Management method and system for effectively judging quality assurance grade of nuclear power plant spare part
Through the combined image recognition and neural network model with dynamic evaluation mechanism, the problem of insufficient accuracy of the evaluation of spare parts warranty level in traditional nuclear power plants and the reliance on manual inspection is solved, and the intelligent and dynamic management of spare parts warranty level is realized, and the evaluation accuracy and efficiency are improved.
Patent Information
- Application Number
- CN202411569607.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-05
- Publication Date
- 2025-02-25
AI Technical Summary
The assessment of the warranty level of spare parts in traditional nuclear power plants relies on static data and cannot be adjusted in real time, resulting in insufficient evaluation accuracy and relying on manual testing, which is costly and inefficient, and lacks accurate quantitative analysis.
By collecting quality historical data of spare parts, using image recognition for automated detection and analysis, building neural network models, establishing dynamic evaluation mechanisms, and re-evaluating the warranty level based on historical data and detection data.
It realizes intelligent and dynamic management of spare parts warranty levels, improves judgment accuracy and efficiency, reduces human intervention, and improves management efficiency.
Smart Images
Figure CN119515160A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent predictive maintenance, and in particular to a management method and system for effectively determining the quality assurance level of spare parts in a nuclear power plant. Background Art
[0002] In the nuclear energy industry, component classification is crucial to ensuring the safe operation of nuclear power plants. Traditional methods rely on manual experience and periodic inspections, which are inefficient and slow to respond. With the advancement of industrial Internet of Things and big data analysis technologies, intelligent monitoring and predictive maintenance have become new trends. The combination of sensor networks and edge computing has realized real-time multimodal data collection and analysis of equipment status, significantly enhancing the reliability and economy of operation.
[0003] However, there are challenges in the current predictive maintenance system. The complexity of multimodal data requires efficient data fusion technology, but the existing methods have limited ability to integrate information, which affects the effect of feature extraction and model training. The component classification assessment does not fully consider the environment, load and aging factors, and the prediction accuracy is limited. In addition, the dynamic threshold adjustment and maintenance strategy formulation rely on fixed rules and lack flexibility and intelligence. Therefore, an innovative component classification determination method is urgently needed. It combines deep learning and integrated learning, uses multimodal data to accurately evaluate the health status of components, intelligently adjusts dynamic thresholds, and promotes the upgrade of predictive maintenance to autonomous intelligence. It introduces feature conversion functions and multimodal deep residual sequence integrated prediction models, coupled with a dynamic threshold adjustment mechanism, to significantly improve the accuracy and efficiency of determination, and to safeguard the safety and economic benefits of nuclear power plants. Summary of the invention
[0004] In view of the above-mentioned problems, the present invention is proposed.
[0005] Therefore, the technical problem solved by the present invention is that the traditional evaluation method is based on static data and cannot adjust the warranty level in real time according to the actual usage status or performance changes of the spare parts, resulting in insufficient accuracy of the warranty level evaluation. In the quality evaluation of nuclear power plant spare parts, surface defect detection is usually done manually, which is costly, inefficient and easily affected by human factors, lacks precise quantitative analysis, and the quality data of spare parts come from various sources and different dimensions, making it difficult to conduct unified evaluation and comparison, which affects the determination of the warranty level.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: A management method for effectively determining the quality assurance level of spare parts in nuclear power plants, which comprises the following steps:
[0007] Collect historical data on spare parts quality; use image recognition to automatically detect and analyze spare parts; build a neural network model; establish a dynamic evaluation mechanism to re-evaluate the warranty level based on the historical data and test data of spare parts.
[0008] As a preferred solution of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant described in the present invention, the quality history data includes the service life and failure rate of the spare parts.
[0009] As a preferred solution of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant described in the present invention, the image recognition includes collecting image data of spare parts through a camera and pre-processing the collected images using image processing technology.
[0010] As a preferred solution of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant described in the present invention, the construction of the neural network model includes inputting the preprocessed spare parts image into the neural network model, performing spare parts image analysis through the ResNet deep CNN architecture, and identifying defects on the spare parts surface through the Faster R-CNN algorithm.
[0011] As a preferred solution of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant described in the present invention, the target classification loss in the Faster R-CNN algorithm is expressed as:
[0012]
[0013] Among them, p i represents the probability that the candidate box i belongs to the target category, Represents the true category label.
[0014] The bounding box regression loss in the Faster R-CNN algorithm is expressed as,
[0015]
[0016] Among them, t i represents the bounding box parameters predicted by the network, represents the true bounding box parameters, Represents the loss function, which is specifically expressed as,
[0017]
[0018] Where x represents the difference between the bounding box parameters of the neural network model and the true bounding box parameters. The total loss function in the Faster R-CNN algorithm is expressed as,
[0019] L=L cls +λL bbox
[0020] Here, λ represents the balance factor.
[0021] As a preferred solution of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant described in the present invention, defect detection is performed on the surface of the spare parts through the neural network model to obtain detection data.
[0022] As a preferred solution of the management method for effectively determining the quality assurance level of nuclear power plant spare parts described in the present invention, the re-evaluation of the quality assurance level includes setting evaluation indicators and assigning weights, standardizing the service life and failure rate data to the range of 0-1, outputting defect information scores based on detection data, deriving a comprehensive score based on the service life, failure rate data and defect information scores, and dividing the spare parts into different quality assurance levels based on the comprehensive score.
[0023] Another object of the present invention is to provide a management system for effectively determining the warranty level of spare parts in nuclear power plants, which can achieve intelligent and dynamic management of spare parts warranty levels through automated data collection, image recognition and analysis, deep learning model evaluation, and dynamic warranty level adjustment, thereby solving the problems of existing warranty level determination relying on human experience, lack of real-time and dynamic adjustment in the evaluation process, low detection accuracy, and inability to optimize warranty management based on real-time status.
[0024] To solve the above technical problems, the present invention provides the following technical solutions: a management system for effectively determining the quality assurance level of spare parts in nuclear power plants, comprising a data acquisition module, an image preprocessing module, a neural network detection module, a defect detection and classification module, and a quality assurance level assessment module.
[0025] The data acquisition module is responsible for acquiring image data of spare parts.
[0026] The image preprocessing module preprocesses the collected spare part image data.
[0027] The neural network detection module analyzes the preprocessed image through a deep learning model to detect defects on the surface of the spare parts.
[0028] The defect detection and classification module classifies the defects on the surface of the spare parts according to the results output by the neural network detection module.
[0029] The quality assurance level assessment module calculates a comprehensive score using a weighted method based on standardized historical data and defect detection scores, and divides the quality assurance levels according to the scores.
[0030] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant as described above are implemented.
[0031] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant as described above.
[0032] Beneficial effects of the present invention: The present invention combines the historical quality data (such as service life, failure rate) and real-time detection data (such as surface defects, wear conditions) of spare parts by establishing a dynamic evaluation mechanism, and automatically evaluates the current quality status of spare parts. The evaluation system dynamically calculates the comprehensive score based on these data, and automatically adjusts the warranty level according to the preset rules, making the warranty evaluation more flexible and real-time. The historical data and detection data of spare parts are standardized, and the data such as service life and failure rate are mapped to the range of 0-1 to unify the data dimension. Then a weighted scoring mechanism is constructed, the standardized data is combined with the detection defect score, and the comprehensive score is calculated according to the preset weight to ensure that the impact of each quality data on the evaluation can be reasonably reflected in the warranty level. An intelligent evaluation system is established, the neural network model is used to automatically process the detection data, and the spare parts warranty level is determined in real time in combination with the dynamic evaluation mechanism. The system automatically calculates and determines the warranty level based on the scoring and rule engine, realizes the intelligent management of the spare parts warranty level, reduces human intervention and improves management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0034] Figure 1 An overall flow chart of a management method for effectively determining the quality assurance level of spare parts of a nuclear power plant provided in the first embodiment of the present invention.
[0035] Figure 2 An overall framework diagram of a management system for effectively determining the quality assurance level of spare parts for nuclear power plants provided in a second embodiment of the present invention. DETAILED DESCRIPTION
[0036] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0037] Example 1, reference Figure 1, which is an embodiment of the present invention, provides a management method for effectively determining the quality assurance level of spare parts in a nuclear power plant, which is characterized by:
[0038] S1: Collect quality history data of spare parts.
[0039] Quality history data includes the service life and failure rate of spare parts.
[0040] S2: Automated inspection and analysis of spare parts using image recognition;
[0041] Image recognition includes collecting image data of spare parts through a camera and pre-processing the collected images using image processing technology.
[0042] S3: Build a neural network model.
[0043] Building a neural network model includes inputting the preprocessed spare parts image into the neural network model, analyzing the spare parts image through the ResNet deep CNN architecture, and identifying defects on the spare parts surface through the Faster R-CNN algorithm.
[0044] The target classification loss in the Faster R-CNN algorithm is expressed as,
[0045]
[0046] Among them, p i represents the probability that the candidate box i belongs to the target category, Represents the true category label.
[0047] The bounding box regression loss in the Faster R-CNN algorithm is expressed as,
[0048]
[0049] Among them, t i represents the bounding box parameters predicted by the network, represents the true bounding box parameters, Represents the loss function, which is specifically expressed as,
[0050]
[0051] Where x represents the difference between the bounding box parameters of the neural network model and the true bounding box parameters. The total loss function in the FasterR-CNN algorithm is expressed as,
[0052] L=L cls +λL bbox
[0053] Here, λ represents the balance factor.
[0054] The neural network model performs defect detection on the spare part surface to obtain detection data.
[0055] The present invention adopts the deep CNN architecture of ResNet and Faster R-CNN to provide an accurate image analysis and defect detection solution. ResNet can extract the deep features of the image to help the model better understand the tiny details of the spare parts image, while Faster R-CNN ensures the accurate positioning of the surface defects of the spare parts through its target classification loss and bounding box regression loss function. Specifically, the target classification loss of Faster R-CNN is used to determine the category of the detection area, and the bounding box regression loss measures the difference between the predicted box and the real box. The final total loss function combines the results of the two to achieve accurate defect detection. The advantage of this step is that the automatic detection of surface defects through neural networks not only greatly improves the accuracy of detection, but also provides objective data support for the dynamic adjustment of subsequent quality assurance levels.
[0056] S4: Establish a dynamic evaluation mechanism to re-evaluate the warranty level based on the historical data and test data of spare parts.
[0057] Re-evaluating the warranty level includes setting evaluation indicators and assigning weights, standardizing the service life and failure rate data to the range of 0-1, outputting the defect information score based on the inspection data, deriving a comprehensive score based on the service life, failure rate data and defect information score, and dividing the spare parts into different warranty levels based on the comprehensive score.
[0058] In the evaluation mechanism, the present invention introduces multiple methods such as standardization, weighted scoring and dynamic adjustment. First, the service life and failure rate data are standardized to the range of 0-1, so that data of different dimensions can be compared. Secondly, the score is generated according to the defect detection data, and the comprehensive score is calculated by weighted calculation. The comprehensive score reflects the overall quality status of the spare parts, thereby dividing the spare parts into different warranty levels. The advantage of the dynamic evaluation mechanism is that it can adjust the warranty level in real time, respond to the latest changes in the spare parts status, and improve the scientificity and flexibility of spare parts warranty management.
[0059] Example 2, reference Figure 2 , which is an embodiment of the present invention, provides a system for managing a method for effectively determining the quality assurance level of spare parts in a nuclear power plant, characterized in that it includes a data acquisition module 100, an image preprocessing module 200, a neural network detection module 300, a defect detection and classification module 400 and a quality assurance level assessment module 500.
[0060] The data acquisition module 100 is responsible for acquiring image data of the spare parts.
[0061] The image preprocessing module 200 preprocesses the collected spare part image data.
[0062] The neural network detection module 300 analyzes the preprocessed image through a deep learning model to detect defects on the surface of the spare parts.
[0063] The defect detection and classification module 400 classifies the defects on the surface of the spare parts according to the results output by the neural network detection module.
[0064] The quality assurance level assessment module 500 calculates a comprehensive score by a weighted method based on standardized historical data and defect detection scores, and divides the quality assurance level according to the score.
[0065] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.
[0066] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0067] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0068] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0069] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
Claims
1. A management method for effectively determining the quality assurance level of spare parts in nuclear power plants, characterized in that: include: Collect quality history data of spare parts; Automated inspection and analysis of spare parts using image recognition; Build a neural network model; Establish a dynamic evaluation mechanism to re-evaluate the warranty level based on the historical data and test data of spare parts.
2. A management method for effectively determining the quality assurance level of nuclear power plant spare parts according to claim 1, characterized in that: The quality history data includes the service life and failure rate of spare parts.
3. A management method for effectively determining the quality assurance level of spare parts for nuclear power plants as claimed in claim 2, characterized in that: The image recognition includes collecting image data of the spare parts through a camera and pre-processing the collected images using image processing technology.
4. A management method for effectively determining the quality assurance level of spare parts for nuclear power plants as claimed in claim 3, characterized in that: The construction of the neural network model includes inputting the preprocessed spare part image into the neural network model, performing spare part image analysis through the ResNet deep CNN architecture, and identifying defects on the spare part surface through the Faster R-CNN algorithm.
5. A management method for effectively determining the quality assurance level of spare parts for nuclear power plants as claimed in claim 4, characterized in that: The target classification loss in the Faster R-CNN algorithm is expressed as, Among them, p i represents the probability that the candidate box i belongs to the target category, Represents the true category label; The bounding box regression loss in the Faster R-CNN algorithm is expressed as, Among them, t i represents the bounding box parameters predicted by the network, represents the true bounding box parameters, Represents the loss function, which is specifically expressed as, Where x represents the difference between the bounding box parameters of the neural network model and the true bounding box parameters; The total loss function in the Faster R-CNN algorithm is expressed as, L=L cls +λL bbox Here, λ represents the balance factor.
6. A management method for effectively determining the quality assurance level of spare parts for nuclear power plants as claimed in claim 5, characterized in that: Defect detection on the surface of spare parts is performed through the neural network model to obtain detection data.
7. A management method for effectively determining the quality assurance level of spare parts for nuclear power plants as claimed in claim 6, characterized in that: The re-evaluation of the warranty level includes setting evaluation indicators and assigning weights, standardizing the service life and failure rate data to the range of 0-1, outputting a defect information score based on the detection data, obtaining a comprehensive score based on the service life, failure rate data and defect information score, and dividing the spare parts into different warranty levels based on the comprehensive score.
8. A system adopting a management method for effectively determining the quality assurance level of nuclear power plant spare parts according to any one of claims 1 to 7, characterized in that: It comprises a data acquisition module (100), an image preprocessing module (200), a neural network detection module (300), a defect detection and classification module (400) and a quality assurance level assessment module (500); The data acquisition module (100) is responsible for acquiring image data of spare parts; The image preprocessing module (200) preprocesses the collected spare part image data; The neural network detection module (300) analyzes the preprocessed image through a deep learning model to detect defects on the surface of the spare part; The defect detection and classification module (400) classifies defects on the surface of the spare part according to the results output by the neural network detection module; The quality assurance level assessment module (500) calculates a comprehensive score by a weighted method based on standardized historical data and defect detection scores, and divides the quality assurance levels according to the scores.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a management method for effectively determining the quality assurance level of spare parts for a nuclear power plant as described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the management method for effectively determining the quality assurance level of spare parts of a nuclear power plant as described in any one of claims 1 to 7 are implemented.