameters, characterized in that: The steps include: Performing compression tests on rock samples to obtain a set of rock mechanics parameters; the compression tests include: uniaxial compression tests and triaxial compression tests under different confining pressures; the rock mechanics parameters include: elastic modulus, Poisson's ratio, internal friction angle and cohesion; Resampling the rock mechanics parameter set by statistical resampling technology to obtain a random sample set; The optimal marginal distribution function of the random sample set is obtained by probability distribution function and KS test; the probability distribution function includes: normal distribution, gamma distribution, half-normal distribution, lognormal distribution, Weibull distribution and extreme value type I distribution; Constructing a joint distribution function according to the optimal marginal distribution function and the Gaussian copula function; Performing a random sampling operation on the random sample set according to the joint distribution function and the random function to obtain an optimal marginal distribution random variable; A finite element model is constructed by using the finite element analysis technique and the optimal marginal distribution random variables.

2. The finite element model construction method considering the joint distribution of rock mechanical parameters according to claim 1 is characterized in that: Compression tests are performed on rock samples to obtain a set of rock mechanics parameters, including: Conduct no less than n groups of uniaxial compression tests and triaxial compression tests under different confining pressures to obtain compression test data; where n≥9; According to the compression test data, a Mohr circle is drawn on a rectangular coordinate system in a two-dimensional plane to obtain n Mohr envelope lines; Randomly select two different Mohr envelopes for data analysis and obtain C n 2 The rock mechanics parameters are grouped into the rock mechanics parameter set; wherein C n 2 ≥36.

3. The finite element model construction method considering the joint distribution of rock mechanics parameters according to claim 1 is characterized in that: The rock mechanics parameter set is resampled by statistical resampling technology to obtain a random sample set, which specifically includes: The rock mechanics parameter set is C n 2 random sampling with replacement, and recording the rock mechanical parameters of each random sampling to obtain a new sample set; Repeat the step of obtaining the new sample set N-2 times to obtain N-1 groups of different new sample sets, and integrate the N-1 groups of different new sample sets and the rock mechanics parameter set to obtain the random sample set; wherein N≥10000.

4. The finite element model construction method considering the joint distribution of rock mechanical parameters according to claim 1 is characterized in that: The calculation formula of the normal distribution is: Among them, f1(x) is the value of the normal distribution function, x is the input variable, is the mean of x, σ is the standard deviation of x; The calculation formula of the gamma distribution is: Where f2(x) is the value of the gamma distribution function, k is the shape parameter, θ is the scale parameter, μ is the location parameter, Γ(k) is the gamma function, x>0; The calculation formula of the half-normal distribution is: Where, f3(x) is the value of the half-normal distribution function, x ≥ μ; The calculation formula of the lognormal distribution is: Where f4(x) is the value of the lognormal distribution function, and s is the shape parameter; The calculation formula of the Weibull distribution is: Where f5(x) is the value of the Weibull distribution function, x ≥ 0; The calculation formula of the extreme value type I distribution is: Among them, f6(x) is the value of the extreme value type I distribution function.

5. The finite element model construction method considering the joint distribution of rock mechanical parameters according to claim 1 is characterized in that: The optimal marginal distribution function of the random sample set is obtained by using the probability distribution function and the KS test, specifically including: The KS statistics of the normal distribution, the gamma distribution, the half-normal distribution, the lognormal distribution, the Weibull distribution and the extreme value type I distribution are obtained respectively through the KS test; Different KS statistics of the same rock mechanics parameter are compared, and the probability distribution function corresponding to the smallest KS statistic is selected as the optimal marginal distribution function.

6. The finite element model construction method considering the joint distribution of rock mechanics parameters according to claim 1 is characterized in that: Constructing a joint distribution function according to the optimal marginal distribution function and the Gaussian link function specifically includes: Obtaining uniformly distributed variables corresponding to different rock mechanical parameters according to the optimal marginal distribution function; Based on the definition of the Gaussian link function, the joint distribution function is constructed by the correlation coefficient matrix of the random sample set and the uniform distribution variable; the calculation formula of the joint distribution function is: C(u)=Φ R (Φ -1 (u1), Φ -1 (u2), Φ -1 (u3), Φ -1 (u4)); where C(u) is the joint distribution function, Φ is the cumulative distribution function of the standard normal distribution, Φ -1 is the inverse cumulative distribution function of the standard normal distribution, R is the correlation coefficient matrix, and u1, u2, u3 and u4 are all uniformly distributed variables.

7. The method for constructing a finite element model considering the joint distribution of rock mechanics parameters according to claim 6, characterized in that: The random sample set is randomly sampled according to the joint distribution function and the random function to obtain an optimal marginal distribution random variable, which specifically includes: Randomly extract the uniformly distributed variable by using a random function to obtain a random number; Converting the random number into a correlation random number by using the Gaussian link function; The correlation random number is converted into the optimal marginal distribution random variable through the inverse cumulative function of the marginal distribution.

8. The finite element model construction method considering the joint distribution of rock mechanics parameters according to claim 1 is characterized in that: The finite element model is constructed by using the finite element analysis technology and the optimal marginal distribution random variables, specifically including: Grouping the random sample set to obtain geometric division units; Meshing the geometrical division units according to the finite element analysis technique, and setting the divided units as a separate set; The parameters of the individual sets are assigned by using the optimal marginal distribution random variables, and the finite element model is constructed by using a parametric modeling technique.

Images