Nuclear power plant spare part inspection method, server and nuclear power plant spare part inspection system

Abstract

The invention provides a nuclear power plant spare part inspection method, a server and a nuclear power plant spare part inspection system. The method comprises the following steps: acquiring physical characteristic data of a spare part to be inspected from sensor detection equipment; performing deviation calculation on the physical characteristic data to obtain a deviation evaluation result; and sending the deviation evaluation result to the terminal equipment, so that the terminal equipment outputs the deviation evaluation result. According to the invention, high efficiency and accuracy of spare part inspection are realized through data acquisition and deviation evaluation. The tedious problem caused by dependence on manual operation in the traditional spare part acceptance inspection process is effectively solved.

Description

Technical Field

[0001] The present application relates to the field of nuclear power plant spare parts inspection, and in particular to a nuclear power plant spare parts inspection method, a server and a nuclear power plant spare parts inspection system. Background Art

[0002] Measurement is a key step to ensure that spare parts meet the specified requirements and quality standards. Through measurement, the key dimensions, shapes, accuracy, performance, etc. of spare parts can be detected and evaluated. This helps to determine whether the spare parts match the original equipment or design specifications to ensure that they can be properly installed and the equipment can operate normally. Traditional measurement and acceptance of spare parts mostly rely on manual operation, which takes a long time to operate, has a heavy burden on personnel, and requires high skills. Traditional acceptance cannot measure some parts, different measurement personnel lead to different measurement results, and data statistics management is complicated. The work process detection efficiency is low, and manual operation and filling in data are prone to errors in the final result. Summary of the invention

[0003] The present application provides a method, a server and a nuclear power plant spare parts inspection system for nuclear power plant, in order to solve the problem of cumbersome manual operation in the traditional measurement spare parts acceptance method.

[0004] The technical solution adopted by the present application to solve the technical problem is: to provide a method for inspecting spare parts of a nuclear power plant, using a server, the method comprising: Step S1: obtaining physical property data of the spare parts to be inspected from the sensor detection equipment; Step S2: performing deviation calculation on the physical property data to obtain a deviation evaluation result; Step S3: Sending the deviation evaluation result to the terminal device so that the terminal device outputs the deviation evaluation result.

[0005] In one embodiment, the step S3 includes: receiving a query instruction from the terminal device, and analyzing and processing the physical characteristic data according to the query instruction to obtain spare part status information; The spare part status information is sent to the terminal device.

[0006] In one embodiment, the step S3 further includes: receiving a statistical instruction from the terminal device, and performing statistical analysis on the physical characteristic data according to the statistical instruction to obtain a statistical analysis result; Send the statistical analysis result to the terminal device.

[0007] In one embodiment, the step S3 further includes: receiving a record export instruction from the terminal device, and performing data sorting according to the physical property data and / or the deviation evaluation result to obtain intermediate data; Performing format conversion processing on the intermediate data to obtain a record file that complies with a preset export format; Sending the record file that complies with the preset export format to the terminal device.

[0008] In one embodiment, the step S1 includes: Preprocessing the physical property data; The preprocessed physical property data is processed using a preset machine learning algorithm to obtain algorithm-processed physical property data.

[0009] In one embodiment, the step S1 further includes: An algorithm evaluation is performed on the preset machine learning algorithm to determine the reliability of the preset machine learning algorithm.

[0010] In one embodiment, the preset machine learning algorithm includes a preset classification learning algorithm: the physical property data processed by the preset machine learning algorithm is processed to obtain the physical property data processed by the algorithm, including: Using the preset classification learning algorithm to classify the preprocessed physical property data, thereby identifying the category of the spare parts; The performing algorithm evaluation on the preset machine learning algorithm to determine the reliability of the preset machine learning algorithm includes: Get the actual spare parts category; Substituting the physical property data into the preset classification learning algorithm to obtain a classification calculation result; Calculate the evaluation index by using the actual spare parts category and the classification calculation result; The reliability of the preset classification learning algorithm is determined according to the evaluation index.

[0011] In one embodiment, the evaluation indicators include precision and recall.

[0012] The present application also provides a server, comprising a processor and a memory storing a computer program, wherein the processor implements the steps of the above-mentioned method for inspecting spare parts of a nuclear power plant when executing the computer program.

[0013] The present application also provides a nuclear power plant spare parts inspection system, comprising: Sensor detection equipment; Terminal equipment; The server described above.

[0014] The implementation of this application has the following beneficial effects: This application provides a method, a server and a system for inspecting spare parts in a nuclear power plant. The steps of the method include obtaining physical property data of spare parts to be inspected from a sensor detection device; performing deviation calculation on the physical property data to obtain a deviation evaluation result; and sending the deviation evaluation result to a terminal device so that the terminal device outputs the deviation evaluation result. The present invention achieves high efficiency and accuracy in spare parts inspection through data collection and deviation evaluation. It effectively solves the cumbersome problems caused by relying on manual operation in the traditional spare parts inspection process. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present application will be further described below with reference to the accompanying drawings and embodiments, in which: Figure 1 It is a flow chart of a method for inspecting spare parts of a nuclear power plant according to the present application; Figure 2 It is a histogram of the physical property data of the spare parts of this application. DETAILED DESCRIPTION

[0016] The present application is further described in detail below by specific embodiments in conjunction with the accompanying drawings. Wherein similar elements in different embodiments adopt associated similar element numbers. In the following embodiments, many detailed descriptions are intended to enable the present application to be better understood. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, and methods. In some cases, some operations related to the present application are not shown or described in the specification, in order to avoid the core part of the present application being overwhelmed by too much description, and for those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0017] like Figure 1 As shown, Figure 1 The present invention is a flowchart of a method for inspecting spare parts of a nuclear power plant.

[0018] The present application provides a method for inspecting spare parts of a nuclear power plant, using a server, and the method includes: Step S1: obtaining physical property data of the spare parts to be inspected from the sensor detection equipment; In this step, it should be noted that the physical property data of the spare parts to be inspected are obtained from the sensor detection equipment. The sensor detection equipment includes but is not limited to high-precision measuring instruments, non-destructive testing equipment, etc., which can detect the key dimensions, shapes, precision, performance, etc. of the spare parts. For example, the dimension data of the spare parts is obtained by a laser rangefinder, and the internal defects of the spare parts are detected by an ultrasonic flaw detector.

[0019] In the process of digital intelligent acceptance, sensor detection equipment is relied on to obtain the physical property data of the spare parts to be accepted. These devices cover a variety of types such as high-precision measuring instruments and non-destructive testing, which can comprehensively detect key indicators such as key dimensions, shapes, accuracy and performance of spare parts. For example, laser rangefinders can accurately measure the dimensions of spare parts, while ultrasonic flaw detectors can effectively detect potential defects inside spare parts. As the core component of this process, the performance of sensors is crucial to ensuring the accuracy and reliability of acceptance data. At present, widely used sensors include photoelectric sensors, pressure sensors, temperature sensors, etc., as well as new sensors such as optical fiber sensors, microwave sensors and image sensors.

[0020] Step S2: performing deviation calculation on the physical property data to obtain a deviation evaluation result; In this step, it should be noted that the deviation calculation is performed on the obtained physical property data to obtain the deviation evaluation result. The deviation calculation is performed through a preset algorithm and model, and the actual measurement data is compared with the design specification or standard value to calculate the deviation value. For example, for a spare part with a design size of 100mm, the actual measured size is 100.5mm, and the deviation value is 0.5mm. The deviation evaluation result includes information such as the size and direction of the deviation value and whether it exceeds the allowable range.

[0021] Step S3: Send the deviation evaluation result to the terminal device so that the terminal device outputs the deviation evaluation result.

[0022] In this step, it should be noted that the deviation assessment results are sent to the terminal device, which can be a computer, tablet computer, smart phone, etc. After receiving the deviation assessment results, the terminal device outputs the results in the form of a graphical interface or report so that the inspector can intuitively understand the quality status of the spare parts. For example, the terminal device can display detailed information such as the size deviation, shape deviation, and performance deviation of the spare parts, and mark whether they meet the acceptance criteria.

[0023] The nuclear power plant spare parts inspection method of the present application significantly improves the inspection efficiency and reduces the time and workload of manual operation by automatically acquiring and processing physical property data. At the same time, the method avoids errors that may occur during manual operation and data filling, thereby improving the accuracy of the inspection results. In addition, the method can obtain the physical property data of spare parts in real time, and perform deviation calculation and evaluation in a timely manner to ensure that the quality of spare parts is always under control. By centrally managing and storing inspection data on the server, the method also improves the convenience and reliability of data management, and facilitates data query, statistics and analysis. Ultimately, the method ensures that spare parts meet design specifications and quality standards, thereby improving the safety and reliability of nuclear power plant equipment.

[0024] In one embodiment, step S3 includes: receiving a query instruction from a terminal device, and analyzing and processing the physical characteristic data according to the query instruction to obtain spare part status information; Send spare part status information to terminal devices.

[0025] It should be noted that the query instruction may include detailed information query for a specific spare part, overall quality assessment of a batch of spare parts, etc. The server further analyzes and processes the stored physical property data according to the query instruction to generate spare part status information. For example, statistical information such as the average deviation value, maximum deviation value, and minimum deviation value of the spare part can be calculated, or the quality of the spare part can be graded and assessed. After that, the spare part status information is sent to the terminal device. After receiving the spare part status information, the terminal device outputs the result in the form of a graphical interface or report so that the inspector can fully understand the status of the spare part. For example, the terminal device can display a detailed quality report of the spare part, including the statistical analysis results of various physical property data, quality assessment grade and other information. In addition, the measurement results and their main statistical value information (such as average value, σ, 3σ, 6σ, Ca, Cp, Cpk, etc.) will be automatically recorded and archived, and the operator can select different screening conditions to extract historical records, thereby realizing comprehensive monitoring and management of spare part quality.

[0026] In one embodiment, after step S3, the method further includes: receiving a statistical instruction from a terminal device, and performing statistical analysis on the physical characteristic data according to the statistical instruction to obtain a statistical analysis result; Send statistical analysis results to terminal devices.

[0027] like Figure 2It should be noted that the statistical instructions may include detailed statistical analysis of specific spare parts, statistical evaluation of the overall quality of a batch of spare parts, etc. According to the statistical instructions, the server further performs statistical analysis and processing on the stored physical property data to generate statistical analysis results. For example, statistical information such as the average deviation value, maximum deviation value, minimum deviation value, σ (standard deviation), 3σ (three times the standard deviation), 6σ (six times the standard deviation), process capability index: Ca (process performance index), Cp (process capability index), Cpk (process capability index) of spare parts can be calculated, or the quality of spare parts can be graded and evaluated. After that, the statistical analysis results are sent to the terminal device. After receiving the statistical analysis results, the terminal device outputs the results in the form of a graphical interface or report so that the inspector can fully understand the quality status of the spare parts. For example, the terminal device can display a detailed quality report of the spare parts, including the statistical analysis results of various physical property data, quality evaluation grades and other information. In addition, statistical analysis plays an important role in controlling the production process and improving product quality. Through SPC (Statistical Process Control) analysis, using statistical methods for quality diagnosis and analysis, it is possible to monitor the changing trend of product quality, play a preventive role in the production process, reduce the waste caused by post-inspection, and thus achieve control of the production process and improve product quality. The histogram can reflect the fluctuation state and distribution of product quality, convey information about the process quality status, and be used to judge and predict product quality and non-conforming rate. The trend chart monitors the abnormalities of the production equipment and production process through the regular change trend of the measured value, such as the monotonic change and periodic change of the measured value.

[0028] In one embodiment, after step S3, the method further includes: receiving a record export instruction from a terminal device, and arranging data according to physical property data and / or deviation evaluation results to obtain intermediate data; Perform format conversion on the intermediate data to obtain a record file that complies with the preset export format; Send the record file that conforms to the preset export format to the terminal device.

[0029] In one embodiment, the record management function includes: measurement records can be imported and exported, users can select the report format, and support users to set report templates. Record files can be saved in pdf, csv and other formats, and print reports can be quickly output with one click.

[0030] In one embodiment, step S1 includes: Preprocessing of physical property data; The preprocessed physical property data is processed using a preset machine learning algorithm to obtain algorithm-processed physical property data.

[0031] In a specific embodiment, the acquired physical property data is preprocessed to remove interference factors such as noise and outliers to improve the quality and reliability of the data. Preprocessing methods include but are not limited to data cleaning, data smoothing, data normalization, etc. For example, outliers are removed by data cleaning, data fluctuations are reduced by data smoothing, and data are scaled to a specific range by data normalization. The preprocessed physical property data is processed using a preset machine learning algorithm to obtain algorithm-processed physical property data. Machine learning algorithms may include but are not limited to support vector machines (SVM), decision trees, random forests, neural networks, etc. For example, data is classified by a support vector machine algorithm, and data is regressed by a random forest algorithm to obtain more accurate physical property data.

[0032] Deviation calculation is performed on the physical property data processed by the algorithm to obtain the deviation evaluation result. Deviation calculation compares the actual measurement data with the design specifications or standard values ​​through the preset algorithm and model to calculate the deviation value. The deviation evaluation result includes information such as the size and direction of the deviation value and whether it exceeds the allowable range. For example, the spare part with a design size of 100mm and an actual measured size of 100.5mm has a deviation value of 0.5mm.

[0033] Send deviation assessment results. Send deviation assessment results to terminal devices, such as computers, tablets, smart phones, etc. After receiving the results, the terminal device outputs them through a graphical interface or report form so that the inspector can intuitively understand the quality status of the spare parts. For example, the terminal device can display detailed information such as the size deviation, shape deviation, performance deviation, etc. of the spare parts, and mark whether they meet the acceptance criteria.

[0034] In one embodiment, after step S1, the method further includes: Perform algorithm evaluation on the preset machine learning algorithm to determine the reliability of the preset machine learning algorithm.

[0035] It should be noted that the preset machine learning algorithm is evaluated to determine its reliability. The algorithm evaluation is performed through preset evaluation indicators and methods, for example, through cross-validation, precision evaluation, recall evaluation and other methods to evaluate the performance and accuracy of the algorithm when processing physical property data. The evaluation results will be used to determine whether the algorithm is suitable for subsequent physical property data processing.

[0036] In one embodiment, the preset machine learning algorithm includes a preset classification learning algorithm: the preprocessed physical property data is processed using the preset machine learning algorithm, and the physical property data processed by the algorithm includes: Using a preset classification learning algorithm to classify the preprocessed physical property data, thereby identifying the category of spare parts; Performing algorithm evaluation on a preset machine learning algorithm to determine the reliability of the preset machine learning algorithm includes: Get the actual spare parts category; Substitute the physical property data into the preset classification learning algorithm to obtain the classification calculation results; Calculate the evaluation index through the actual spare parts category and the classification calculation results; Determine the reliability of the preset classification learning algorithm based on the evaluation indicators.

[0037] In a specific embodiment, the preprocessed physical property data is processed using a preset machine learning algorithm to obtain the physical property data processed by the algorithm. Specifically, the preprocessed physical property data is first classified using a preset classification learning algorithm to identify the category of the spare parts. For example, an SVM (support vector machine algorithm) can be used to classify the data to determine the type of the spare parts.

[0038] In order to ensure the reliability of the preset machine learning algorithm, it is necessary to evaluate the algorithm. The evaluation process includes the following steps: first, obtain the actual spare parts category, which can be done by manual labeling or obtaining the standard data of spare parts from the manufacturer; second, substitute the preprocessed physical property data into the preset classification learning algorithm to obtain the classification calculation result; then, by comparing the actual spare parts category with the classification calculation result, calculate the evaluation indicators such as accuracy, recall rate, F1 score, etc.; finally, determine the reliability of the preset classification learning algorithm based on these evaluation indicators. If the evaluation indicators show that the performance and reliability of the algorithm meet the requirements, the algorithm can be considered suitable for the identification of spare parts categories.

[0039] In one embodiment, the evaluation indicators include precision and recall.

[0040] It should be noted that the evaluation indicators include one or more of accuracy, recall, F1 score, and precision.

[0041] In one embodiment, before the acceptance of the machine learning algorithm, a data set containing representative samples needs to be collected and prepared to fully reflect the application effect of the algorithm in the actual scenario. The data set should be of high quality and fully annotated to facilitate the evaluation of the algorithm performance. For different tasks, corresponding performance evaluation indicators are used, such as accuracy, precision, recall, F1-score (F1 score) for classification tasks, and mean square error (MSE) or mean absolute error (MAE) for regression tasks. Among them, F1-score is an indicator used to measure the performance of classification models, especially widely used in binary classification problems. Its Chinese meaning is "F1 score" or "F1 score". In order to reduce the deviation of the evaluation results, cross-validation can be used to divide the data set into multiple subsets, repeat training and testing, and finally take the average result to effectively evaluate the generalization ability of the algorithm. In addition, comparing the algorithm to be evaluated with the existing benchmark algorithm can more objectively evaluate its advantages and disadvantages. After completing the algorithm verification experiment, the results need to be analyzed and interpreted to clarify the advantages and disadvantages of the algorithm and the room for improvement. For example, manual imagers rely on the operator's proficiency to perform learning tasks by manually taking points; fully automatic imagers use software to identify the edge of the product and automatically take points. Each time, the measurement needs to be fixed, and the software completes the learning task; the intelligent automatic measurement and acceptance device uses a camera to take pictures, and the software captures pixels to identify the edge of the product and calculates the number of pixels. The product can be placed in any direction to complete the learning task. By comparing the measurement data collected by different machines, the same product can be measured multiple times at fixed positions and different positions to evaluate indicators such as accuracy, recall rate and F1-score. In addition, the multiple measurement data of different products at fixed positions and different positions collected by different machines can calculate the mean square error or mean absolute error to evaluate the stability of machine measurement. Among the collected measurement data, some size data can be selected for cross-validation to avoid excessive verification due to too much product data. The benchmark algorithm comparison is based on the true value of the product. After the sensor equipment obtains the spare parts data, it first uses the basic algorithm comparison to understand the deviation between the data and the true value, and then uses cross-validation to evaluate the measurement stability of the equipment. Finally, the basic algorithm comparison and cross-validation data are summarized to evaluate the reliability and stability of the product data.

[0042] The present application also provides a server, including a processor and a memory storing a computer program, and the processor implements the steps of the above-mentioned nuclear power plant spare parts inspection method when executing the computer program.

[0043] The present application also provides a nuclear power plant spare parts inspection system, comprising: Sensor detection equipment; Terminal equipment; The above server.

[0044] Implementing this application has the following advantages: 1. Spare parts acceptance efficiency: Intelligent digital acceptance devices can quickly obtain measurement results, usually only a few seconds or even shorter. Using a vernier caliper for measurement is relatively slow.

[0045] 2. Automation and intelligence of spare parts acceptance: Intelligent digital acceptance devices are usually automated and intelligent, and can automatically identify, measure and analyze the characteristics of objects, reducing the errors and tediousness of manual operations.

[0046] 3. Non-contact measurement: Using non-contact measurement technology will not cause damage or deformation to the object being measured.

[0047] 4. High precision and accuracy: The measurement precision and accuracy of the intelligent digital acceptance device are higher than those of traditional acceptance tools, and can provide more accurate measurement data.

[0048] 5. Data processing and recording: Equipped with data processing and recording functions, it can display the measurement results in real time and generate detailed measurement reports to facilitate data analysis and preservation.

[0049] 6. Multifunctional measurement: Intelligent digital acceptance devices can measure multiple parameters, such as size, shape, angle, position, etc., while the functions of traditional acceptance tools are relatively single.

[0050] 7. Adaptability and flexibility: The intelligent digital acceptance device can adapt to the measurement of objects of different shapes and sizes, and has wider applicability and flexibility.

[0051] The purpose of the present invention is to solve the above problems, improve the efficiency of spare parts acceptance, reduce the errors and tediousness of human operation, adopt non-contact measurement and acceptance technology that will not cause damage or deformation to the measured object, and can provide more accurate measurement data. Equipped with data processing and recording functions, the measurement results can be displayed in real time, and detailed measurement reports can be generated to facilitate data analysis and preservation. The present invention relies on advanced Internet / big data / cloud computer and other technologies to realize visual one-key acceptance, and can quickly and accurately identify objects and process data. Secondly, high-precision sensors and algorithms are used to ensure the accuracy of spare parts information, reduce human errors, speed up spare parts acceptance efficiency, improve spare parts acceptance quality, reduce labor costs, and improve the operation safety and economy of power plants.

Claims

1. A method for inspecting spare parts of a nuclear power plant, using a server, characterized in that: The method comprises: Step S1: obtaining physical property data of the spare parts to be inspected from the sensor detection equipment; Step S2: performing deviation calculation on the physical property data to obtain a deviation evaluation result; Step S3: Sending the deviation evaluation result to the terminal device so that the terminal device outputs the deviation evaluation result.

2. The method for inspecting spare parts of a nuclear power plant according to claim 1, characterized in that: The step S3 then includes: receiving a query instruction from the terminal device, and analyzing and processing the physical characteristic data according to the query instruction to obtain spare part status information; The spare part status information is sent to the terminal device.

3. The method for inspecting spare parts of a nuclear power plant according to claim 1, characterized in that: After step S3, the following steps are also included: receiving a statistical instruction from the terminal device, and performing statistical analysis on the physical characteristic data according to the statistical instruction to obtain a statistical analysis result; Send the statistical analysis result to the terminal device.

4. The method for inspecting spare parts of a nuclear power plant according to claim 1, characterized in that: After step S3, the following steps are also included: receiving a record export instruction from the terminal device, and performing data sorting according to the physical property data and / or the deviation evaluation result to obtain intermediate data; Performing format conversion processing on the intermediate data to obtain a record file that complies with a preset export format; Sending the record file that complies with the preset export format to the terminal device.

5. The method for inspecting spare parts of a nuclear power plant according to claim 1, characterized in that: The step S1 then includes: Preprocessing the physical property data; The preprocessed physical property data is processed using a preset machine learning algorithm to obtain algorithm-processed physical property data.

6. The method for inspecting spare parts of a nuclear power plant according to claim 5, characterized in that: After step S1, the following steps are also included: An algorithm evaluation is performed on the preset machine learning algorithm to determine the reliability of the preset machine learning algorithm.

7. The method for inspecting spare parts of a nuclear power plant according to claim 6, characterized in that: The preset machine learning algorithm includes a preset classification learning algorithm: the physical property data processed by the preset machine learning algorithm is processed to obtain the physical property data processed by the algorithm, including: Using the preset classification learning algorithm to classify the preprocessed physical property data, thereby identifying the category of the spare parts; The performing algorithm evaluation on the preset machine learning algorithm to determine the reliability of the preset machine learning algorithm includes: Get the actual spare parts category; Substituting the physical property data into the preset classification learning algorithm to obtain a classification calculation result; Calculate the evaluation index by using the actual spare parts category and the classification calculation result; The reliability of the preset classification learning algorithm is determined according to the evaluation index.

8. The method for inspecting spare parts of a nuclear power plant according to claim 7, characterized in that: The evaluation indicators include precision and recall.

9. A server comprising a processor and a memory storing a computer program, characterized in that: The processor implements the steps of the method for inspecting spare parts of a nuclear power plant as described in any one of claims 1 to 8 when executing the computer program.

10. A nuclear power plant spare parts inspection system, characterized in that: include: Sensor detection equipment; Terminal equipment; The server of claim 9.

Images