Random field modeling method for quantizing initial defects in spherical shell structures

Abstract

The present invention discloses a random field modeling method for quantizing initial defects of a spherical shell structure, the method includes: determining simulation spherical shell data corresponding to a target spherical shell, including spherical shell geometric information and spherical shell material information; determining a target correlation function and a representative discrete point corresponding to the target spherical shell according to the spherical shell geometric information and the spherical shell material information; determining a correlation coefficient matrix according to the target correlation function and the representative discrete point; decomposing the correlation coefficient matrix feature value to obtain a standard random field; and performing random field transformation on the standard random field according to the random field transformation formula to obtain a target random field corresponding to the target spherical shell. The present invention can accurately quantize the inherent correlation of the initial defect of the spherical shell by determining the target correlation function and the representative discrete point according to the simulation spherical shell data, introducing the geodesic distance into the target correlation function, and establishing the spherical shell random field according to the feature value decomposition method, and further improve the accuracy of the spherical shell performance analysis.

Description

[Technical field]

[0001] The present invention relates to the technical field of structural engineering, specifically to a random field modeling method for quantizing initial imperfections in a spherical shell structure. [Background technology]

[0002] Due to its reasonable force-bearing performance, the spherical shell structure is widely used in the fields of architecture, aerospace, industry, etc. However, in actual construction, various initial defects, such as geometric defects and material defects, inevitably exist in the spherical shell, which significantly reduces the overall bearing capacity, stability, fatigue life, etc. of the spherical shell structure.

[0003] Since the distribution of the initial defects of the spherical shell structure is random, it is generally characterized as a random field on the spherical shell. In the prior art, many random field simulation methods have been developed, but these methods are only applicable to Euclidean spaces, such as planes, cubes, etc., and are not applicable to curved structures such as spherical shells. When the methods in the prior art are used to generate the random field on the spherical shell, the actual correlation between any two points on the spherical shell cannot be characterized, and the initial defects of the spherical shell cannot be accurately quantized, which affects the performance analysis and evaluation of the spherical shell. Summary of the Invention [Problem to be solved by the invention]

[0004] In view of the above deficiencies of the prior art, the technical problem solved by the present invention is how to improve the quantization accuracy of initial defects of a spherical shell. [Means for solving the problem]

[0005] In order to solve at least one of the technical problems proposed above, the present invention discloses a random field modeling method for quantizing initial defects in a spherical shell structure.

[0006] According to a first aspect of the present disclosure, there is provided a random field modeling method for quantizing an initial defect in a spherical shell structure, the method comprising: Determining simulation shell data for constructing a target shell, the simulation shell data including shell geometric information, shell material information, and shell defect information; determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information; determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information; performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field; and performing random field transformation on the standard random field based on a random field transformation formula to obtain a target random field corresponding to the spherical shell defect information.

[0007] In some possible embodiments, the step of determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information includes: performing a modeling process based on the geometric information of the spherical shell and the spherical shell material information to obtain the target spherical shell; The method includes dividing the target shell into a mesh and obtaining the representative discrete points.

[0008] In some possible embodiments, the step of determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information includes: obtaining a target length parameter corresponding to the target shell based on the spherical shell defect information, the target length parameter characterizing range information corresponding to the target random field; calculating a target distance parameter based on the geometric information of the spherical shell; and constructing the target correlation function based on the target length parameter and the target distance parameter.

[0009] In some possible embodiments, the step of determining the representative discrete points, the spherical shell defect information, and the correlation coefficient matrix based on the target correlation function includes: performing a probability distribution type judgment on the spherical shell defect information to obtain a type judgment result; When the type determination result indicates that the spherical shell defect information corresponds to a normal distribution, sequentially calculating a first correlation coefficient between any two representative discrete points among a plurality of representative discrete points based on the target correlation function; and constructing the correlation coefficient matrix based on first correlation coefficients corresponding to the plurality of representative discrete points.

[0010] In some possible embodiments, the method further comprises: determining a correlation coefficient calculation formula based on the target correlation function when the type determination result indicates that the spherical shell defect information corresponds to a log-normal distribution; calculating a second correlation coefficient between any two of the representative discrete points based on the correlation coefficient calculation formula; and constructing the correlation coefficient matrix based on second correlation coefficients corresponding to the plurality of representative discrete points.

[0011] In some possible embodiments, the step of performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field includes the steps of: performing feature value decomposition on the correlation coefficient matrix to obtain a feature value matrix and a feature vector matrix corresponding to the correlation coefficient matrix; Determining a target number of random variables based on a reconstruction threshold calculation formula and a reconstruction threshold parameter, wherein the target number of random variables is smaller than the number of the representative discrete points; determining a target feature value matrix, a target feature vector matrix, and target random vector parameters based on the target number of random variables, the feature value matrix, and the feature vector matrix; and constructing the standard random field based on the target feature value matrix, the target feature vector matrix, and the target random vector parameters.

[0012] In some possible embodiments, performing a random field transformation on the standard random field based on a random field transformation formula to obtain a target random field corresponding to the target shell includes: Obtaining average values ​​and standard differences corresponding to the shell defect information; inputting the mean value, the standard difference and the standard random field into the random field transformation formula to obtain the target random field.

[0013] According to a second aspect of the present disclosure, there is provided a random field modeling apparatus for quantizing an initial defect in a spherical shell structure, the apparatus comprising: a simulation information acquisition module for acquiring simulation shell data, including shell geometric information, shell material information, and shell defect information, for constructing a target shell; a spherical shell parameter determination module for determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information; a correlation coefficient matrix determination module for determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information; a feature value decomposition module for performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field; and a random field determining module for performing random field transformation on the standard random field based on a random field transformation formula to obtain a target random field corresponding to the spherical shell defect information.

[0014] According to a third aspect of the present disclosure, there is provided an electronic device including a processor and a memory, wherein the memory stores at least one instruction and at least one program, and the at least one instruction and the at least one program are loaded and executed by the processor to realize the random field modeling method for quantizing an initial defect in a spherical shell structure described above.

[0015] According to a fourth aspect of the present disclosure, there is provided a computer storage medium having at least one instruction and at least one program stored therein, the at least one instruction and the at least one program being loaded and executed by a processor to realize the random field modeling method for quantizing an initial defect in a spherical shell structure described above. Effect of the Invention

[0016] By implementing the present invention, the following beneficial effects are obtained.

[0017] In the present invention, a target correlation function and a representative discrete point are determined based on the simulation spherical shell data corresponding to the target spherical shell, where the target distance parameter related to the target correlation function is determined based on the specific simulation spherical shell data, and a correlation coefficient matrix is ​​determined based on the target correlation function and the representative discrete point, and the correlation coefficient matrix feature value is further decomposed, so that the inherent correlation of the initial defect of the target spherical shell can be accurately quantized, and different random field transformation formulas are used for different initial probability distribution types, so as to ensure the association and matching degree between the random field and the spherical shell defect, thereby improving the accuracy of the initial defect quantization of the spherical shell, and further improving the reliability of the spherical shell performance analysis. [Brief description of the drawings]

[0018] [Figure 1] 1 is a flowchart corresponding to a random field modeling method according to an embodiment of the present invention. [Diagram 2] 4 is a flow chart corresponding to determining representative discrete points according to an embodiment of the present invention; [Diagram 3] 4 is a flow chart corresponding to determining a target correlation function according to an embodiment of the present invention. [Figure 4] 4 is a first flowchart corresponding to determining a correlation coefficient matrix according to an embodiment of the present invention; [Diagram 5] 4 is a second flowchart corresponding to determining a correlation coefficient matrix according to an embodiment of the present invention; [Figure 6] 4 is a flow chart corresponding to determining a standard random field according to an embodiment of the present invention. [Figure 7] 4 is a flowchart corresponding to determining a target random field according to an embodiment of the present invention. [Figure 8] FIG. 2 is a first schematic diagram of a spherical shell normal distribution random field according to one embodiment of the present invention; [Figure 9] FIG. 2 is a second schematic diagram of a spherical shell normal distribution random field according to one specific embodiment of the present invention; [Figure 10] FIG. 13 is a third schematic diagram of a spherical shell normal distribution random field according to a specific embodiment of the present invention. [Figure 11] FIG. 2 is a first schematic diagram of a spherical shell log-normal distribution random field according to one specific embodiment of the present invention; [Figure 12] FIG. 2 is a second schematic diagram of a spherical shell log-normal distribution random field according to one specific embodiment of the present invention. [Figure 13] FIG. 13 is a third schematic diagram of a spherical shell log-normal distribution random field according to one specific embodiment of the present invention. [Figure 14] 1 is a structural schematic diagram corresponding to a random field modeling apparatus according to an embodiment of the present invention; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] 1 is a flowchart corresponding to a random field modeling method according to an embodiment of the present invention, and the execution body can be any terminal or electronic device that can realize the random field modeling method for quantizing initial defects of a spherical shell structure, such as any device with computing power and capable of realizing the random field modeling method, such as a computer, a server, an embedded system, a smart phone, a tablet computer or a server cluster. In one specific embodiment, as shown in FIG. 1, the random field modeling method for quantizing initial defects of a spherical shell structure includes the following steps:

[0020] Step S101: Determine simulation shell data for constructing a target shell, including shell geometric information, shell material information, and shell defect information; In one specific embodiment, the simulation shell data may include spherical shell geometric information and spherical shell material information, and may further include spherical shell defect information, where the spherical shell geometric information may include at least the spherical shell radius r and the spherical shell thickness t, the spherical shell material information may include the spherical shell material information and the spherical shell elastic modulus E, and the spherical shell defect information may be initial defect information, which may be determined before constructing the target spherical shell, and may be determined by methods such as numerical simulation, statistical analysis, etc.

[0021] Step S102: determining a target correlation function and a representative discrete point corresponding to the target spherical shell based on the geometric information of the spherical shell, the spherical shell material information, and the spherical shell defect information; In one specific embodiment, a target spherical shell may be constructed first based on the geometric information of the spherical shell and the material information of the spherical shell, and then the corresponding target correlation function and the representative discrete points may be obtained based on the target spherical shell.

[0022] How to determine the target correlation function and the representative discrete points can be described in conjunction with Figures 2 and 3 respectively. In particular, Figure 2 is a flowchart corresponding to determining the target correlation function according to an embodiment of the present invention. As shown in Figure 2, for determining the representative discrete points, step S102 may include:

[0023] Step S201: A modeling process is performed based on the geometric information of the spherical shell and the spherical shell material information to obtain the target spherical shell; Specifically, the ABAQUS finite element software can be used to combine the radius, thickness, material and elastic modulus of the spherical shell to realize the construction of the target spherical shell.

[0024] Step S202: The target spherical shell is divided into meshes to obtain the representative discrete points.

[0025] JPEG0007650559000002.jpg21165

[0026] Specifically, the target spherical shell is divided into a mesh, and the target spherical shell is divided into a plurality of geometric units, and each vertex of the geometric unit becomes a node, and each node may be determined as a corresponding representative discrete point, and the number and positions of the nodes may depend on the density of the mesh division and the complexity of the target spherical shell.

[0027] Furthermore, FIG. 3 shows a flowchart corresponding to determining the target correlation function according to an embodiment of the present invention, and as shown in FIG. 3, for determining the target correlation function, step S102 may include:

[0028] Step S301: obtain a target length parameter corresponding to the target spherical shell according to the spherical shell defect information, and the target length parameter characterizes range information corresponding to the target random field; In one specific embodiment, the target length parameter c may be a correlation length, and the target length parameter may characterize range information corresponding to the target random field, i.e., the correlation length c may be used to represent the correlation range of the random field, and the target length parameter may be preset or adaptively modified according to different simulation shell data.

[0029] Step S302: Calculate a target distance parameter based on the geometric information of the spherical shell; In one specific embodiment, the target distance parameter g may be a geodesic distance, and the geodesic distance between any two points on the target spherical shell is the shortest arc length passing through the two points on the target spherical shell. A coordinate system may be established based on the spherical shell radius r and the spherical shell thickness t to determine the center position of the spherical shell, and further any two points on the target spherical shell may be determined, and the geodesic distance, i.e., the target distance parameter g, may be obtained by determining the shortest arc length based on the coordinate values ​​of the any two points.

[0030] Step S303: Construct the target correlation function based on the target length parameter and the target distance parameter.

[0031] JPEG0007650559000003.jpg16165JPEG0007650559000004.jpg5164where g is the target distance parameter and c is the target length parameter.

[0032] Step S103: determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information; In one specific embodiment, the correlation coefficient matrix may be a correlation coefficient between any two representative discrete points, and the correlation coefficient may be determined based on the target correlation function and the geodesic distance between any two representative discrete points. The correlation coefficient matrix R may be as follows: JPEG0007650559000005.jpg20164JPEG0007650559000006.jpg6165

[0033] Specifically, FIG. 4 is a first flowchart corresponding to determining the correlation coefficient matrix according to an embodiment of the present invention. As shown in FIG. 4, for determining the correlation coefficient matrix, step S103 may include:

[0034] Step S401: Perform a probability distribution type judgment on the shell defect information to obtain a type judgment result; In one specific embodiment, the shell defect information may characterize an initial defect of the target shell, which may be the geometry of the target shell (the geometry is the distance that the actual position of the target shell deviates from the ideal position of the target shell), the elastic modulus of the target shell, and the shell thickness of the target shell, etc.

[0035] The initial defects of the target shell may include the probability distribution type, the mean value, the standard difference, the target correlation function and the correlation length, etc., and may generally be determined by examining the literature and existing research. The shell defect information of the target shell may be determined after the target shell is constructed.

[0036] JPEG0007650559000007.jpg26165

[0037] In one specific embodiment, the probability distribution type of the spherical shell defect information may include at least a normal distribution and a log-normal distribution, and since different probability distribution types correspond to different correlation coefficient determination methods, before calculating the correlation coefficient between any two representative discrete points, it is necessary to first determine the probability distribution type contained in the spherical shell defect information.

[0038] Step S402: when the type determination result indicates that the spherical shell defect information corresponds to a normal distribution, sequentially calculate a first correlation coefficient between any two representative discrete points among a plurality of representative discrete points based on the target correlation function; In one specific embodiment, when the type determination result indicates that the probability distribution type corresponding to the spherical shell defect information is a normal distribution, a correlation coefficient may be determined using a target correlation function, and the correlation coefficient determined using the target correlation function is a first correlation coefficient, where the first correlation coefficient and the second correlation coefficient are used to represent correlation coefficients corresponding to different probability distributions, and do not distinguish the correlation coefficients in terms of procedure or any other meaning.

[0039] JPEG0007650559000008.jpg14165JPEG0007650559000009.jpg5163JPEG0007650559000010.jpg6165JPEG0007650559000011.jpg6163Step S403: Construct the correlation coefficient matrix based on first correlation coefficients corresponding to the plurality of representative discrete points.

[0040] Specifically, after determining first correlation coefficients corresponding to any two of the plurality of representative discrete points, a correlation coefficient matrix is ​​constructed based on the plurality of first correlation coefficients.

[0041] In another specific embodiment, the probability distribution type corresponding to the spherical shell defect information may be a log-normal distribution, and Figure 5 is a second flowchart corresponding to determining the correlation coefficient matrix according to an embodiment of the present invention. As shown in Figure 5, when the type judgment result indicates that the probability distribution type corresponding to the spherical shell defect information is a log-normal distribution, the method further includes:

[0042] Step S501: when the type determination result indicates that the spherical shell defect information corresponds to a log-normal distribution, a correlation coefficient calculation formula is determined based on the target correlation function; JPEG0007650559000012.jpg21165JPEG0007650559000013.jpg12163JPEG0007650559000014.jpg10165

[0043] Step S502: Sequentially calculating a second correlation coefficient between any two representative discrete points among the plurality of representative discrete points based on the correlation coefficient calculation formula; Step S503: Construct the correlation coefficient matrix based on second correlation coefficients corresponding to the plurality of representative discrete points.

[0044] Specifically, the correlation coefficient determined using the correlation coefficient formula is a second correlation coefficient, and the second correlation coefficient is a correlation coefficient between any two representative discrete points among the plurality of representative discrete points. After obtaining the second correlation coefficient, a correlation coefficient matrix is ​​constructed based on the second correlation coefficient.

[0045] Step S104: Perform feature value decomposition on the correlation coefficient matrix to obtain a standard random field; In one specific embodiment, for the correlation coefficient matrix R, the following matrix calculation formula may exist: JPEG0007650559000015.jpg16163JPEG0007650559000016.jpg21165

[0046] Specifically, FIG. 6 is a flowchart corresponding to determining a standard random field according to an embodiment of the present invention. As shown in FIG. 6, step S104 may include:

[0047] Step S601: Perform feature value decomposition on the correlation coefficient matrix to obtain a feature value matrix and a feature vector matrix corresponding to the correlation coefficient matrix; In one specific embodiment, the above matrix calculation formula is used to perform feature value decomposition on the correlation coefficient matrix R to obtain a feature value matrix and a feature vector matrix corresponding to the correlation coefficient matrix, where the feature vector matrix may include a first feature vector matrix and a second feature vector matrix, and the second feature vector matrix is ​​a transpose matrix of the first feature vector matrix.

[0048] Step S602: determine a target number of the random variables according to a reconstruction threshold calculation formula and a reconstruction threshold parameter, and the target number of the random variables is smaller than the number of the representative discrete points; In one specific embodiment, there is a reconstruction threshold calculation formula: JPEG0007650559000017.jpg11163JPEG0007650559000018.jpg32165

[0049] Step S603: Determine a target feature value matrix, a target feature vector matrix and a target random vector parameter according to the target number of random variables, the feature value matrix and the feature vector matrix; JPEG0007650559000019.jpg38165

[0050] Step S604: Construct the standard random field according to the target feature value matrix, the target feature vector matrix and the target vector parameters.

[0051] JPEG0007650559000020.jpg16167JPEG0007650559000021.jpg6163The standard random field Y0 is a vector with n rows and 1 column, and corresponds one-to-one to the n representative discrete points of the target shell. The random field value is assigned to the corresponding node or integral point in the finite element model by the discrete point number, thereby realizing the generation of a finite element model that takes the random field into consideration. By repeatedly generating the column vector of the standard normal distribution several times, different standard normal distribution random fields can be obtained.

[0052] It should be noted that, of course, when the probability distribution type corresponding to the spherical shell defect information is normal distribution or log-normal distribution, the standard random field may be determined using the above steps S601 to S604 in either case.

[0053] Step S105: According to a random field transformation formula, perform random field transformation on the standard random field to obtain a target random field corresponding to the spherical shell defect information.

[0054] In one specific embodiment, the random field transformation formula may include a first random field transformation formula and a second random field transformation formula, and the corresponding transformation formula may be determined according to a probability distribution type corresponding to the spherical shell defect information to generate a target random field corresponding to the target spherical shell, where the probability distribution type corresponding to the first random field transformation formula is a normal distribution, and the probability distribution type corresponding to the second random field transformation formula is a log-normal distribution.

[0055] Specifically, FIG. 7 is a flowchart corresponding to determining a target random field according to an embodiment of the present invention. As shown in FIG. 7, step S105 may include:

[0056] Step S701: Obtain the average value and standard difference corresponding to the spherical shell defect information; Step S702: The mean value, the standard difference and the standard random field are input into the random field transformation formula to obtain the target random field.

[0057] In one specific embodiment, when the probability distribution type is normal distribution, a first random field transformation formula needs to be used to determine the target random field Y. The first random field transformation formula is as follows: In another specific embodiment, when the probability distribution type is a log-normal distribution, the second random field transformation formula needs to be used to determine the target random field Z. The second random field transformation formula is as follows: JPEG0007650559000023.jpg16164JPEG0007650559000024.jpg10165

[0058] 8 to 13 are examples of the realization of the initial defect random field of spherical shell normal distribution and log-normal distribution corresponding to the specific embodiment of the present invention, each situation includes three sets of different correlation lengths, and each correlation length includes two random sample realizations. In the following, the random field modeling method for quantizing the initial defect of the spherical shell structure of the present invention will be described in conjunction with FIG. 8 to FIG. 13, taking the normal distribution and the log-normal distribution as specific embodiments respectively.

[0059] Example 1: When the probability distribution type of the spherical shell defect information corresponding to the target spherical shell is normal distribution, 8 to 10 are examples of realizing a random field corresponding to a spherical shell when the probability distribution type (all normal distributions), the mean value and standard difference are the same, and the correlation lengths c are different, where each numerical unit may correspond to mm, the correlation length c in FIG. 8 may be 100, the correlation length c in FIG. 9 may be 200, and the correlation length c in FIG. 10 may be 400, and the mean values ​​in FIGS. 8 to 10 are all 0 and the standard differences are all 2. As can be seen from FIGS. 8 to 10, different correlation lengths have different effects on the random field corresponding to the spherical shell. Referring to FIGS. 8 to 10, a random field modeling method for quantizing initial defects of a spherical shell structure as a normal distribution according to an embodiment of the present invention includes the following steps:

[0060] Step 1: Determine the geometric and material information of the spherical shell to be simulated. Take the pressure shell of a manned submersible as an example. According to literature research, the radius of this spherical shell is r=1050mm, the thickness t=21mm, the material is 980 steel, and the elastic modulus is 2×105MPa.

[0061] JPEG0007650559000025.jpg37165JPEG0007650559000026.jpg5163Here, the correlation length c may be set to 200 mm.

[0062] Step 3: Determine the representative discrete points of the spherical shell, and take the ABAQUS finite element analysis software as the platform to establish a finite element model of the spherical shell to be simulated, divide the model into a triangular mesh, the mesh size is about 50 mm, the model after discretization contains 5722 nodes and 11440 units, the defect type is a geometric defect, and select the node as the representative discrete point.

[0063] JPEG0007650559000027.jpg14165JPEG0007650559000028.jpg6163JPEG0007650559000029.jpg10165JPEG0007650559000030.jpg5163The correlation coefficient matrix is ​​a logarithmic matrix of (5722 × 5722), and is specifically as follows: JPEG0007650559000031.jpg20163Step 5: Perform feature value decomposition based on the correlation coefficient matrix to generate a standard normal distribution random field corresponding to the spherical shell.

[0064] JPEG0007650559000032.jpg22165

[0065] The reconstruction threshold value β is determined to be 95%. When c is 200 mm, the following equation is satisfied: JPEG0007650559000033.jpg11163JPEG0007650559000034.jpg10165JPEG0007650559000035.jpg6163JPEG0007650559000036.jpg15165

[0066] JPEG0007650559000037.jpg16165

[0067] Example 2: When the probability distribution type of the shell defect information corresponding to the target shell is a log-normal distribution, 11 to 13 are examples of realizing a random field corresponding to a spherical shell when the probability distribution type (all are log-normal distribution), the mean value and standard difference are the same, and the correlation length c is different. Each numerical unit may correspond to MPa. The correlation length c in FIG. 11 may be 100 mm, the correlation length c in FIG. 12 may be 200 mm, and the correlation length c in FIG. 13 may be 400 mm. Furthermore, the mean values ​​in FIG. 11 to FIG. 13 are all 200,000, and the standard differences are all 10,000. As can be seen from FIG. 11 to FIG. 13, different correlation lengths have different effects on the random field corresponding to the spherical shell. Referring to FIG. 11 to FIG. 13, a random field modeling method for quantizing initial defects of a spherical shell structure as a normal distribution according to an embodiment of the present invention includes the following steps.

[0068] Step 1: Determine the geometric and material information of the spherical shell to be simulated. Take the pressure shell of a manned submersible as an example. According to literature research, the radius of this spherical shell is r=1050mm, the thickness is t=21mm, the material is 980 steel, and the elastic modulus is 2×105MPa.

[0069] JPEG0007650559000038.jpg32165JPEG0007650559000039.jpg5163Here, the correlation length c may be set to 100 mm.

[0070] Step 3: Determine the representative discrete point of the spherical shell. Take ABAQUS finite element analysis software as the platform to establish a finite element model of the spherical shell. Divide the model into a triangular mesh, the mesh size is about 50 mm, the model after discretization contains 5722 nodes, 11440 units, the defect type is elastic modulus, and select the unit integration point as the representative discrete point.

[0071] JPEG0007650559000040.jpg14165JPEG0007650559000041.jpg11163JPEG0007650559000042.jpg5165JPEG0007650559000043.jpg6163The correlation coefficient matrix is ​​a logarithmic matrix of (11440 x 11440), and is specifically as follows: JPEG0007650559000044.jpg20163Step 5: Decompose the correlation coefficient matrix feature values ​​and generate a standard normal distribution random field.

[0072] JPEG0007650559000045.jpg22165

[0073] The reconstruction threshold value β is determined to be 95%. When c = 100 mm, the following equation is satisfied: JPEG0007650559000046.jpg12163JPEG0007650559000047.jpg10165JPEG0007650559000048.jpg6163JPEG0007650559000049.jpg15165

[0074] Step 6: Transform the standard normal distribution random field generated in step 5 to generate a target distribution random field of the initial defect, that is, the log-normal distribution random field of the elastic modulus initial defect is as follows: JPEG0007650559000050.jpg27165

[0075] Compared with the prior art, the technical solution disclosed in the present invention has the following beneficial effects: 1. By introducing geodesic distance, a spherical shell random field can be established based on the feature value decomposition method, and the inherent correlation of the initial defects of the spherical shell can be accurately quantized. 2. By setting the reconstruction threshold, the feature value retention can be selected according to the actual situation, and the number of random variables used in the random field numerical simulation can be greatly reduced. 3. By generating a non-normal distribution random field through direct equal probability transformation, the correlation function of the random field is changed. The present invention directly realizes the simulation of a log-normal random field by combining the change of the cumulative distribution function through the transformation of the correlation function form, and both the probability distribution and the correlation function can be consistent with the target.

[0076] An embodiment of the present invention further provides a random field modeling apparatus for quantizing an initial defect of a spherical shell structure, as shown in FIG. 14, the apparatus includes: a simulation information acquisition module 1410 for acquiring simulation shell data, including shell geometric information, shell material information, and shell defect information, for constructing a target shell; a spherical shell parameter determination module 1420 for determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information; a correlation matrix determination module 1430 for determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information; a feature value decomposition module 1440 for performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field; and a random field determination module 1450 for performing random field transformation on the standard random field based on a random field transformation formula to obtain a target random field corresponding to the spherical shell defect information.

[0077] In some other embodiments, the spherical shell parameter determination module 1420 comprises: a first parameter determination module for obtaining a target length parameter corresponding to the target spherical shell based on the spherical shell defect information, the target length parameter characterizing range information corresponding to the target random field; a second parameter determination module for calculating a target distance parameter based on the geometric information of the spherical shell; and a correlation function determination module for constructing the target correlation function based on the target length parameter and the target distance parameter.

[0078] In some other embodiments, the spherical shell parameter determination module 1420 comprises: a spherical shell modeling processing module for performing a modeling process based on the geometric information of the spherical shell and the spherical shell material information to obtain the target spherical shell; and a discrete point determination module for dividing the target spherical shell into a mesh and obtaining the representative discrete points.

[0079] In some other embodiments, the correlation matrix determination module 1430 comprises: a distribution type judgment module for performing a probability distribution type judgment on the spherical shell defect information to obtain a type judgment result; a first coefficient determination module for sequentially calculating a first correlation coefficient between any two representative discrete points among a plurality of representative discrete points based on the target correlation function when the type determination result indicates that the spherical shell defect information corresponds to a normal distribution; and a first matrix construction module for constructing the correlation coefficient matrix based on first correlation coefficients corresponding to the plurality of representative discrete points.

[0080] In some further embodiments, the apparatus further comprises: a formula determination module for determining a correlation coefficient calculation formula based on the target correlation function when the type determination result indicates that the spherical shell defect information corresponds to a log-normal distribution; a second coefficient determining module for sequentially calculating a second correlation coefficient between any two representative discrete points among the plurality of representative discrete points based on the correlation coefficient calculation formula; and a second matrix construction module for constructing the correlation coefficient matrix based on second correlation coefficients corresponding to the plurality of representative discrete points.

[0081] In some other embodiments, the feature value decomposition module 1440 comprises: a matrix determination module for performing feature value decomposition on the correlation coefficient matrix to obtain a feature value matrix and a feature vector matrix corresponding to the correlation coefficient matrix; A random variable number determination module for determining a target number of random variables based on a reconstruction threshold calculation formula and a reconstruction threshold parameter, the target number of random variables being smaller than the number of the representative discrete points; a third parameter determination module for determining a target feature value matrix, a target feature vector matrix, and a target random vector parameter based on the target number of random variables, the feature value matrix, and the feature vector matrix; and a first random field construction module for constructing the standard random field based on the target feature value matrix, the target feature vector matrix and the target random vector parameters.

[0082] In some other embodiments, the random field determination module 1450 comprises: a numerical determination module for obtaining a mean value and a standard difference corresponding to the spherical shell defect information; and a second random field construction module for inputting the mean value, the standard difference and the standard random field into the random field transformation formula to obtain the target random field.

[0083] The apparatus and method embodiments in the apparatus embodiment are based on the same inventive concept for implementing the random field modeling method for quantizing the initial defects of the spherical shell structure described above.

[0084] An embodiment of the present invention further provides an electronic device, the device including a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the at least one instruction, the at least one program, code set or instruction set being loaded and executed by the processor to realize a random field modeling method for quantizing an initial defect in a spherical shell structure as described in any one of the method embodiments.

[0085] An embodiment of the present invention further provides a storage medium, which may be installed on a server for storing at least one instruction, at least one program, code set or instruction set for realizing the above-mentioned method embodiment, and the at least one instruction, the at least one program, code set or instruction set is loaded and executed by the processor to realize the random field modeling method for quantizing initial defects in a spherical shell structure described in any one of the method embodiments.

[0086] The computer-readable storage medium may be a tangible device capable of maintaining and storing instructions for use by an instruction execution device. The computer-readable storage medium may be, for example, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include portable computer disks, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), static random access memories (SRAMs), portable compressed disk read-only memories (CD-ROMs), digital versatile disks (DVDs), memory sticks, floppy disks, mechanically encoded devices such as puncture cards or groove-in-projection structures on which instructions are stored, and any suitable combination of the above. As used herein, computer-readable storage medium is not to be construed as a momentary signal per se, such as an electric wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., light pulses through a fiber optic cable), or an electrical signal transmitted through an electrical wire.

[0087] The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium into each computing / processing device, or may be downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0088] Computer program instructions for carrying out the operations of the present disclosure may include assembly instructions, instruction set architecture instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or target code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C+, and the like, and general procedural programming languages ​​such as "C" or similar. The computer readable program instructions may be executed entirely on the user computer, partially on the user computer, as a separate software package, partially on the user computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., connected via the Internet using an Internet Service Provider). In some embodiments, state information of the computer readable program instructions can be used to personalize and customize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), such that the electronic circuit can execute the computer readable program instructions, thereby implementing aspects of the present disclosure.

[0089] As can be seen from the above embodiments of the present invention, in the present invention, a target correlation function and a representative discrete point are determined based on the simulation spherical shell data corresponding to the target spherical shell, where the target distance parameter related to the target correlation function is determined based on the specific simulation spherical shell data, a correlation coefficient matrix is ​​determined based on the target correlation function and the representative discrete point, and the correlation coefficient matrix feature value is further decomposed, so that the inherent correlation of the initial defect of the target spherical shell can be accurately quantized; and by using different random field transformation formulas for different initial probability distribution types, the association and matching degree between the random field and the spherical shell defect can be guaranteed, so that the accuracy of the initial defect quantization of the spherical shell can be improved, and the reliability of the performance analysis of the spherical shell can be further improved.

Claims

1. A processor-implemented random field modeling method for quantizing an initial defect in a spherical shell structure, comprising: Determining simulation shell data for constructing a target shell, the simulation shell data including shell geometric information, shell material information, and shell defect information; Determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information, where g is the target distance parameter and c is the target length parameter; determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information, The correlation coefficient matrix R may be: performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field; performing random field transformation on the standard random field based on a random field transformation formula to obtain a target random field corresponding to the spherical shell defect information; Including, the number of the representative discrete points is plural, Determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information includes: performing a probability distribution type judgment on the spherical shell defect information to obtain a type judgment result; When the type determination result indicates that the spherical shell defect information corresponds to a normal distribution, a first correlation coefficient between any two representative discrete points among a plurality of representative discrete points is calculated in sequence based on the target correlation function, r is the shell radius; constructing the correlation coefficient matrix based on first correlation coefficients corresponding to the plurality of representative discrete points; When the type determination result indicates that the spherical shell defect information corresponds to a log-normal distribution, a correlation coefficient calculation formula is determined based on the target correlation function, calculating a second correlation coefficient between any two of the representative discrete points based on the correlation coefficient calculation formula; constructing the correlation coefficient matrix based on second correlation coefficients corresponding to the plurality of representative discrete points; Including, The step of performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field includes the steps of: performing feature value decomposition on the correlation coefficient matrix to obtain a feature value matrix and a feature vector matrix corresponding to the correlation coefficient matrix; Determining a target number of random variables based on a reconstruction threshold calculation formula and a reconstruction threshold parameter, wherein the target number of random variables is smaller than the number of the representative discrete points; determining a target feature value matrix, a target feature vector matrix, and target random vector parameters based on the target number of random variables, the feature value matrix, and the feature vector matrix; constructing the standard random field based on the target feature value matrix, the target feature vector matrix and the target random vector parameters; Including, According to the random field transformation formula, performing random field transformation on the standard random field to obtain a target random field corresponding to the target shell, Obtaining average values ​​and standard differences corresponding to the shell defect information; inputting said mean value, said standard difference and said standard random field into said random field transformation formula to obtain said target random field; Including, A random field modeling method for quantizing an initial defect in a spherical shell structure, comprising:

2. Determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information includes: performing a modeling process based on the geometric information of the spherical shell and the spherical shell material information to obtain the target spherical shell; Dividing the target shell into a mesh and obtaining the representative discrete points; Including, 2. The random field modeling method for quantizing an initial defect in a spherical shell structure according to claim 1.

3. Determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information includes: obtaining a target length parameter corresponding to the target shell based on the spherical shell defect information, the target length parameter characterizing range information corresponding to the target random field; calculating a target distance parameter based on the geometric information of the spherical shell; constructing the target correlation function based on the target length parameter and the target distance parameter; Including, The random field modeling method for quantizing initial defects in a spherical shell structure according to claim 2.

4. A random field modeling device for quantizing an initial defect in a spherical shell structure, comprising: a simulation information acquisition module for acquiring simulation shell data, including shell geometric information, shell material information, and shell defect information, for constructing a target shell; a spherical shell parameter determination module for determining a target correlation function and a representative discrete point corresponding to the spherical shell defect information based on the spherical shell geometric information, the spherical shell material information, and the spherical shell defect information; a correlation coefficient matrix determination module for determining a correlation coefficient matrix corresponding to the spherical shell defect information based on the target correlation function, the representative discrete points, and the spherical shell defect information; a feature value decomposition module for performing feature value decomposition on the correlation coefficient matrix to obtain a standard random field; a random field determination module for performing random field transformation on the standard random field according to a random field transformation formula to obtain a target random field corresponding to the spherical shell defect information; Including, A random field modeling device for quantizing initial defects in a spherical shell structure.

5. An electronic device comprising a processor and a memory, wherein the memory stores at least one instruction and at least one program, and the at least one instruction and the at least one program are loaded and executed by the processor to realize a random field modeling method for quantizing an initial defect in a spherical shell structure described in any one of claims 1 to 3.

6. A computer storage medium having at least one instruction and at least one program stored therein, the at least one instruction and the at least one program being loaded and executed by a processor to realize a random field modeling method for quantizing an initial defect in a spherical shell structure as described in any one of claims 1 to 3.

Images