Automatic trend analysis method and system for nuclear data critical reference test, storage medium and equipment

Abstract

The invention relates to a nuclear data critical reference test automatic trend analysis method and system, a storage medium and equipment, and the method comprises the steps: (1) selecting critical reference experiment feature information and the nuclear reaction section sensitivity of a critical reference experiment effective proliferation factor keff to each nuclide-reaction channel, and carrying out the screening of frequent item sets through association rule analysis; (2) establishing a linear regression model and analyzing by taking the nuclear reaction section sensitivity of the corresponding nuclide-reaction channel under each screened frequent item set condition as an independent variable and the C / E value of keff as a dependent variable, judging whether the C / E value of keff meets a trend analysis requirement or not, and screening out all effective data combinations meeting the requirement; and (3) drawing all the obtained effective data results into a chart and outputting the chart. According to the method, the critical keff automatic trend analysis based on nuclide-reaction channel sensitivity is realized, and the efficiency bottleneck of large-scale critical reference analysis is broken through.

Description

Technical Field

[0001] The present invention belongs to a macroscopic inspection method for nuclear data, and in particular relates to an automated trend analysis method, system, storage medium and equipment for critical benchmark inspection of nuclear data. Background Art

[0002] Nuclear data is the cornerstone of nuclear engineering applications. Nuclear data must undergo macroscopic verification to confirm their reliability before they can be used in nuclear engineering. Critical benchmark verification is one of the macroscopic verification methods for nuclear data conducted using critical experiments. Nuclear data critical benchmark verification uses critical benchmark experiments to simulate experiments using reliable calculation methods and calculate the effective proliferation factor k of critical experimental devices. eff , the calculated value (abbreviated as C value) is compared with the evaluated benchmark value (abbreviated as experimental value or E value) to check the deviation of the calculated value, thereby verifying the accuracy and reliability of a nuclear database in the specific type of application represented by the benchmark experiment. In order to effectively use the benchmark test results to improve nuclear data, trend analysis, sensitivity analysis and other methods have been developed at home and abroad to analyze k eff The possible causes of calculation deviations (referred to as C / E values) are analyzed to provide feedback for evaluating improvements to the nuclear database.

[0003] The so-called trend analysis of critical benchmark inspection is a method of studying the critical device k eff The method is to determine whether there is a linear relationship between the C / E value of the calculated value and a certain quantitative characteristic parameter of a series of benchmark experiments, so as to determine whether the calculation deviation is related to a certain quantitative experimental characteristic. The quantitative parameter here can be the enrichment of the fuel, the thickness of the critical device reflector material, the energy spectrum index that characterizes the hardness and softness of the neutron energy spectrum, k eff The purpose is to study a series of critical experiments k eff The C / E value changes with the experimental characteristic parameters, and the correlation between the calculated deviation and the materials in the critical device is established. The relationship between the calculated deviation and the nuclear data is further judged by combining physical analysis.

[0004] Sensitivity analysis is an important method in the macroscopic examination of nuclear data. Sensitivity analysis is used to evaluate the effect of changes in various parameters in the nuclear system (such as the microscopic cross-section, geometry, density, etc. of the material) on the integral quantity (such as k eff , reaction rate, reaction rate ratio, accumulation, neutron leakage spectrum, etc. Sensitivity analysis can directly establish the correlation between the integral and the parameter, and calculate the sensitivity coefficient by calculating the effect of a small change in the parameter on the integral. In the critical benchmark test, the calculation of k eff Sensitivity to nuclear data is a way to help researchers or engineers understand how the calculated bias is related to the nuclear data.

[0005] In order to carry out critical benchmark inspection more effectively, China Nuclear Data Center has established a trend analysis method based on nuclear data sensitivity. This method combines the advantages of sensitivity analysis and critical benchmark inspection trend analysis. On the basis of trend analysis, sensitivity is used as an analysis indicator, and linear regression is used to analyze k eff The correlation between the calculated deviation and different sensitivity data (such as nuclides, nuclides-reaction channel combinations, nuclides-reaction channel energy groups) is established to establish the relationship between nuclear data sensitivity and k eff The relationship model between the calculation deviation. This method can help researchers effectively analyze k eff Calculate the source of bias.

[0006] However, this method has encountered an efficiency bottleneck in practical application due to the increase in the scale of critical benchmark verification. At present, the China Nuclear Data Center has established the Evaluation Nuclear Data Systematization Macro-Verification System (ENDITS), which includes the critical benchmark experiment feature information database ENDITS-EXP containing 2237 critical benchmark experiments and the critical benchmark experiment sensitivity database ENDITS-SEN-2.1. eff The sensitivity of each nuclide-reaction channel in the device has been completed, and the critical benchmark test calculations of each major evaluation nuclear database have been completed, which has enabled the development of critical k based on the sensitivity of nuclide-reaction channel. eff The basis of trend analysis. However, in such a large-scale critical benchmark test trend analysis, it is difficult to rely solely on manpower to find all the feature correlations in thousands of critical benchmark experiments. The current sensitivity-based trend analysis requires manual screening of the benchmark experiment sequence according to the sensitive nuclides of the experiment, and then screening various experimental feature information (experimental materials, energy spectrum indicators, etc.) to find possible linear trends. This method is inefficient and difficult to traverse all possible trends. Therefore, it is necessary to implement critical benchmark test trend analysis based on nuclear data sensitivity through an efficient and automated algorithm. Related algorithms based on machine learning can largely meet the above requirements, provide more powerful data processing and understanding capabilities, and help break through the efficiency bottleneck of large-scale critical benchmark test trend analysis. Summary of the Invention

[0007] The purpose of the present invention is to address the defects in the prior art and provide a method, system, storage medium and device for automatic trend analysis of nuclear data criticality benchmark inspection, which can realize the critical k based on nuclide-reaction channel sensitivity through association rule analysis and linear regression analysis. eff Automated trend analysis breaks through efficiency bottlenecks in large-scale critical benchmark analysis.

[0008] To achieve the above object, the technical solution of the present invention is as follows:

[0009] A method for automated trend analysis of nuclear data criticality benchmark verification comprises the following steps:

[0010] (1) Selecting critical benchmark experiment characteristic information and critical benchmark experiment effective proliferation factor k eff For the nuclear reaction cross section sensitivity of each nuclide-reaction channel, frequent item sets are screened through association rule analysis to select benchmark experimental combinations with certain correlation and meeting the minimum support threshold requirement;

[0011] (2) Taking the nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened as the independent variable, k eff The C / E value (calculation deviation) is the dependent variable, a linear regression model is established and analyzed to determine k eff Whether the C / E value meets the trend analysis requirements, and screen out all valid data combinations that meet the requirements;

[0012] (3) All valid data results are plotted into charts for output.

[0013] Furthermore, in a specific embodiment, in the above-mentioned nuclear data critical benchmark test automated trend analysis method, in step (1), based on the critical benchmark experiment feature information database ENDITS-EXP and the critical benchmark experiment sensitivity database ENDITS-SEN-2.1 established by the China Nuclear Data Center, the critical benchmark experiment feature information and critical benchmark experiment effective proliferation factor k related to all material types are selected. eff The nuclear reaction cross section sensitivity of each nuclide-reaction channel is analyzed and processed using association rule mining tools.

[0014] Furthermore, in step (1), frequent itemsets are screened using the April algorithm.

[0015] Furthermore, in a specific embodiment, in the above-mentioned method for automated trend analysis of critical benchmark inspection of nuclear data, in step (2), the linear regression determination factor R is calculated by performing regression analysis on the frequent item sets selected in step (1). 2 , according to R 2 The size of determines whether it meets the trend analysis requirements.

[0016] Furthermore, in step (2), when the linear regression factor R 2 When it is greater than 0.5, the analyzed data is considered to have a certain linear trend, and the corresponding data combination is saved.

[0017] Furthermore, in a specific embodiment, in the above-mentioned automated trend analysis method for critical benchmark inspection of nuclear data, in step (3), data visualization is performed to draw a trend analysis image.

[0018] Furthermore, in step (3), the trend analysis graph drawn uses the nuclear reaction cross section sensitivity data of the nuclide-reaction channel as the x-axis, k eff The C / E value is on the y-axis, and the content includes nuclide-reaction channel, data points, fitting line, regression equation, regression coefficient, and frequent item set.

[0019] Based on the above-mentioned method for automated trend analysis of nuclear data criticality benchmark verification, the present invention further provides a software system for implementing the method, comprising:

[0020] Association rule analysis module is used to combine the selected critical benchmark experiment feature information and the critical benchmark experiment effective proliferation factor k eff For the nuclear reaction cross section sensitivity of each nuclide-reaction channel, frequent item sets are screened through association rule analysis to select benchmark experimental combinations with certain correlation and meeting the minimum support threshold requirement;

[0021] The linear regression analysis module is used to take the nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened as the independent variable, k eff The C / E value (calculation deviation) is the dependent variable, a linear regression model is established and analyzed to determine k eff Whether the C / E value meets the trend analysis requirements, and screen out all valid data combinations that meet the requirements;

[0022] The visualization output module is used to draw all valid data results into charts for output.

[0023] Further, in a specific embodiment, in the nuclear data critical benchmark inspection automated trend analysis system as described above, the association rule analysis module is implemented by the mlxtend library in Python software, the linear regression analysis module is implemented by the sciki t-learn library in Python software, and the visualization output module is implemented by the matplot lib library in Python software.

[0024] The present invention also provides a storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for automated trend analysis of nuclear data criticality benchmark inspection.

[0025] The present invention further provides a device comprising a storage medium and a processor, wherein the storage medium stores a computer program, and the processor implements the above-mentioned method for automated trend analysis of nuclear data criticality benchmark inspection when executing the computer program.

[0026] The present invention has the following beneficial effects: The automated trend analysis method for critical benchmark testing based on nuclide-reaction channel sensitivity provides powerful data screening capabilities, rapidly identifying correlated item sets (frequent item sets) from thousands of benchmark experiments and outputting those frequent item sets that satisfy linear regression criteria as visualization images for trend analysis. The analysis time required for a single nuclide-reaction channel sensitivity combination is approximately 10 seconds, several orders of magnitude faster than manual analysis. This significantly improves the efficiency of critical benchmark testing trend analysis and overcomes the efficiency bottleneck of large-scale critical benchmark testing analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is the critical k based on the nuclide-reaction channel sensitivity in the specific embodiment of the present invention eff Automated trend analysis flow chart. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0029] The overall technical idea of the present invention is that the user first selects the characteristic information of all critical benchmark experiments and the critical benchmark experiment k eff For the sensitivity of a group of nuclear reaction cross sections (nuclide-reaction channel), the benchmark experiments are automatically grouped by association rule analysis instead of expert manual analysis; then, the k eff Sensitivity to nuclear reaction cross section and k eff Calculate the deviation (C / E value) and perform linear regression analysis. When the linear regression factor R 2 When the judgment conditions are met, the obtained data are considered to have a certain linear trend, and the experimental grouping is saved and plotted.

[0030] Specifically, the present invention provides a method for automated trend analysis of nuclear data criticality benchmark inspection. Figure 1 As shown, the following steps are included:

[0031] (1) Screening of frequent item sets through association rule analysis

[0032] First, check whether the information of the input file is legal; then, select the critical benchmark experiment feature information and the critical benchmark experiment effective proliferation factor k according to the read file data. effThe sensitivity of the nuclear reaction cross section of each nuclide-reaction channel is determined; a minimum support threshold is set (for example, the minimum support threshold is set so that each frequent item set can screen out no less than 3 experiments, which is convenient for subsequent regression analysis), and then the frequent item sets are screened through association rule analysis to select benchmark experiment combinations with certain correlations that meet the minimum support threshold requirements.

[0033] (2) Establish a linear regression model and perform trend analysis

[0034] The nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened in step (1) is used as the independent variable, k eff The C / E value (calculated deviation) is the dependent variable, a linear regression model is established and analyzed, the minimum linear regression factor is set, and k is judged. eff Whether the C / E value meets the trend analysis requirements, and filter out all valid data combinations that meet the requirements.

[0035] (3) Visual output of results

[0036] All valid data results obtained in step (2) are plotted into a graph for output. The trend analysis graph is plotted with the nuclear reaction cross section sensitivity data of the nuclide-reaction channel as the x-axis, k eff The C / E value is on the y-axis, and the content includes nuclide-reaction channel, data points, fitting line, regression equation, regression coefficient, and frequent item set.

[0037] Furthermore, the present invention also provides a nuclear data criticality benchmark verification automated trend analysis system, storage medium, and device. The software system corresponding to the above-mentioned nuclear data criticality benchmark verification automated trend analysis method includes:

[0038] Association rule analysis module is used to combine the selected critical benchmark experiment feature information and the critical benchmark experiment effective proliferation factor k eff For the nuclear reaction cross section sensitivity of each nuclide-reaction channel, frequent item sets are screened through association rule analysis to select benchmark experimental combinations with certain correlation and meeting the minimum support threshold requirement;

[0039] The linear regression analysis module is used to take the nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened as the independent variable, k eff The C / E value (calculation deviation) is the dependent variable, a linear regression model is established and analyzed to determine k eff Whether the C / E value meets the trend analysis requirements, and screen out all valid data combinations that meet the requirements;

[0040] The visualization output module is used to draw all valid data results into charts for output.

[0041] In a specific implementation, the association rule analysis module can be implemented by the mlxtend library in Python software, the linear regression analysis module can be implemented by the sciki t-learn library in Python software, and the visualization output module can be implemented by the matplotlib library in Python software.

[0042] The computer program of the software system is stored in a storage medium, and when the computer program is executed by a processor, the automated trend analysis method for criticality benchmark inspection of nuclear data as described above is implemented.

[0043] A hardware device comprises the above-mentioned storage medium and a processor, wherein the storage medium stores a computer program, and when the processor executes the computer program, the above-mentioned nuclear data criticality benchmark verification automated trend analysis method is implemented.

[0044] Example

[0045] This embodiment describes an automated trend analysis software for critical benchmark inspection of nuclear data based on machine learning algorithms. The software architecture includes an association rule analysis module, a linear regression analysis module, and a visualization output module. The association rule analysis module selects the characteristic information of critical benchmark experiments related to all material types and the critical benchmark experiment k eff For the sensitivity of a group of nuclear reaction cross sections (nuclide-reaction channel), the critical benchmark experiments are automatically grouped by association rule analysis; the linear regression analysis module is used to calculate the k of each group. eff Sensitivity to nuclear reaction cross section and k eff Calculate the deviation (C / E value) and perform linear regression analysis to screen out experimental group data with a certain linear trend; the visualization output module plots all the valid data results obtained into charts for output.

[0046] The process of using this software to perform sensitivity-based trend analysis is as follows:

[0047] Step 1: Screen frequent itemsets based on the characteristics of critical benchmark experiments.

[0048] The China Nuclear Data Center has established the Evaluation Nuclear Data System Macro-Verification System (ENDITS), which includes the critical benchmark experiment feature information database ENDITS-EXP containing 2237 critical benchmark experiments, the critical benchmark experiment sensitivity database ENDITS-SEN-2.1, and provides the k eff Sensitivity to each nuclide-reaction channel within the device.

[0049] This embodiment is based on the critical benchmark experiment feature information database ENDITS-EXP and the critical benchmark experiment sensitivity database ENDITS-SEN-2.1 established by the China Nuclear Data Center. The characteristic parameters related to the material types in ENDITS-EXP are selected as items, all benchmark experiments in ENDITS-EXP are classified, and benchmark experiment combinations (frequent item sets) with certain correlations and meeting the minimum support threshold requirements are selected.

[0050] Specifically, after completing the inspection and reading of the input file, this embodiment uses the April algorithm of the mlxtend library in Python software to mine and process the data, screen out all frequent item sets that exceed the minimum support threshold, check the legitimacy of the frequent item sets, calculate the association rules and store the results, and output all frequent item set combinations that meet the minimum support threshold conditions and the corresponding association rule files.

[0051] Association rule calculation involves three aspects: item set support, confidence, and lift. Support indicates how frequently an item set (e.g., {A, B}) appears in the entire dataset. High support indicates that the item set {A, B} appears frequently and is universal. To filter frequent item sets, a minimum support threshold is typically set to filter out low-frequency itemsets. In this example, the minimum support threshold is set so that each frequent item set can be filtered out for at least three experiments.

[0052]

[0053] Confidence measurement refers to the conditional probability that itemset B is also included in a transaction containing itemset A. A high confidence level indicates that when A occurs, the probability of B occurring is high.

[0054]

[0055] Lift measures the independence between item sets A and B. A lift greater than 1 indicates a positive correlation between A and B; a lift of 1 indicates independence; and a lift less than 1 indicates a negative correlation. This can reflect the effectiveness of a rule and avoid the potential misleading effects of relying solely on confidence.

[0056]

[0057] This embodiment only uses support to screen frequent itemsets, and confidence and lift are mainly used to assist subsequent expert analysis.

[0058] mlxtend (Machine Learning Extensions) is a Python library for machine learning and data science. It provides many practical tools and functions for tasks such as data preprocessing, model evaluation, feature selection, and visualization. mlxtend.frequent_patterns provides tools for association rule mining, such as the Apriori algorithm and the FP-Growth algorithm. This example uses the mlxtend library to calculate association rules and perform data mining.

[0059] The minimum support threshold (Min Support Threshold) refers to the minimum support value used to determine which itemsets are considered frequent itemsets in data mining. The purpose of setting the minimum support threshold is to filter out itemsets that appear frequently in the dataset, thereby reducing the amount of computation and complexity, and helping to find meaningful association rules more quickly. Taking the Apriori algorithm for frequent itemset screening as an example, first, a minimum support threshold is set to determine whether an itemset is a frequent itemset. Support refers to the frequency with which an itemset appears in the entire dataset. Then, the dataset is scanned, the support of each candidate itemset is calculated, and itemsets with support greater than or equal to the minimum support threshold are filtered out.

[0060] Step 2: Establish a linear regression model and perform trend analysis.

[0061] This step is to determine k by establishing a linear regression model. eff The calculation deviation satisfies a certain trend. If it does, it is considered a valid output result. Specifically, after filtering out all frequent item sets (if any), this embodiment uses the scikit-learn library in Python software to perform regression analysis.

[0062] Scikit-learn (sklearn) is a widely used machine learning library in Python. It provides simple and efficient tools for data mining and analysis, supporting a variety of machine learning tasks, including classification, regression, clustering, dimensionality reduction, model selection, and preprocessing. Scikit-learn is built on NumPy, SciPy, and Matplotlib, with a consistent API design that makes it easy to use and powerful.

[0063] When using the sciki t-learn library for regression analysis, the imported data set must contain features (independent variables) and target variables (dependent variables). This example uses the nuclear reaction cross section sensitivity data of the corresponding nuclide-reaction channel under each frequent item set as the independent variable, k effThe C / E value (calculated deviation) is used as the dependent variable, the data set is divided into a training set and a test set, and a simple linear regression is performed using the LinearRegression model in sciki t-learn to fit the established linear regression model; the trained model is used to predict the test set, and the mean square error (MSE) and linear regression determination coefficient (R) are used to calculate the C / E value (calculated deviation). 2 Evaluate model performance.

[0064] In this example, the linear regression coefficient of determination R 2 , to screen data combinations with linear trends. R 2 The value range is [0,1], which indicates the model's ability to explain the target variable. 2 =1 means the model can perfectly fit the data, R 2 =0 means that the model cannot explain the changes of the target variable. 2 All data combinations with a value greater than 0.5 are considered to have a certain linear trend and can be used as valid output results.

[0065] Step 3: Visual output of the results.

[0066] In this embodiment, all the valid results obtained in step 2 are plotted into graphs for output. The specific method is to use the matplotlib library in Python software to visualize the data and draw trend analysis graphs (sensitivity data is on the x-axis, k is on the k-axis). eff The C / E value is on the y-axis), and the content includes nuclide-reaction channel, data points, fitting line, regression equation, regression coefficient, frequent item set, and all images are output as PDF files.

[0067] Matplotlib is one of the most commonly used data visualization libraries in Python software. It provides rich drawing functions, supports multiple chart types and is highly customizable. The basic operations for generating charts using matplotlib include: importing libraries, preparing data, creating graphs and subgraphs, drawing charts, customizing charts, displaying and saving. Using matplotlib, you can easily draw trend analysis images. Trend analysis is usually used to show how data changes with ordered variables. Common trend analysis charts include line charts and trend lines. This embodiment can use trend analysis images drawn by the matplotlib library to display k eff The C / E value changes with the sensitivity of the nuclear reaction cross section of the corresponding nuclide-reaction channel to reflect the linear relationship between the data.

[0068] Those skilled in the art will readily identify other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. Thus, if such variations, uses, or adaptations of the present invention fall within the scope of the claims and their equivalents, the present invention is intended to encompass such variations and uses.

[0069] The above embodiments are merely illustrative of the present invention, and the present invention may also be implemented in other specific ways or in other specific forms without departing from the gist or essential features of the present invention. Therefore, the described embodiments should be considered in all respects as illustrative rather than restrictive. The scope of protection of the present invention should be described by the claims, and any variations that are equivalent to the intent and scope of the claims should also be included within the scope of the present invention.

Claims

1. A method for automated trend analysis of nuclear data criticality benchmark verification, characterized in that: The steps include: (1) Selecting critical benchmark experiment characteristic information and critical benchmark experiment effective proliferation factor k eff For the nuclear reaction cross section sensitivity of each nuclide-reaction channel, frequent item sets are screened through association rule analysis to select benchmark experimental combinations with certain correlation and meeting the minimum support threshold requirement; (2) Taking the nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened as the independent variable, k eff The C / E value is the dependent variable, a linear regression model is established and analyzed to determine k eff Whether the C / E value meets the trend analysis requirements, and screen out all valid data combinations that meet the requirements; (3) All valid data results obtained are plotted into charts for output.

2. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 1, wherein: In step (1), based on the critical benchmark experiment characteristic information database ENDITS-EXP and the critical benchmark experiment sensitivity database ENDITS-SEN-2.1 established by the China Nuclear Data Center, the critical benchmark experiment characteristic information and critical benchmark experiment effective proliferation factor k related to all material types are selected. eff The nuclear reaction cross section sensitivity of each nuclide-reaction channel is analyzed and processed using association rule mining tools.

3. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 2, wherein: In step (1), frequent itemsets are screened using the Apriori algorithm.

4. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 1, wherein: In step (2), the linear regression determination factor R is calculated by performing regression analysis on the frequent item sets selected in step (1). 2 , according to R 2 The size of is used to determine whether the trend analysis requirements are met.

5. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 4, wherein: In step (2), when the linear regression factor R 2 When it is greater than 0.5, the analyzed data is considered to have a certain linear trend, and the corresponding data combination is saved.

6. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 1, wherein: In step (3), the data is visualized and a trend analysis graph is drawn.

7. The method for automated trend analysis of nuclear data criticality benchmark verification according to claim 6, wherein: In step (3), the trend analysis graph is drawn with the nuclear reaction cross section sensitivity data of the nuclide-reaction channel as the x-axis, k eff The C / E value is on the y-axis, and the content includes nuclide-reaction channel, data points, fitting line, regression equation, regression coefficient, and frequent item set.

8. A nuclear data criticality benchmark verification automated trend analysis system for implementing the method of claim 1, characterized in that: include: Association rule analysis module is used to combine the selected critical benchmark experiment feature information and the critical benchmark experiment effective proliferation factor k eff For the nuclear reaction cross section sensitivity of each nuclide-reaction channel, frequent item sets are screened through association rule analysis to select benchmark experimental combinations with certain correlation and meeting the minimum support threshold requirement; The linear regression analysis module is used to take the nuclear reaction cross section sensitivity of the corresponding nuclide-reaction channel under the conditions of each frequent item set screened as the independent variable, k eff The C / E value is the dependent variable, a linear regression model is established and analyzed to determine k eff Whether the C / E value meets the trend analysis requirements, and screen out all valid data combinations that meet the requirements; The visualization output module is used to draw all valid data results into charts for output.

9. The nuclear data criticality benchmark verification automated trend analysis system according to claim 8, characterized in that: The association rule analysis module is implemented by the mlxtend library in Python software, the linear regression analysis module is implemented by the scikit-learn library in Python software, and the visualization output module is implemented by the matplotlib library in Python software.

10. A storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for automated trend analysis of nuclear data criticality benchmark inspection according to any one of claims 1 to 7 is implemented.

11. A device comprising a storage medium and a processor, wherein the storage medium stores a computer program, wherein: When the processor executes the computer program, the method for automated trend analysis of nuclear data criticality benchmark inspection as described in any one of claims 1 to 7 is implemented.

Images