Cyclone separator reliability analysis method based on Kriging model analysis high-order moment and maximum entropy method
Through the reliability analysis of cyclone separator based on the Kriging model and maximum entropy method, the problem of separation efficiency uncertainty caused by cyclone separator due to size changes in the nuclear industrial environment is solved, and its stability and reliability are improved, reducing the stability and reliability of the cyclone separator. failure risk and maintenance costs.
Patent Information
- Application Number
- CN202411148483.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-12-06
AI Technical Summary
In the nuclear industrial environment, the separation efficiency of the cyclone separator is affected by the uncertainty of the size changes of the collection barrel and cone table, resulting in an increased risk of failure in the radiation environment, high maintenance costs, and affecting the safety of nuclear operations.
Using the reliability analysis method based on the Kriging model and maximum entropy method, the structural function function of the cyclone separator is constructed by analyzing uncertain factors, the high-order moment and probability density functions are calculated, and the separation efficiency and structural reliability of the cyclone separator are evaluated. Spend.
It improves the stability and reliability of the cyclone separator, reduces maintenance costs, reduces the risk of failure, and provides more efficient reliability analysis and calculation, suitable for reliability analysis of any structure.
Smart Images

Figure CN119089774A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of reliability analysis, and in particular relates to a reliability analysis method for a cyclone separator based on a Kriging model to analyze high-order moments and a maximum entropy method. Background Art
[0002] At present, industrial robots are gradually being used in nuclear power plants to replace workers to complete a series of operations to keep workers away from the impact of nuclear radiation. As an important equipment for post-processing processes in the nuclear industry, the residue cleaning robot has a key piece of equipment, the cyclone separator. Since the residue is highly radioactive, effective measures must be taken to separate, process and store it to ensure the safety of personnel and the environment. Once a cyclone separator fails in a radiation environment, it will not only incur high maintenance costs, but will also cause related nuclear operations to stop, and even increase the risk of safety accidents. Therefore, in order to ensure the normal operation of the cyclone separator in a radiation environment, a reliability analysis is required to ensure that its reliability meets the requirements.
[0003] The cyclone separator mainly forms centrifugal force through the high-speed rotation of the cyclone barrel, depositing heavy particles at the bottom, while the clean air is discharged through the gas outlet, thereby achieving the purpose of separation. In the nuclear industry environment, the working state of the cyclone separator is affected by many factors. Among them, the changes in the manufacturing dimensions of the collection barrel, cone, etc. will cause a certain dispersion in its separation efficiency, thereby affecting the accuracy and stability of the operation. Therefore, it is necessary to consider the influence of the uncertainty propagation of the radius of the cyclone separator collection barrel and the height of the cone on the separation efficiency of the cyclone separator, and evaluate the reliability of the cyclone separator. Summary of the invention
[0004] In response to the above problems, the present invention takes into account the uncertain factors that affect the separation efficiency of the cyclone separator. After obtaining the influence law of the uncertain factors on the separation efficiency of the cyclone separator, certain uncertain factors are controlled in a targeted manner according to this law, providing theoretical guidance for the subsequent optimization of the cyclone separator design, so as to improve the stability and reliability of the cyclone separator in actual operation.
[0005] The technical solution adopted by the present invention is: a reliability analysis method of a cyclone separator based on the analytical high-order moment of the Kriging model and the maximum entropy method, comprising:
[0006] S1 Analyze the uncertain factors existing in the actual working process of cyclone separator;
[0007] S2 obtains sample points based on ANSYS simulation;
[0008] S3 uses sample points to train and obtain the Kriging model of the cyclone separator structure function function;
[0009] S4 uses the algorithm in this paper to analyze the n-order origin moment of the Kriging model;
[0010] S5 uses the maximum entropy method to obtain the probability density function of the cyclone separation efficiency using the obtained n-order origin moment;
[0011] S6 solves the structural reliability of the cyclone separator according to the given lower limit threshold of separation efficiency.
[0012] The beneficial effects of the present invention are:
[0013] (1) The present invention takes into account the uncertainty factors of the cyclone separator system parameters, which is consistent with the actual situation, and obtains the probability density function and cumulative distribution function of the cyclone separator separation effect, accurately and objectively characterizes the uncertainty existing in the actual working process of the cyclone separator, realizes the reliability analysis of the cyclone separator, and provides a theoretical reference for subsequent reliability tests, structural optimization design, etc.;
[0014] (2) Compared with the traditional reliability analysis method, the method proposed in the present invention can improve the reliability analysis calculation efficiency while ensuring the analysis accuracy;
[0015] (3) Compared with other statistical moment calculation methods, the present invention uses analytical methods to directly calculate high-order moments, calls the function fewer times, and the calculation error is still very small, while ensuring the efficiency and accuracy of the calculation.
[0016] (4) The cyclone separator reliability analysis method proposed in the present invention is applicable to the reliability analysis of any structure, has no special requirements for the structure, and has a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 A flow chart of a cyclone separator reliability analysis method based on the analytical high-order moments of the Kriging model and the maximum entropy method provided in an embodiment of the present invention.
[0018] Figure 2 Ansys simulation model of a cyclone separator provided in an embodiment of the present invention.
[0019] Figure 3 A Kriging model of the structural function of a cyclone separator provided in an embodiment of the present invention.
[0020] Figure 4 It is the probability density function of the separation efficiency of the cyclone separator calculated according to the embodiment of the present invention.
[0021] Figure 5 It is the cumulative distribution function of the separation efficiency of the cyclone separator calculated in the embodiment of the present invention. DETAILED DESCRIPTION
[0022] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.
[0023] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.
[0024] like Figure 1 As shown, the present invention provides a cyclone separator reliability analysis method based on the analytical high-order moment of the Kriging model and the maximum entropy method, comprising the following steps;
[0025] 1. Combined with actual parameters and based on Ansys simulation analysis of the uncertainty in the working process of the cyclone separator, the structural function function of the cyclone separator is determined, and the function is:
[0026] Y=G(X) (1)
[0027] Where X=(X1,X2,…,X K ) T (K is the number of parameters) are the uncertain input variables that affect its performance. As shown in Table 1, they are the radius of the collection bucket and the height of the cone, respectively. These two variables are recorded as X1 and X2. The three-dimensional model of the collection bucket is shown in Figure 2 As shown;
[0028] Table 1 Uncertainty variable parameters
[0029] variable Distribution Type Mean Coefficient of variation Radius of collection bucket (mm) normal distribution 100 1% Height of cone (mm) normal distribution 180 1%
[0030] 2. Convert the input variable into a standard normal variable. That is, convert the input variable X into a standard normal variable U through formula (2) to obtain the structural function function Y=G(U).
[0031]
[0032] in, For X i The cumulative distribution function, Φ -1 is the inverse function of the cumulative distribution function of the standard normal distribution. The two non-standard normal distribution variables are transformed into standard normal distribution variables through mapping transformation, specifically denoted as U1 and U2;
[0033] 3. Obtain 10 non-standard normal input sample points through Latin hypercube sampling, and convert them into standard normal input sample points. Specifically, obtain the standard normal distribution input variable matrix u through the S2 transformationin ,in
[0034]
[0035] Furthermore, according to the extracted training sample points, the corresponding performance function response vector g is calculated by ANSYS simulation:
[0036] g=[g (1) ,g (2) , …, g (N) ] T (4)
[0037] The Gaussian kernel function is further selected as the correlation function of the Kriging model, as shown in the following formula:
[0038]
[0039] Among them, u (i) is the i-th sample point of U (i=1,2,…,N), u k (i) for u (i) The k-th dimension component (k = 1, 2, ..., K), l k is the correlation length of U in the k-th dimension, L is a K×K symmetric matrix with the main diagonal elements being l k 2 ,c(u (i) ,u (j) ; L -1 ) is called the unit coefficient independent Gaussian kernel.
[0040] Finally, the regression function is selected as a constant polynomial, and the random field is trained through sample points to obtain the Kriging model of the cyclone separator function function:
[0041]
[0042] Where μ is The mean of , r is the correlation vector between u and all training sample points, R is the correlation matrix of training sample points, and g is The function value at the training sample point.
[0043] Table 2 Latin hypercube sample points
[0044] Non-standard normal input Standard normal input Response value (100.00,180.00) (0.00,0.00) 1.0000 (99.31,176.29) (-0.69,-2.06) 0.9859 (101.37,178.76) (-1.37,-0.69) 0.9589 (97.25,175.06) (-2.75,-2.75) 0.9863 (100.69,180.00) (0.69,0.00) 0.9859 (97.94,183.71) (-2.06,2.06) 0.9565 (102.75,177.53) (2.75,-1.37) 0.9324 (100.00,184.94) (0.00,2.75) 0.9589 (102.06,182.47) (2.06,1.37) 0.9589 (98.63,181.24) (-1.37,0.69) 0.9861
[0045] 4. Select the regression function as a constant polynomial, the correlation function as a Gaussian kernel function, and the Kriging model obtained by training the sample points is shown in formula (20). The Kriging model of the cyclone separator function function is as follows: Figure 3 shown.
[0046]
[0047] Specifically, let α = R -1 (g-μ), Then we can get The Qth-order origin moment m Q The expression is:
[0048]
[0049] Where f(u) is the joint probability density function of U.
[0050] Assuming that the input variables are independent of each other, according to the polynomial theorem, we can get (r T α) Q The expression is:
[0051]
[0052] in, β (d) is the corresponding coefficient (d=1,2,…,D), h d (u) is the product of Q independent Gaussian kernels with unit coefficients, h d The training points of the qth unit coefficient independent Gaussian kernel in (u) (q=1,2,…,Q).
[0053] Further calculation of h d (u), we can get:
[0054] h d (u) = λ (d) c(u,M (d) / Q;QL -1 ) (9)
[0055] in,
[0056] And since f(u)=(2π) -0.5K c(u,0;I K ), where 0 is the K-dimensional zero vector, I K is the K-dimensional unit matrix, and we can further get m Q The expression is:
[0057]
[0058] Further calculation of w (d) , we can get:
[0059] w (d) =η (d) c(u,(QL -1 +IK ) -1 L -1 M (d) ;(QL -1 +I K )) (8)
[0060] in,
[0061] Therefore w (d) is also a Gaussian kernel function.
[0062] By constructing the following Gaussian density function, we avoid direct calculation The integral of
[0063]
[0064] m Q for:
[0065]
[0066] 5. Calculate the n-order origin moment of the Kriging model, that is
[0067]
[0068] Among them, l=0,1,…,Q.
[0069] Specifically, we use formula (21) to calculate The origin moment m Q :
[0070]
[0071] Further calculation using formula (22) yields The origin moment M Q :
[0072]
[0073] The origin moments of various orders of cyclone separator separation efficiency are shown in Table 3.
[0074] Table 3 Separation efficiency of each order of origin moment
[0075] Degree 1 2 3 4 5 6 origin moment 0.9842 0.9689 0.9540 0.9396 0.9255 0.9118
[0076] 6. Use the n-order origin moment to obtain the probability density function of the Kriging model through the maximum entropy method. Specifically, let The analytical expression of the probability density function ρ(z) of z is obtained by the maximum entropy method as shown in formula (23). The probability density function curve of the cyclone separator separation efficiency is shown in Figure 4 As shown, the cumulative distribution function curve is as follows Figure 5 as shown
[0077]
[0078] wherein let Considering the first n-order origin moments of the random variable z as the constraint conditions, an optimization problem for solving the probability density function ρ(z) of z can be obtained:
[0079]
[0080] Using the Lagrange multiplier method, the above optimization problem is converted into the following unconstrained optimization problem:
[0081]
[0082] An analytical expression of ρ(z) can be obtained by solving the optimization problem of S52:
[0083]
[0084] where λ i can be solved and obtained by Equation (18):
[0085]
[0086] 7. Calculate the structural reliability of the cyclone separator according to the lower threshold of the separation efficiency. Given the minimum threshold z0 of the separation efficiency of the cyclone separator, when z < z0, it is considered that the cyclone separator fails, and the reliability R of the cyclone separator is obtained through Equation (19):
[0087]
[0088] Specifically, the minimum threshold z0 of the separation efficiency of the cyclone separator is 0.95, and the reliability of the cyclone separator calculated by Equation (24) is 0.9641.
[0089]
[0090] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.
Claims
1. A reliability analysis method for cyclone separator based on Kriging model. It is characterized in that include: S1 Analyze the uncertain factors existing in the actual working process of cyclone separator; S2 obtains sample points based on ANSYS simulation; S3 uses sample points to train and obtain the Kriging model of the cyclone separator structure function function; S4 uses the algorithm in this paper to analyze the n-order origin moment of the Kriging proxy model; S5 uses the maximum entropy method to obtain the probability density function of the cyclone separation efficiency using the obtained n-order origin moment; S6 solves the structural reliability of the cyclone separator according to the given lower limit threshold of separation efficiency.
2. The cyclone separator reliability analysis method based on Kriging method according to claim 1, Features: The S1 analyzes the uncertain factors existing in the actual working process of the cyclone separator by determining the structural function function of the cyclone separator, and the function is: Y=G(X) (1) Where X = (X 1 ,X 2 ,…,X K ) T (K is the number of parameters) is the input variable that affects the uncertainty of its performance. The input variable X is converted into a standard normal variable U by formula (2), and the structural function function Y=G(U) is obtained. in, For X i The cumulative distribution function, Φ -1 is the inverse of the cumulative distribution function of the standard normal distribution.
3. The reliability analysis method of cyclone separator based on Kriging method according to claim 1, Features: S2 is to obtain a number of training sample points and calculate the response value through the Latin hypercube sampling method. The specific process is to obtain the standard normal distribution input variable matrix u through S1 transformation in , According to the extracted training sample points, the corresponding functional response vector g is calculated through ANSYS simulation: g=[g (1) ,g (2) ,…,g (N) ] T (4) The correlation function of the Kriging model is selected as the Gaussian kernel function, as shown in the formula: Among them, u (i) is the i-th sample point of U (i=1,2,...,N), u k (i) for u (i) The k-th dimension component (k=1,2,...,K), l k is the correlation length of U in the k-th dimension, L is a K×K symmetric matrix with the main diagonal elements being l k 2 ,c(u (i) ,u (j) ; L -1 ) is called the unit coefficient independent Gaussian kernel. Finally, the regression function is selected as a constant polynomial. The Kriging model of the cyclone separator function function is obtained by training the random field with sample points: Where μ is The mean of , r is the correlation vector between u and all training sample points, R is the correlation matrix of training sample points, and g is The function value at the training sample point.
4. The reliability analysis method of cyclone separator based on Kriging model according to claim 1, Features: The Kriging model of S3 is as follows: -1 (g-μ), Then we can get The Qth-order origin moment m Q The expression is: Where f(u) is the joint probability density function of U. Assuming that the input variables are independent of each other, according to the polynomial theorem, we can get (r T α) Q The expression is: in, β (d) is the corresponding coefficient (d=1,2,…,D), h d (u) is the product of Q independent Gaussian kernels with unit coefficients, h d The training points of the qth unit coefficient independent Gaussian kernel in (u) (q=1,2,…,Q). Further calculation of h d (u), we can get: h d (u)=λ (d) c(u,M (d) / Q;QL -1 ) (9) in, Since f(u)=(2π) -0.5K c(u,0;I K ), where 0 is the K-dimensional zero vector, I K is the K-dimensional unit matrix, and we can further get m Q The expression is: Further calculation of w (d) , we can get: w (d) =η (d) c(u,(QL -1 +I K ) -1 L -1 M (d) ;(QL -1 +I K )) (1) in, Therefore w (d) is also a Gaussian kernel function. By constructing the following Gaussian density function, we avoid direct calculation The integral of m Q for:
5. The reliability analysis method of cyclone separator based on Kriging method according to claim 1, Features: The n-order origin moment of the S4 Kriging model: Among them, l=0,1,…,Q.
6. The reliability analysis method of cyclone separator based on Kriging model according to claim 1, Features: The S5 is command Considering the first n-order origin moments of the random variable z as constraints, we can get the optimization problem of solving the probability density function ρ(z) of z: Using the Lagrange multiplier method, the above optimization problem is transformed into the following unconstrained optimization problem: By solving the optimization problem of S52, the analytical expression of ρ(z) can be obtained: Among them, λ i It can be solved by formula (18):
7. The reliability analysis method of cyclone separator based on Kriging method according to claim 1, Features: S6 is to set the minimum threshold z 0 , when z <z 0 When , the cyclone separator is considered to be failed, and the reliability R of the cyclone separator is obtained by formula (19):
