A CFD software credibility quantitative evaluation method based on data mining

Abstract

The application provides a CFD software credibility quantitative evaluation method based on data mining, comprising: establishing a correlation data set between simulation data and test data of a CFD software credibility evaluation item to be evaluated; calculating absolute errors of each type of data sequence; calculating a maximum value and a minimum value of the absolute errors; calculating a correlation coefficient between each type of data sequence through the maximum value and the minimum value, and then calculating a correlation degree through the correlation coefficient; and finally calculating a correlation degree weight through the correlation degree; calculating a proportion mean of the absolute errors between each type of data sequence, and calculating a score of each type of data sequence through the proportion mean; calculating a credibility quantitative score through the correlation degree weight and the score of each type of evaluation data sequence, and obtaining a numerical simulation software credibility quantitative result. The application realizes quantitative evaluation of CFD software credibility, and provides data support for a decision of a simulation software selection of an aircraft designer.

Description

Technical Field

[0001] The present invention relates to the field of numerical simulation software credibility evaluation, and in particular to a CFD software credibility quantitative evaluation method based on data mining. Background Art

[0002] Computational fluid dynamics (CFD) is playing an increasingly important role in aircraft design, becoming the third most important design tool after wind tunnel experiments and theoretical analysis. It plays a crucial role in all stages of aircraft design. Consequently, the credibility of CFD software simulation results is attracting increasing attention. However, for any advanced numerical simulation technology to truly be effective in practical engineering, its accuracy, precision, and applicability must be evaluated to ensure the reliability and usability of the simulation results. This means evaluating the credibility of the calculation results.

[0003] Existing research on CFD software credibility indicates that the primary focus is on evaluating a specific capability of the software using a number of relevant case studies. For example, AIAA drag prediction utilizes three transport configurations: DLR-F4, DLR-F6, and CRM; and AIAA high-lift prediction utilizes Traping-Wing semi-extended / fully extended configurations. These case-based evaluation methods, such as grayscale comprehensive evaluation, realism-based assessment, and gray comprehensive evaluation, generally rely on expert experience to assign weighted scores. This incorporation of experts into the evaluation process is often influenced by human factors, such as their expertise, experience, and even emotions, leading to uncertainty in the evaluation results, which in turn impacts aircraft designers' decisions regarding simulation software. Summary of the Invention

[0004] In order to eliminate the uncertainty of evaluation results caused by human factors in existing CFD software credibility evaluation, an embodiment of the present application provides a CFD software credibility quantitative evaluation method based on data mining. The method aims to mine the potential correlation information within the CFD numerical simulation result data, realize the quantitative evaluation of CFD software credibility, and provide data support for aircraft designers to select simulation software.

[0005] The present application provides the following technical solution: a data mining-based quantitative evaluation method for CFD software credibility, comprising:

[0006] Step 1: Establish a correlation data set between the simulation data and the corresponding test data of the CFD software credibility evaluation items to be evaluated;

[0007] Step 2: Calculate the absolute error between the simulation data sequence and the corresponding test data sequence in the correlation data set to establish an absolute error data set;

[0008] Step 3: Calculate the maximum and minimum values ​​in the absolute error data set;

[0009] Step 4: Calculate the correlation coefficient between each type of data sequence based on the maximum value and the minimum value, then calculate the correlation degree based on the correlation coefficient, and finally calculate the correlation degree weight based on the correlation degree;

[0010] Step 5: Calculate the average of the proportions of absolute errors between each type of data sequence, and calculate the score of each type of data sequence based on the average of the proportions;

[0011] Step 6: Calculate the credibility quantification score through the correlation weight and score of each type of data sequence to obtain the credibility quantification result of the numerical simulation software.

[0012] According to one embodiment, the calculation formula of the correlation coefficient is:

[0013]

[0014] Where γ is the correlation coefficient, (i = 1, 2, ... n; j = 1, 2, ... m), i represents the number of subjects of numerical simulation / experiment used for the evaluation item, j represents the data sequence of the i-th group of numerical simulation / experiment sequence, ξ is the resolution coefficient, Δ max is the maximum value in the absolute error data set, Δ min is the minimum value in the absolute error data set.

[0015] According to one embodiment, the calculation formula of the correlation degree is:

[0016]

[0017] Where κ is the correlation degree, (i=1,2,…n; j=1,2,…m).

[0018] 4. The CFD software credibility quantitative evaluation method according to claim 3, wherein the calculation formula of the correlation weight is:

[0019]

[0020] Among them, ν i is the association weight of each type of data sequence, (i=1,2,...,n).

[0021] According to one embodiment, the formula for calculating the mean value of the specific gravity is:

[0022]

[0023] Among them, ωi It represents the average value of the absolute error proportion of each type of data sequence, i represents the number of subjects of numerical simulation / test used for the evaluation item, j represents the data sequence of the i-th group of numerical simulation / test sequence, and T represents the test data.

[0024] According to one embodiment, the score calculation formula is:

[0025] S i =1-ω i

[0026] Among them, S i Represents the score of each type of data sequence, which is between [0, 1].

[0027] According to one embodiment, the calculation formula of the credibility quantification score is:

[0028]

[0029] Among them, S is the score quantified by the credibility, ν i is the correlation weight of each type of data sequence, S i The score of each type of data series.

[0030] Compared with the prior art, the at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects:

[0031] 1. The present invention provides a quantitative evaluation method for CFD software credibility based on data mining. The method aims to start from the simulation result data itself, mine the potential correlation information within the data, realize the quantitative evaluation of CFD software credibility results, and eliminate the influence of human factors in existing evaluation technologies (such as correlation analysis and error comparison methods).

[0032] 2. This invention mines the potential information within CFD numerical simulation result data to identify the correlation between the data, thereby achieving a quantitative evaluation of the credibility of CFD software, thereby providing data support for aircraft designers' decision-making in selecting simulation software. This is not only suitable for CFD software, but can also be further extended to the credibility evaluation of other numerical simulation software in scientific computing. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0034] Figure 11 is a flow chart of a method for quantitatively evaluating the credibility of CFD software according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0036] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of this application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments, and the technical solutions of the present invention will be clearly and completely described. Obviously, the embodiments described 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 making creative work are within the scope of protection of the present invention.

[0037] An embodiment of the present invention provides a quantitative evaluation method for CFD software credibility based on data mining, which aims to eliminate the uncertainty of evaluation results caused by human factors in existing CFD software credibility evaluation, mine the potential information within the data of CFD numerical simulation results, establish the correlation between data, realize the quantitative evaluation of CFD software credibility, and provide data support for aircraft designers' decision-making in selecting simulation software.

[0038] like Figure 1 As shown, the CFD software credibility quantitative evaluation method based on data mining provided by the embodiment of the present invention specifically includes the following steps:

[0039] Step 1: Create a dataset to associate simulation data with test data

[0040] The numerical simulation results and experimental data of several evaluation cases used to evaluate the credibility of CFD software were collected, and a correlation data set between the two types of data was established, as shown in Table 1.

[0041] Table 1 List of associated datasets of simulation data and experimental data

[0042]

[0043] Step 2: Calculate the absolute error between the comparison data series

[0044] Calculate the absolute error between the CFD numerical simulation results and the corresponding experimental data.

[0045] The absolute error calculation formula is: Δ ij =|T ij -X ij|, (i = 1, 2, ... n; j = 1, 2, ... m), where i represents the case / test sequence, j represents the data sequence of the i-th case / test sequence, T represents the test data, and X represents the numerical simulation results. The comparison sequences must correspond one to one. A list of absolute error sets between the data sets is established, as shown in Table 2.

[0046] Table 2 List of absolute error sets between simulation data and test data

[0047]

[0048] Step 3: Calculate the maximum and minimum values ​​in the absolute error data set;

[0049] According to the maximum value calculation formula and the minimum value calculation formula, find the maximum error value and the minimum error value from the absolute error data set.

[0050] Maximum value calculation formula:

[0051] Minimum value calculation formula:

[0052] Step 4: Calculate the correlation coefficient between the simulation data sequence and the test data sequence through the maximum value and the minimum value, then calculate the correlation degree through the correlation coefficient, and finally calculate the correlation degree weight through the correlation degree.

[0053] (1) Calculate the correlation coefficient γ of each evaluation data sequence.

[0054] The calculation formula is: Wherein, (i=1, 2, ... n; j=1, 2, ... m), ξ is the resolution coefficient, preferably 0.5, to reduce the influence of the maximum absolute error being too large and causing distortion of the result.

[0055] (2) Calculate the correlation

[0056] Calculate the correlation degree κ of each type of evaluation data sequence.

[0057] The calculation formula is: (i=1,2,…n;j=1,2,…m) to obtain the correlation between various evaluation data.

[0058] (3) Calculate the association weight

[0059] Calculate the weight ν of each type of evaluation data sequence.

[0060] According to the formula: i=1,2,...,n calculate the weight ν of the correlation degree of the data sequence i .

[0061] Step 5: Calculate the average of the proportions of absolute errors between the simulation data sequence and the test data sequence, and calculate the score of each type of data sequence based on the average of the proportions;

[0062] (1) Calculate the average value of the absolute error proportion of each type of evaluation data series

[0063] According to the formula: (i=1,2,…n;j=1,2,…m) calculate the average value of the absolute deviation of each evaluation data, where ω i It represents the average error proportion of each type of evaluation data series.

[0064] (2) Calculate the scores of various evaluation data sequences

[0065] According to the formula: S i =1-ω i , the score of each type of data sequence is calculated according to the error deviation method, and the score is between [0, 1], where S i Indicates the score of each type of data sequence.

[0066] Step 6: Calculate the credibility quantification score through the correlation weight and score of each type of evaluation data sequence, and obtain the credibility quantification result of the numerical simulation software.

[0067] According to the formula: The credibility quantification score is calculated to obtain the final credibility quantification result of the numerical simulation software.

[0068] The method of this embodiment quantitatively evaluates the credibility of CFD software by mining potential correlation information within the data. The degree of correlation between the numerical simulation and experimental data sequences is determined based on the magnitude of the error between the two. The smaller the error, the closer the comparison data sequences are and the greater the correlation, while the smaller the error, the closer the correlation is. By analyzing the correlation data, a set of weights for software evaluation items (or evaluation indicators) is determined, thereby achieving a quantitative evaluation of the software's credibility.

[0069] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A CFD software credibility quantitative evaluation method based on data mining, characterized by: include: Step 1: Establish a correlation data set between the simulation data and the corresponding test data of the CFD software credibility evaluation items to be evaluated; Step 2: Calculate the absolute error between the simulation data sequence and the corresponding test data sequence in the correlation data set to establish an absolute error data set; Step 3: Calculate the maximum and minimum values ​​in the absolute error data set; Step 4: Calculate the correlation coefficient between each type of data sequence based on the maximum value and the minimum value, then calculate the correlation degree based on the correlation coefficient, and finally calculate the correlation degree weight based on the correlation degree; Step 5: Calculate the average of the proportions of absolute errors between each type of data sequence, and calculate the score of each type of data sequence based on the average of the proportions; Step 6: Calculate the credibility quantification score through the correlation weight and score of each type of data sequence, and obtain the credibility quantification result of the numerical simulation software; The mean value calculation formula of the proportion is: , in, Represents the average value of the absolute error proportion of each type of data sequence, represents the number of subjects of numerical simulation / experiment used in the evaluation item, j represents the Data series of numerical simulation / test series, represents the test data; The calculation formula of the score is: in, Represents the score of each type of data sequence, which is between [0, 1]; The calculation formula of the credibility quantification score is: ； Among them, S is the score of credibility quantification, is the correlation weight of each type of data sequence, The score of each type of data series.

2. The CFD software credibility quantitative evaluation method according to claim 1, characterized in that: The calculation formula of the correlation coefficient is: in, is the correlation coefficient, , represents the number of subjects of numerical simulation / experiment used in the evaluation item, j represents the Data series of numerical simulation / test series, is the resolution coefficient, is the maximum value in the absolute error data set, is the minimum value in the absolute error data set.

3. The CFD software credibility quantitative evaluation method according to claim 2, characterized in that: The calculation formula of the correlation degree is: , in, is the correlation, .

4. The CFD software credibility quantitative evaluation method according to claim 3, characterized in that: The calculation formula of the association weight is: in, is the correlation weight of each type of data sequence (i=1,2,...,n).

Images