Ray digital imaging defect analysis system for welding seam of end plug of nuclear fuel rod

Abstract

The invention belongs to the technical field of nuclear power, and particularly relates to a nuclear rod end plug welding seam ray digital imaging defect analysis system. The invention provides a novel and innovative on-line automatic defect analysis system aiming at the defects of an existing nuclear fuel rod end plug welding seam ray digital imaging defect analysis system. By introducing an advanced image processing technology and a deep learning algorithm, automatic analysis and defect identification of the ray digital image are realized, manual intervention and errors are greatly reduced, and the analysis efficiency and accuracy are improved. By using an efficient algorithm and an optimized calculation process, the online automatic analysis system disclosed by the invention can process a large number of ray images in real time, so that the analysis time is greatly shortened, and the analysis speed is improved. All defect data are stored through the database module, later tracing and analysis are facilitated, and powerful support is provided for quality management and improvement.

Description

Technical Field

[0001] The present invention belongs to the field of nuclear power technology, and in particular relates to a nuclear fuel rod end plug weld ray digital imaging defect analysis system. Background Art

[0002] With the widespread application of nuclear energy, radiographic digital imaging detection of nuclear fuel rod end plug welds has become an important means to ensure the safe operation of nuclear facilities. However, the current defect analysis of radiographic digital imaging mainly relies on manual operation, which is not only inefficient, but also prone to misjudgment or missed detection due to human factors. Therefore, the development of a system that can automatically and accurately analyze radiographic digital imaging defects of nuclear fuel rod end plug welds is of great significance to improving the safety and operation efficiency of nuclear facilities.

[0003] At present, the evaluation of the internal defects of nuclear power fuel rod weld end plugs by radiographic detection is mainly done manually, which not only requires a large workload, but also has high detection costs and a long cycle. The result assessors need to obtain the corresponding test qualification certificates regularly, and the assessment level is greatly affected by the subjective factors of the personnel. With the continuous development of the domestic nuclear industry, it is an inevitable development trend to promote weld inspection from traditional manual defect assessment methods to computer-based automatic weld defect detection.

[0004] The development of defect recognition technology using machine vision abroad was relatively early. Okayama University in Japan used fuzzy reasoning to realize computer automatic analysis of weld X-ray films and computational recognition of welding defects, and its recognition effect reached the level of professional film reviewers; BAM in Berlin, Germany used image processing and neural network technology to realize automatic recognition of weld crack defects in X-ray films; Nafa et al. in Algeria used digital image processing technology to realize contour tracking and defect classification of weld boundaries; France, Ukraine and the United Kingdom have all made outstanding research in this field, and held a conference on "CT Application in Industry and Image Processing in X-ray Photography" in Berlin, and published relevant research papers. In the field of nuclear fuel, Philips Industrial X-ray Company in Germany produced the MU80F semi-automatic X-ray inspection system for OJSC MSZ in Russia, and it has been installed in the nuclear fuel rod production line to detect the size and number of weld defects on nuclear fuel rods.

[0005] The application of relevant technologies in China is later than that in foreign countries. Tsinghua University and Tianjin University have conducted research in this field for many years. They have used digital image processing technology and neural network technology to establish a weld X-ray real-time imaging automatic analysis system, which has been applied to actual weld detection and achieved good results. Guangdong Yuehai Company has also conducted some research in this regard and developed image enhancement real-time imaging auxiliary evaluation software for image-assisted evaluation of weld detection, which greatly improves the accuracy and speed of weld defect evaluation, and the evaluation results are included in the detection database management. In the field of nuclear fuel, China National Nuclear Corporation North and China National Nuclear Corporation Construction have a certain foundation and have conducted in-depth research on the digital imaging technology of fuel rod welds, but the automatic identification of defects based on image processing needs further development. Based on digital imaging, China National Nuclear Corporation Seventh Institute used image feature extraction technology to conduct research on a small amount of data and achieved certain results.

[0006] In the existing technology, although there are some online automatic analysis systems, these systems often cannot fully meet the needs of nuclear facilities for efficient, accurate and reliable defect analysis. The main problems include:

[0007] Inability to conduct real-time online analysis: Most systems require digital images to be downloaded locally for offline analysis, which not only wastes time but also increases the possibility of misoperation.

[0008] Low degree of automation: Some systems still require manual intervention for defect identification and classification, which greatly reduces work efficiency.

[0009] Low accuracy: Due to the complexity of digital images and the diversity of defect types, existing systems often cannot accurately identify and classify various defects.

[0010] Low degree of centralization: Most of them use automatic analysis systems at the edge, and the analysis results and data are stored on the edge. There is no remote diagnosis and lack of centralized evaluation and data analysis capabilities.

[0011] Lack of verification of detection accuracy: For internal defects in nuclear fuel rod welds, only the statistical detection accuracy of sample test results is used. This may have uncertainties in the actual production environment, and the detection effect of the production environment needs to be verified.

[0012] Therefore, there is an urgent need for a new online automatic analysis system for radiographic digital imaging defects of nuclear fuel rod end plug welds that can meet the needs of nuclear facilities. Summary of the invention

[0013] In order to overcome the problems existing in the related art, a nuclear fuel rod end plug weld radiographic digital imaging defect analysis system is provided.

[0014] According to one aspect of an embodiment of the present disclosure, a nuclear fuel rod end plug weld radiographic digital imaging defect analysis system is provided, the system comprising: a nuclear fuel rod weld online evaluation application system, a nuclear fuel rod weld data image processing and analysis center, and a nuclear fuel rod weld quality verification three-dimensional imaging system;

[0015] The nuclear fuel rod weld online evaluation application system is used to realize image acquisition and image management, and integrate the digital image analysis algorithm model of the fuel rod end plug weld, to perform online defect recognition and evaluation on the acquired image data, and finally to uniformly report the analyzed fuel rod data to the nuclear fuel rod weld data image processing and analysis center;

[0016] The nuclear fuel rod weld data image processing and analysis center is used for centralized monitoring management and dispatch control. It is deployed on the secondary nodes of each production line for intelligent analysis, realizes the management of all fuel rod intelligent analysis data assets, and can support subsequent data asset sharing and distribution, initially forming data resource and assetization;

[0017] The nuclear fuel rod weld quality verification three-dimensional imaging system is used to effectively verify the quality of the fuel rod weld defect intelligent analysis system without damaging the inspected fuel rods; the nuclear fuel rod weld quality verification three-dimensional imaging system can perform non-destructive detection and non-destructive quality evaluation of internal defects and structural abnormalities in the welding parts of the fuel rods.

[0018] In a possible implementation, the nuclear fuel rod weld online evaluation application system includes: a digital image acquisition module, a digital image intelligent analysis and evaluation module, a digital image management module, an intelligent analysis algorithm parameter management module, a user management module, a manual review module, and a defect data reporting module;

[0019] The digital image acquisition module is used to collect the digital image function of the radiographic detection system in real time according to the image source of each fuel rod production line, and input the image to the digital image intelligent analysis and evaluation component module for defect detection;

[0020] The digital image intelligent analysis and evaluation module is used to define and integrate the algorithm models of digital image processing and measurement, and digital image defect intelligent analysis and evaluation. It implements the fuel rod weld defect algorithm detection scheduling based on the image data input by the digital image acquisition module, and identifies and outputs the defect detection results.

[0021] Digital image management module, used to realize local storage, viewing and screening of fuel rod weld defect data;

[0022] Intelligent analysis algorithm parameter management module, which is used to manage the algorithm model parameters for intelligent analysis of fuel rod weld defects, synchronize the parameters issued by the nuclear fuel rod weld data image processing and analysis center, and support local custom parameter configuration and update maintenance;

[0023] User management module, used to implement functions such as user creation, editing, status change, and permission management. For example, system user roles can include operators and administrators;

[0024] The manual review module is used to manually review and modify the defect type for any objections to the images processed by the image analysis and evaluation module using the administrator role

[0025] The defect data reporting module is used to encapsulate the structured defect data of the intelligent analysis of fuel rod weld defects, the image data marked with specific defects, and the operating status, and report them to the nuclear fuel rod weld data image processing and analysis center in real time.

[0026] In a possible implementation, the nuclear fuel rod weld data image processing and analysis center includes: a weld defect data access module, an intelligent analysis data center module, a weld defect real-time monitoring module, an image data acquisition management module, a defect data management query module, a multi-dimensional data statistical analysis module, a defect sample acquisition management module, and a system background management module;

[0027] The weld defect data access module is used to access the defect detection data of the nuclear fuel rod weld online evaluation application system, realize the digital storage of the fuel rod weld defect data, and form a nuclear fuel rod weld database;

[0028] Intelligent analysis data center module, used for statistical analysis of fuel rod weld defect data, real-time defect prompts and other visual displays;

[0029] The real-time monitoring module for weld defects is used to realize real-time prompt of defects, and can display the statistical data of the day's inspection, as well as the real-time inspection status and defect information;

[0030] Image data acquisition management module, used to manage and query preview all image data collected by the nuclear fuel rod weld online evaluation application system;

[0031] Defect data management query module is used to query and review the defect data detected by the nuclear fuel rod weld online evaluation application system; and can export the defect data in batches;

[0032] The multi-dimensional data statistical analysis module is used to analyze data based on defect intelligence and use big data analysis technology to establish multi-dimensional statistical analysis of fuel rod weld inspection results. It uses visualization technologies such as graphs and tables to present nuclear fuel rod defect data, image data and sample data in a centralized, intuitive and organized manner, forming data analysis capabilities, and through various comparative analyses, data presentations and linkage queries;

[0033] Defect sample collection and management module, which is used to realize the classification definition and management of samples based on the image data collected and analyzed by the fuel rod weld online evaluation application system;

[0034] The system background management module is used to add system users and configure their permissions.

[0035] In a possible implementation, a nuclear fuel rod weld quality verification three-dimensional imaging system includes: a radiation source subsystem, a detector subsystem, a scanning device subsystem, a scanning control subsystem, a radiation safety protection subsystem, and a software subsystem;

[0036] The radiation source subsystem is used to use a micro-focus X-ray machine as an X-ray source to generate X-rays in a controlled manner; it includes an open X-ray tube, a high-voltage generator, a cooler, a vacuum pump, cables, an electric control box and other components, and is equipped with a filter accessory device;

[0037] The detector subsystem collects X-rays based on the area array detector to generate digital images, which are then transmitted to the software subsystem workstation for processing;

[0038] The scanning device subsystem adopts a mechanical scanning system to carry the radiation source and detector to complete the movement required for DR / CT scanning;

[0039] The scanning control subsystem adopts a motion control system and accepts instructions from the software subsystem to achieve precise motion control of the mechanical system, real-time position measurement, and synchronous triggering of the radiation source and detector. It is also responsible for system power distribution and safety interlocking.

[0040] Radiation safety protection subsystem, used to prevent radiation accidents, reduce the severity of radiation accidents, and ensure the safety of workers, the public, and the environment; the radiation safety protection subsystem includes safety interlocking equipment, safety warning equipment, and video surveillance equipment;

[0041] The software subsystem is the control center and human-machine interface of the entire system. It is interconnected with the X-ray source, detector, scanning control and other subsystems. The operator controls the entire system through the human-machine interface.

[0042] In a possible implementation, the following steps are used to operate the nuclear fuel rod weld online evaluation application system:

[0043] Step 11, login: the user enters the user name and password to log in to the system;

[0044] Step 12: Image import; the system supports automatic acquisition of images taken by the X-ray system, and also supports manual selection and import by the operator;

[0045] Step 13, image verification: pre-process and verify the collected image according to the image quality inspection model issued by the system. If the verification is passed, the next step can be entered. If it fails, the operator needs to manually operate the imaging system to re-collect the image and import it into the system;

[0046] Step 14, defect detection: Defect detection is performed according to the defect detection model built into the system. Images without defects are imported into the image management module in real time. For defective images, the operator needs to make a manual judgment. If the operator believes that the system judgment is correct, the image directly enters the image management module. If the operator believes that the system judgment is wrong, the image is sent to the administrator for manual review.

[0047] In a possible implementation, the following steps are used to operate the nuclear fuel rod weld data image processing and analysis center:

[0048] Step 21, login: the user enters the user name and password to log in to the system;

[0049] Step 22, data access; the system automatically receives the image data and defect data reported by the automatic analysis system; the image data without defects enters the image management module;

[0050] Step 23, defect monitoring; the system automatically prompts the received defect data, and can be comprehensively displayed in the data center module, and queried and retrieved in the data query module;

[0051] Step 24, defect review: review defect data information and defect images.

[0052] In a possible implementation, the operation steps of the nuclear fuel rod weld quality verification three-dimensional imaging system after powering on are as follows:

[0053] Step 31, loading the workpiece; manually placing the workpiece at a designated location;

[0054] Step 32, close the lead room door; manually close the lead room door;

[0055] Step 33, setting parameters: setting relevant scanning parameters through the software subsystem;

[0056] Step 34, DR scanning: the system performs DR scanning on the workpiece according to the set scanning parameters;

[0057] Step 35, image analysis: After the DR scan is completed, image analysis is automatically performed, and a CT scan is performed on the area of ​​interest. After the test is completed, a test report is generated.

[0058] The beneficial effects of the present disclosure are as follows: the present disclosure proposes a new and innovative online automatic defect analysis system to address the shortcomings of the existing nuclear fuel rod end plug weld radiographic digital imaging defect analysis system. The innovations of the system are mainly reflected in the following aspects:

[0059] High degree of automation: This disclosure realizes automatic analysis and defect identification of X-ray digital images by introducing advanced image processing technology and deep learning algorithms, which greatly reduces human intervention and errors and improves the efficiency and accuracy of analysis.

[0060] Fast analysis speed: By using efficient algorithms and optimized calculation processes, the online automatic analysis system disclosed in the present invention can process a large number of radiographic images in real time, greatly shortening the analysis time and improving the analysis speed.

[0061] High analysis accuracy: The use of deep learning and rule model algorithms can extract and classify images more accurately, thereby more accurately identifying and classifying defects, improving the accuracy and reliability of analysis.

[0062] Strong traceability: The present invention saves all defect data through the database module, which is convenient for later traceability and analysis, and provides strong support for quality management and improvement.

[0063] Strong user-friendliness: Through the user interface module, users can easily interact and display results. The operation is simple and easy to understand, which greatly improves the user experience.

[0064] Improve detection efficiency: Support automatic analysis systems connected to multiple production lines, realize centralized management of defects and remote expert diagnosis, greatly shorten evaluation time, improve comprehensive analysis and judgment capabilities of defects, and thus improve product quality.

[0065] Non-destructive verification with detection accuracy: The present invention utilizes industrial three-dimensional imaging technology with micron-level gap resolution to achieve effective verification of defect detection quality without damaging the inspected fuel rods, thereby ensuring the detection quality of the automatic analysis system. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a block diagram of an online automatic evaluation application system for a fuel rod weld according to an exemplary embodiment.

[0067] Figure 2 The present invention is an operation flow chart of an online automatic evaluation application system for a fuel rod weld according to an exemplary embodiment.

[0068] Figure 3 It is a block diagram of a nuclear fuel rod weld data image processing and analysis center according to an exemplary embodiment.

[0069] Figure 4 It is an operation flow chart of a nuclear fuel rod weld data image processing and analysis center according to an exemplary embodiment.

[0070] Figure 5 The present invention is a block diagram of a three-dimensional imaging system for verifying the quality of a nuclear fuel rod weld according to an exemplary embodiment.

[0071] Figure 6 The present invention is an operation flow chart of a three-dimensional imaging system for verifying the weld quality of a nuclear fuel rod according to an exemplary embodiment. DETAILED DESCRIPTION

[0072] The present disclosure is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0073] The disclosed nuclear fuel rod end plug weld radiographic digital imaging defect analysis system is based on the defect intelligent analysis algorithm, establishes online intelligent detection and monitoring for each fuel rod production line, undertakes the overall integration and overall control tasks of the entire system, and realizes the classification management of the radiographic digital images of each fuel rod weld through the system, and establishes an online radiographic digital image acquisition, analysis, evaluation and reporting pipeline. The disclosed system includes an online automatic evaluation application system for nuclear fuel rod welds, a nuclear fuel rod weld data image processing and analysis center, and a nuclear fuel rod weld quality verification three-dimensional imaging system.

[0074] Figure 1 is a block diagram of an online automatic evaluation application system for a fuel rod weld according to an exemplary embodiment. Figure 1 As shown in the figure, the online automatic evaluation application system for nuclear fuel rod welds is built based on the fuel rod weld radiographic detection system on the nuclear fuel rod production line. As the secondary deployed edge analysis node of the whole system, it realizes image acquisition and image management, and integrates the automatic analysis algorithm model of digital images of fuel rod end plug welds. The collected image data is used for online defect recognition and evaluation, and finally the analyzed fuel rod data is uniformly reported to the nuclear fuel rod weld data image processing and analysis center.

[0075] The online automatic evaluation application system for nuclear fuel rod welds includes: digital image acquisition module, digital image intelligent analysis and evaluation module, digital image management module, intelligent analysis algorithm parameter management module, user management module, manual review module, and defect data reporting module.

[0076] The digital image acquisition module is used to automatically acquire the digital image function of the X-ray detection system in real time according to the image source of each fuel rod production line, and input the image to the digital image intelligent analysis and evaluation component module for defect detection.

[0077] The digital image intelligent analysis and evaluation module is used to define and integrate the algorithm models of digital image processing and measurement, and digital image defect intelligent analysis and evaluation. It implements the fuel rod weld defect algorithm detection scheduling based on the image data input by the digital image acquisition module, and automatically identifies and outputs the defect detection results.

[0078] The digital image management module is used to realize the local storage, viewing and screening of fuel rod weld defect data.

[0079] The intelligent analysis algorithm parameter management module is used to manage the algorithm model parameters for intelligent analysis of fuel rod weld defects, synchronize the parameters issued by the nuclear fuel rod weld data image processing and analysis center, and support local custom parameter configuration and update maintenance.

[0080] The user management module is used to implement functions such as user creation, editing, status change, and permission management. For example, system user roles can include operators and administrators.

[0081] The manual review module is used to manually review and modify the defect type for any objections to the images processed by the image analysis and evaluation module using the administrator role

[0082] The defect data reporting module is used to encapsulate the structured defect data of the intelligent analysis of fuel rod weld defects, the image data marked with specific defects, and the operating status, and report them to the nuclear fuel rod weld data image processing and analysis center in real time.

[0083] Figure 2 is an operation flow chart of an online automatic evaluation application system for a fuel rod weld according to an exemplary embodiment. Figure 2 As shown in the figure, the user roles of the system can be operator and administrator. Operator input is performed using the following steps.

[0084] Step 11, log in. The user enters the user name and password to log in to the system.

[0085] Step 12: Image import. The system supports automatic acquisition of images taken by the X-ray system, and also supports manual selection and import by the operator.

[0086] Step 13: Image verification. According to the image quality inspection model built into the system, the collected images are preprocessed and verified. If they pass the verification, they can proceed to the next step. If they fail, the operator needs to manually operate the imaging system to re-collect the images and import them into the system.

[0087] Step 14, defect detection. Defect detection is performed according to the defect detection model built into the system. Images without defects are imported into the image management module in real time. For images with defects, the operator needs to make a manual judgment. If the operator believes that the system judgment is correct, the image directly enters the image management module. If the operator believes that the system judgment is wrong, the image is sent to the administrator for manual review.

[0088] The nuclear fuel rod weld data image processing and analysis center is designed for nuclear fuel rod weld defects. It is based on the nuclear fuel rod weld online automatic evaluation application system and adopts B / S architecture. The core is to establish a fuel rod weld defect database, defect intelligent prompts, defect sample classification management, and storage, collection, management and analysis of all basic data and business data. As the first-level deployment node of the entire system, it is responsible for centralized monitoring management and dispatching control of the second-level nodes deployed in each production line for intelligent analysis. It has high flexibility, versatility and stability, and realizes one-stop, full-link and refined management of all fuel rod intelligent analysis data assets, and can support subsequent data asset sharing and distribution, and initially form data resource and assetization.

[0089] Figure 3 is a block diagram of a nuclear fuel rod weld data image processing and analysis center according to an exemplary embodiment. Figure 3 As shown, the nuclear fuel rod weld data image processing and analysis center includes: weld defect data access module, intelligent analysis data center module, weld defect real-time monitoring module, image data acquisition management module, defect data management query module, multi-dimensional data statistical analysis module, defect sample acquisition management module, and system background management module.

[0090] The weld defect data access module is used to access the defect detection data of the nuclear fuel rod weld online automatic evaluation application system, realize the automatic storage of the fuel rod weld defect data, and form a nuclear fuel rod weld database.

[0091] For example, the weld defect data access module can provide the following functions.

[0092] An interface for receiving the fuel rod being inspected automatically receives and monitors the fuel rod image data inspected by the nuclear fuel rod weld online automatic evaluation application system, including the original image and all corresponding single-rod images.

[0093] The interface for receiving defect data automatically receives the defect data detected and uploaded by the online automated evaluation application system for nuclear fuel rod welds, and automatically stores it in the database according to the rules, including pictures of detected defects, as well as the type of each defect and the position coordinates of the defect in the picture.

[0094] The intelligent analysis data center module is used to realize statistical analysis, real-time defect prompts and other visual displays of fuel rod weld defect data.

[0095] For example, the intelligent analysis data center module can provide the following functions.

[0096] An overview of defect data for the day: based on the data access module, statistics are provided on the total number of rods tested, the number of rods with detected defects, the number of rods with various types of defects and the number of defects according to the classification of nuclear fuel rod defects, and the detected defect data can be viewed.

[0097] Annual defect data statistics, based on the data access module, count the total number of fuel rods inspected annually, the number of qualified products and the product qualification rate; monthly statistics are based on the number of fuel rods inspected each month, the number of qualified products and the number of defects.

[0098] Basic data collection statistics, based on the image management and sample management modules, count the number of images and samples collected historically, as well as the number of samples of each type of defect by defect type, and generate a production line qualification ranking.

[0099] The real-time monitoring module for weld defects is used to realize automatic real-time prompting of defects. It can display the statistical data of the day's inspection, as well as real-time inspection status and defect information.

[0100] For example, the weld defect real-time monitoring module can achieve the following functions.

[0101] Statistics of overall inspection results are collected, including the total number of inspected fuel rods, the total number of qualified fuel rods, the total number of unqualified fuel rods, and the product qualification rate.

[0102] The real-time monitoring module for weld defects counts and displays the inspection status of each production line, including the number of inspection image groups, the total number of detected defects, the number of unqualified products, the product qualification rate, and the number of defects detected by defect type, and supports automatic update of real-time inspection results.

[0103] Statistics are collected on the total number of fuel rods tested annually, the number of qualified rods, and the qualified rate, as well as comparative analysis of the number of rods tested, the number of defective rods detected, and the number of qualified rods each month. The data is automatically updated when it changes.

[0104] After receiving defect data, the defect list is automatically refreshed, and you can further view the details, including structured data: production line, image name, number of groups, number of bars, number of defects, defects, acquisition time, and unstructured defect image data: original image, all single bar images. Batch export of defect data is supported.

[0105] The image data acquisition management module is used to manage, query and preview all image data collected by the nuclear fuel rod weld online automatic evaluation application system.

[0106] For example, the image data acquisition management module can provide the following functions.

[0107] The image data uploaded by the online automatic evaluation application system for nuclear fuel rod welds is queried and received according to rule conditions such as time period and image name.

[0108] Realize previewing image data, downloading image data, and exporting image data individually or in batches.

[0109] The defect data management query module is used to query and review the defect data detected by the nuclear fuel rod weld online automatic evaluation application system; and can export the defect data in batches.

[0110] For example, the defect data management query module can provide the following functions.

[0111] According to the query conditions such as time period, production line, fuel rod code, etc., the defect data can be classified and retrieved, and the specific situation of the defect can be further checked.

[0112] Achieve single or batch export of defect data, including structured and unstructured data;

[0113] View defect statistics for a single bar and view based on defect location.

[0114] The multi-dimensional data statistical analysis module is used to analyze data based on defect intelligence and use big data analysis technology to establish multi-dimensional statistical analysis of fuel rod weld inspection results. By comprehensively utilizing visualization technologies such as graphs and tables, the nuclear fuel rod defect data, image data and sample data are presented in a centralized, intuitive and organized manner, forming data analysis capabilities, and through various comparative analyses, data presentations and linkage queries.

[0115] For example, the multidimensional data statistical analysis module can provide the following functions.

[0116] Statistics and display of product qualification rate, comparative analysis of fuel rod defect detection in each production line, and comparative analysis of various types of defective products.

[0117] Trend analysis of each type of defect data detected each month and comparative analysis of the amount of each type of defect data;

[0118] Realize comparative analysis of monthly collected images and comparative analysis of historical sample data of each defect type collected by each component factory.

[0119] The defect sample collection and management module is used to realize the classification definition and management of samples based on the image data collected and analyzed by the online automatic evaluation application system of the fuel rod weld.

[0120] For example, the multidimensional data statistical analysis module can provide the following functions.

[0121] Query sample data according to each dimension and view annotated images.

[0122] Implement sample data management functions, import sample data in batches according to standard templates, automatically parse imported data set classification, sample name, and sample classification, store structured data in the database, and store unstructured image data in the specified storage directory;

[0123] It can realize downloading sample images, exporting compressed files of certain sample data, and exporting sample data in batches, including exporting only structured data or exporting both structured data and unstructured data.

[0124] Implement the invalidation or deletion function for sample data that is imported incorrectly or no longer used due to algorithm upgrades and optimizations.

[0125] The system background management module is used to add system users and configure their permissions.

[0126] Figure 4 is an operation flow chart of a nuclear fuel rod weld data image processing and analysis center according to an exemplary embodiment. Figure 4 As shown, the system is divided into two roles, monitor and administrator. After the monitor logs in, the steps are as follows:

[0127] Step 21, log in. The user enters the user name and password to log in to the system.

[0128] Step 22, data access. The system automatically receives the image data and defect data reported by the automatic analysis system; defect-free image data enters the image management module.

[0129] Step 23, defect monitoring. The system automatically prompts the received defect data, and can display it comprehensively in the data center module, and query and read it in the data query module.

[0130] Step 24: Check the defect data and defect pictures.

[0131] The nuclear fuel rod weld quality verification 3D imaging system is used to effectively verify the quality of the fuel rod weld defect intelligent analysis system without damaging the inspected fuel rod. The nuclear fuel rod weld quality verification 3D imaging system can perform non-destructive testing and non-destructive quality evaluation on the internal defects of the fuel rod welding parts (such as pores, incomplete penetration, bulging, tungsten inclusions, inclusions, cracks, looseness, porosity, etc.), structural anomalies, etc.

[0132] Figure 5 is a block diagram of a nuclear fuel rod weld quality verification three-dimensional imaging system according to an exemplary embodiment. Figure 5As shown, the nuclear fuel rod weld quality verification three-dimensional imaging system includes: a radiation source subsystem, a detector subsystem, a scanning device subsystem, a scanning control subsystem, a radiation safety protection subsystem, and a software subsystem.

[0133] The ray source subsystem is used to use a micro-focus X-ray machine as an X-ray source to generate X-rays in a controlled manner. It includes an open X-ray tube, a high-voltage generator, a cooler, a vacuum pump, cables, an electric control box, and other components, and is equipped with a filter accessory.

[0134] The detector subsystem, based on the area array detector, collects X-rays to generate digital images, which are then transmitted to the software subsystem workstation for processing. The functions of the detector subsystem can be implemented based on the area array detector, and the detector subsystem can be composed of components such as the detector panel, power supply, acquisition card, and cables. A structure in which a scintillating material is coupled to a two-dimensional photodiode array can be used. The scintillating material absorbs X-rays and converts them into light, which is then converted into charge by an amorphous silicon photodiode array. During the X-ray irradiation process, the charge accumulates continuously, and is then read out by a thin film transistor (TFT) integrated circuit, and the analog-to-digital conversion is completed to generate a digital image, which is then transmitted to the software subsystem workstation for processing.

[0135] The scanning device subsystem adopts a high-precision mechanical scanning system to carry the radiation source and detector to complete the movement required for DR / CT scanning. The subsystem includes a synchronous rotation motion device for the radiation source and detector, a translation motion device for the radiation source and detector (along the axial direction of the workpiece), and a translation motion device for the radiation source and detector (along the radial direction of the workpiece).

[0136] The scanning control subsystem adopts a high-precision motion control system and accepts instructions from the software subsystem to achieve precise motion control of the mechanical system, real-time position measurement, and synchronous triggering of the radiation source and detector. It is also responsible for system power distribution and safety interlocking.

[0137] The subsystem consists of a safety interlocking power distribution cabinet, a scanning device control cabinet, a motion control system, a system console, a trigger module, etc. The subsystem has functions such as real-time position measurement feedback, system status monitoring, fault self-diagnosis, and system initialization, and can also realize system power distribution, safety interlocking, and other functions.

[0138] The radiation safety protection subsystem is used to prevent radiation accidents, reduce the severity of radiation accidents, and ensure the safety of workers, the public, and the environment. The subsystem consists of safety interlocking equipment, safety warning equipment, video surveillance equipment, etc.

[0139] The software subsystem is the control center and human-machine interface of the entire detection system. It is interconnected with the X-ray source, detector, scanning control and other subsystems. The operator controls the entire system through the human-machine interface.

[0140] Figure 6 is an operation flow chart of a nuclear fuel rod weld quality verification three-dimensional imaging system according to an exemplary embodiment. Figure 6 As shown, the operation steps after the system is powered on are as follows:

[0141] Step 31, loading the workpiece: manually placing the workpiece at the designated location.

[0142] Step 32, close the lead room door. Close the lead room door manually.

[0143] Step 33, setting parameters: Setting relevant scanning parameters through the software subsystem.

[0144] Step 34, DR scanning: The system performs DR scanning on the workpiece according to the set scanning parameters.

[0145] Step 35, image analysis: After the DR scan is completed, image analysis is automatically performed, and a CT scan is performed on the area of ​​interest. After the test is completed, a test report is generated.

[0146] The advantages of the present invention compared with the prior art are:

[0147] High degree of automation: This disclosure realizes automatic analysis and defect identification of X-ray digital images by introducing advanced image processing technology and deep learning algorithms, which greatly reduces human intervention and errors and improves the efficiency and accuracy of analysis.

[0148] Fast analysis speed: By using efficient algorithms and optimized calculation processes, the online automatic analysis system disclosed in the present invention can process a large number of radiographic images in real time, greatly shortening the analysis time and improving the analysis speed.

[0149] High analysis accuracy: The use of deep learning and rule model algorithms can extract and classify images more accurately, thereby more accurately identifying and classifying defects, improving the accuracy and reliability of analysis.

[0150] Strong traceability: The present invention saves all defect data through the database module, which is convenient for later traceability and analysis, and provides strong support for quality management and improvement.

[0151] Strong user-friendliness: Through the user interface module, users can easily interact and display results. The operation is simple and easy to understand, which greatly improves the user experience.

[0152] Improve detection efficiency: Support automatic analysis systems connected to multiple production lines, realize centralized management of defects and remote expert diagnosis, greatly shorten evaluation time, improve comprehensive analysis and judgment capabilities of defects, and thus improve product quality.

[0153] Non-destructive verification with detection accuracy: The present invention utilizes industrial three-dimensional imaging technology with micron-level gap resolution to achieve effective verification of defect detection quality without damaging the inspected fuel rods, thereby ensuring the detection quality of the automatic analysis system.

[0154] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A nuclear fuel rod end plug weld radiographic digital imaging defect analysis system, characterized in that: The system includes: an online evaluation application system for nuclear fuel rod welds, a nuclear fuel rod weld data image processing and analysis center, and a nuclear fuel rod weld quality verification three-dimensional imaging system; The nuclear fuel rod weld online evaluation application system is used to realize image acquisition and image management, and integrate the digital image analysis algorithm model of the fuel rod end plug weld, to perform online defect recognition and evaluation on the acquired image data, and finally to uniformly report the analyzed fuel rod data to the nuclear fuel rod weld data image processing and analysis center; The nuclear fuel rod weld data image processing and analysis center is used for centralized monitoring management and dispatch control. It is deployed on the secondary nodes of each production line for intelligent analysis, realizes the management of all fuel rod intelligent analysis data assets, and can support subsequent data asset sharing and distribution, initially forming data resource and assetization; The nuclear fuel rod weld quality verification three-dimensional imaging system is used to effectively verify the quality of the fuel rod weld defect intelligent analysis system without damaging the inspected fuel rods; the nuclear fuel rod weld quality verification three-dimensional imaging system can perform non-destructive detection and non-destructive quality evaluation of internal defects and structural abnormalities in the welding parts of the fuel rods.

2. The system according to claim 1, characterized in that The nuclear fuel rod weld online evaluation application system includes: digital image acquisition module, digital image intelligent analysis and evaluation module, digital image management module, intelligent analysis algorithm parameter management module, user management module, manual review module, and defect data reporting module; The digital image acquisition module is used to collect the digital image function of the radiographic detection system in real time according to the image source of each fuel rod production line, and input the image to the digital image intelligent analysis and evaluation component module for defect detection; The digital image intelligent analysis and evaluation module is used to define and integrate the algorithm models of digital image processing and measurement, and digital image defect intelligent analysis and evaluation. It implements the fuel rod weld defect algorithm detection scheduling based on the image data input by the digital image acquisition module, and identifies and outputs the defect detection results. Digital image management module, used to realize local storage, viewing and screening of fuel rod weld defect data; Intelligent analysis algorithm parameter management module, which is used to manage the algorithm model parameters for intelligent analysis of fuel rod weld defects, synchronize the parameters issued by the nuclear fuel rod weld data image processing and analysis center, and support local custom parameter configuration and update maintenance; User management module, used to implement functions such as user creation, editing, status change, and permission management. For example, system user roles can include operators and administrators; The manual review module is used to manually review and modify the defect type for any objections to the images processed by the image analysis and evaluation module using the administrator role The defect data reporting module is used to encapsulate the structured defect data of the intelligent analysis of fuel rod weld defects, the image data marked with specific defects, and the operating status, and report them to the nuclear fuel rod weld data image processing and analysis center in real time.

3. The system according to claim 1, characterized in that The nuclear fuel rod weld data image processing and analysis center includes: weld defect data access module, intelligent analysis data center module, weld defect real-time monitoring module, image data acquisition management module, defect data management query module, multi-dimensional data statistical analysis module, defect sample acquisition management module, and system background management module; The weld defect data access module is used to access the defect detection data of the nuclear fuel rod weld online evaluation application system, realize the digital storage of the fuel rod weld defect data, and form a nuclear fuel rod weld database; Intelligent analysis data center module, used to realize statistical analysis of fuel rod weld defect data, real-time defect prompts and other visual displays; The real-time monitoring module for weld defects is used to realize real-time prompt of defects, and can display the statistical data of the day's inspection, as well as the real-time inspection status and defect information; Image data acquisition management module, used to manage and query preview all image data collected by the nuclear fuel rod weld online evaluation application system; Defect data management query module is used to query and review the defect data detected by the nuclear fuel rod weld online evaluation application system; and can export the defect data in batches; The multi-dimensional data statistical analysis module is used to analyze data based on defect intelligence and use big data analysis technology to establish multi-dimensional statistical analysis of fuel rod weld inspection results. It uses visualization technologies such as graphs and tables to present nuclear fuel rod defect data, image data and sample data in a centralized, intuitive and organized manner, forming data analysis capabilities, and through various comparative analyses, data presentations and linkage queries; Defect sample collection and management module, which is used to realize the classification definition and management of samples based on the image data collected and analyzed by the fuel rod weld online evaluation application system; The system background management module is used to add system users and configure their permissions.

4. The system according to claim 1, characterized in that The nuclear fuel rod weld quality verification 3D imaging system includes: radiation source subsystem, detector subsystem, scanning device subsystem, scanning control subsystem, radiation safety protection subsystem, and software subsystem; The radiation source subsystem is used to use a micro-focus X-ray machine as an X-ray source to generate X-rays in a controlled manner; it includes an open X-ray tube, a high-voltage generator, a cooler, a vacuum pump, cables, an electric control box and other components, and is equipped with a filter accessory device; The detector subsystem collects X-rays based on the area array detector to generate digital images, which are then transmitted to the software subsystem workstation for processing; The scanning device subsystem adopts a mechanical scanning system to carry the radiation source and detector to complete the movement required for DR / CT scanning; The scanning control subsystem adopts a motion control system and accepts instructions from the software subsystem to achieve precise motion control of the mechanical system, real-time position measurement, and synchronous triggering of the radiation source and detector. It is also responsible for system power distribution and safety interlocking. Radiation safety protection subsystem, used to prevent radiation accidents, reduce the severity of radiation accidents, and ensure the safety of workers, the public, and the environment; the radiation safety protection subsystem includes safety interlocking equipment, safety warning equipment, and video surveillance equipment; The software subsystem is the control center and human-machine interface of the entire system. It is interconnected with the X-ray source, detector, scanning control and other subsystems. The operator controls the entire system through the human-machine interface.

5. The system according to claim 1, characterized in that Use the following steps to operate the nuclear fuel rod weld online evaluation application system: Step 11, login: the user enters the user name and password to log in to the system; Step 12, image import; the system supports automatic acquisition of images taken by the X-ray system, and also supports manual selection and import by the operator; Step 13, image verification; According to the image quality inspection model issued by the system, the collected images are preprocessed and verified. If they pass the verification, they can proceed to the next step. If they fail, the operator needs to manually operate the imaging system to re-collect the images and import them into the system. Step 14, defect detection: Defect detection is performed according to the defect detection model built into the system. Images without defects are imported into the image management module in real time. For defective images, the operator needs to make a manual judgment. If the operator believes that the system judgment is correct, the image directly enters the image management module. If the operator believes that the system judgment is wrong, the image is sent to the administrator for manual review.

6. The system according to claim 1, characterized in that Use the following steps to operate the nuclear fuel rod weld data image processing and analysis center: Step 21, login: the user enters the user name and password to log in to the system; Step 22, data access; the system automatically receives the image data and defect data reported by the automatic analysis system; the defect-free image data enters the image management module; Step 23, defect monitoring; the system automatically prompts the received defect data, and can be comprehensively displayed in the data center module, and queried and retrieved in the data query module; Step 24, defect check; View defect data information and defect images.

7. The system according to claim 1, characterized in that The operation steps of the nuclear fuel rod weld quality verification 3D imaging system after powering on are as follows: Step 31, loading the workpiece; manually placing the workpiece at a designated location; Step 32, close the lead room door; manually close the lead room door; Step 33, setting parameters: setting relevant scanning parameters through the software subsystem; Step 34, DR scanning: the system performs DR scanning on the workpiece according to the set scanning parameters; Step 35, image analysis: After the DR scan is completed, image analysis is automatically performed, and a CT scan is performed on the area of ​​interest. After the test is completed, a test report is generated.

Images