Fuel assembly full-cycle data management system, method, apparatus, medium, and product

Abstract

This disclosure presents a fuel assembly full-lifecycle data management system, method, device, medium, and product. The system includes multiple data management subsystems: a data acquisition module, an operating condition prediction module, and a data encryption module. The data acquisition module collects operating data and environmental data of the fuel assembly during the current sampling period. The operating condition prediction module obtains the current operating status and predicted operating data based on the operating data, environmental data, and an operating condition prediction model, and adjusts the data management strategy based on the current operating status and predicted operating data. The data encryption module encrypts the operating data and environmental data to obtain encrypted data, and stores the encrypted data on a blockchain platform. The technical solution of this disclosure enables real-time and accurate monitoring of the operating status of nuclear power plant fuel assemblies throughout their entire lifecycle and rapid capture of sudden changes in operating conditions, improving the timeliness and accuracy of condition detection.

Description

Technical Field

[0001] This disclosure relates to the field of nuclear power plant data management technology, and in particular to a fuel assembly full-lifecycle data management system, method, device, medium and product. Background Technology

[0002] As high-risk, high-safety-level energy facilities, nuclear power plants' core equipment, fuel assemblies, endure extreme operating conditions such as high temperature, high pressure, and strong radiation during operation. Fuel assemblies undergo the entire lifecycle of the reactor core, from loading, operation, shutdown to decommissioning. Their safety and integrity directly affect the stable operation of the reactor and the controlled release of radioactive materials. Therefore, conducting full-process status monitoring and precise management of fuel assemblies has become a key technological means to ensure the safe operation of nuclear power plants.

[0003] Currently, the detection of fuel assembly status mostly adopts a combination of manual periodic inspections and offline data collection by single-point sensors. This means that maintenance personnel record limited parameters such as temperature and pressure on-site, or collect data periodically through fixed-installation single-type sensors and store it on a local terminal. Subsequently, technicians review the data offline and manually draw operating condition curves. It is only possible to determine whether there is an anomaly by simple fixed numerical thresholds (such as temperature exceeding a certain constant value).

[0004] However, this detection method has several drawbacks: First, it is extremely time-sensitive, with long intervals between manual inspections and delayed offline data analysis, making it impossible to capture sudden changes and gradual trends in the operating conditions of fuel components in a timely manner, and making it difficult to provide timely warnings of potential risks. Second, it lacks fault tolerance and self-healing capabilities, and abnormal data caused by radiation interference or data transmission failures cannot be identified or corrected. Once data deviations occur, manual re-collection and verification are the only options. Third, it relies entirely on human experience to judge the operating status, which is not only inefficient but also highly susceptible to safety hazards due to human error. Summary of the Invention

[0005] This invention provides a fuel assembly full-lifecycle data management system, method, equipment, medium, and product to achieve real-time and accurate monitoring of the operating status of nuclear power plant fuel assemblies throughout their entire lifecycle and rapid capture of sudden changes in operating conditions, thereby improving the timeliness and accuracy of condition detection.

[0006] According to one aspect of the present invention, a fuel assembly full-lifecycle data management system is provided, the system comprising: multiple distributed data management subsystems, each of the data management subsystems comprising: a data acquisition module, an operating condition prediction module, and a data encryption module, wherein, The data acquisition module is used to acquire the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located during the current sampling period; The operating condition prediction module is used to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data of the next sampling period based on the collected current operating data, the current environmental data and the pre-trained operating condition prediction model, and to adjust the data management strategy of the data management subsystem based on the current operating state and the predicted operating data. The data encryption module is used to encrypt the collected current running data and the current environment data to obtain encrypted data corresponding to the current collection period, and to store the encrypted data and the current sampling period together in the blockchain platform.

[0007] According to another aspect of the present invention, a method for full-cycle data management of fuel assemblies is provided, the method being applied to a data management subsystem, comprising: The data acquisition module collects the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located within the current sampling period. The operating condition prediction module obtains the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period based on the collected current operating data, the current environmental data, and the pre-trained operating condition prediction model. The data management strategy of the data management subsystem is then adjusted based on the current operating state and the predicted operating data. The data encryption module encrypts the collected current running data and current environment data to obtain encrypted data corresponding to the current sampling period, and then stores the encrypted data and the current sampling period together on the blockchain platform.

[0008] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: One or more processors; Storage device for storing one or more programs. When one or more programs are executed by one or more processors, the one or more processors implement a fuel component full-cycle data management method as described in any of the embodiments of this disclosure.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement any of the fuel component full-cycle data management methods of the present invention.

[0010] According to another aspect of the present disclosure, a computer program product is provided, which, when executed by a processor, implements a fuel component full-cycle data management method as described in any of the embodiments of the present disclosure.

[0011] This disclosure provides a fuel assembly full-cycle data management system, including multiple data management subsystems. Each data management subsystem includes: a data acquisition module, an operating condition prediction module, and a data encryption module. The data acquisition module is used to acquire current operating data of the fuel assembly and current environmental data of the environment in which the fuel assembly is located within the current sampling period. The operating condition prediction module is used to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period based on the acquired current operating data, current environmental data, and a pre-trained operating condition prediction model. It also adjusts the data management strategy of the data management subsystem based on the current operating state and the predicted operating data. The encryption module is used to encrypt the collected current operating data and current environmental data to obtain encrypted data corresponding to the current collection cycle. The encrypted data and the current sampling cycle are then associated and stored on the blockchain platform. This solves the problems of poor timeliness in fuel assembly condition detection in related technologies, which makes it impossible to capture sudden changes in operating conditions during fuel assembly operation in a timely manner, resulting in low accuracy and efficiency in fuel assembly condition detection. It enables real-time and accurate monitoring of the operating status of nuclear power plant fuel assemblies throughout their entire life cycle and rapid capture of sudden changes in operating conditions, improving the timeliness and accuracy of condition detection. At the same time, relying on data encryption and evidence storage capabilities, it ensures the continuity, credibility, and security of data, ultimately significantly enhancing the operational safety of fuel assemblies and the controllability of nuclear power plant operation.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used 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 those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A schematic diagram of the structure of a fuel assembly lifecycle data management system provided in this embodiment of the disclosure; Figure 2 A flowchart illustrating a fuel assembly lifecycle data management method provided in this embodiment of the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0015] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0016] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0017] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0018] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0019] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0020] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0021] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0022] Figure 1 This is a schematic diagram of a fuel assembly lifecycle data management system provided in an embodiment of this disclosure. This embodiment is applicable to managing the entire lifecycle data of nuclear power plant fuel assemblies. Figure 1 As shown, the fuel assembly lifecycle data management system provided in this embodiment includes multiple data management subsystems 10. Each data management subsystem 10 includes: a data acquisition module 101, an operating condition prediction module 102, and a data encryption module 103. The structural composition of the fuel assembly lifecycle data management system 10 provided in this embodiment will be described in detail below.

[0023] The data acquisition module 101 is used to acquire the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located within the current sampling period; the operating condition prediction module 102 is used to obtain the current operating state of the fuel assembly in the current sampling period and / or the predicted operating data of the next sampling period based on the acquired current operating data, current environmental data and a pre-trained operating condition prediction model, and to adjust the data management strategy of the data management subsystem 10 based on the current operating state and / or predicted operating data; the data encryption module 103 is used to encrypt the acquired current operating data and current environmental data to obtain encrypted data corresponding to the current sampling period, and to associate and store the encrypted data and the current sampling period on the blockchain platform.

[0024] The fuel assembly lifecycle data management system can be understood as an integrated management platform that collects operational and environmental data, predicts operating conditions, encrypts data, and stores information on the blockchain throughout the entire lifecycle of nuclear fuel assemblies, from factory loading, in-reactor operation, shutdown maintenance, to decommissioning. The system consists of multiple distributed data management subsystems 10. The core objective of the fuel assembly lifecycle data management system is to ensure the reliability and traceability of fuel assembly lifecycle data, as well as the real-time and intelligent monitoring of operating conditions. Each data management subsystem 10 can be understood as a distributed functional unit of the fuel assembly lifecycle data management system, serving as the core carrier for localized management of fuel assembly data. Each data management subsystem 10 can be independently deployed in the monitoring area where the fuel assembly is located (e.g., different zones of the reactor core). Each data management subsystem 10 possesses complete capabilities for data acquisition, operating condition prediction, and data encryption, enabling autonomous local data processing and decision-making. It also supports collaborative interaction with other subsystems and the blockchain platform, avoiding the latency and single-point-of-failure risks associated with centralized architectures.

[0025] It should be noted that the distributed architecture of multiple data management subsystems 10 in the fuel assembly lifecycle data management system has the following advantages: each data management subsystem 10 can independently complete the collection, condition prediction, and encrypted storage of local fuel assembly operating data and environmental data, without relying on a centralized server for data transfer and command issuance, significantly reducing data transmission latency and enabling real-time monitoring and rapid response to the fuel assembly's operating status; at the same time, when a single data management subsystem 10 experiences sensor failures, radiation interference, or other anomalies, it will not affect the normal operation of other data management subsystems 10, and adjacent data management subsystems 10 can share... By leveraging non-sensitive feature data to collaboratively complete operational condition analysis, the risk of single-point failure in centralized architectures is effectively avoided, improving the overall fault tolerance and anti-disturbance capability of the system. In addition, data is encrypted on local nodes before being uploaded to the blockchain platform, eliminating the need for cross-node transmission of core operational data, reducing the risk of data leakage and tampering from the source, and ensuring data privacy and security. Furthermore, when adding monitoring areas, only the corresponding data management subsystem 10 needs to be installed, without requiring large-scale modifications to the existing system architecture, significantly improving the system's scalability and deployment flexibility, and adapting to the refined management needs of the entire lifecycle of fuel assemblies from factory loading, in-pile operation to decommissioning.

[0026] It should also be noted that the data management subsystem 10 can perform full lifecycle data management for a single fuel assembly or multiple fuel assemblies. When the data management subsystem 10 performs full lifecycle data management for multiple fuel assemblies, the technical solutions provided in this embodiment can be used for data acquisition, operational condition prediction, and data encryption. The following description uses the full lifecycle data management process of a single fuel assembly by the data management subsystem 10 as an example to illustrate the technical solutions provided in this embodiment.

[0027] The data acquisition module 101 can be understood as a hardware and software integrated unit in the data management subsystem 10 responsible for acquiring raw data. The data acquisition module 101 may include at least one sensor, which may include at least one of the following: temperature sensor, pressure sensor, vibration sensor, radiation dose sensor, etc. The data acquisition module 101 may also include a data acquisition driver and / or a module integrating filtering and noise reduction algorithms. The core function of the data acquisition module 101 is to acquire the operating data of the fuel assembly and the environmental data of the surrounding environment of the fuel assembly according to a preset sampling period, providing a basic data source for subsequent operating condition prediction and data encryption. The fuel assembly can be understood as the detection object of the data management subsystem 10. It can be understood that the fuel assembly can be the core structural unit in the reactor core of a nuclear power plant used to realize nuclear fission reactions and release nuclear energy, typically assembled from components such as fuel rods, positioning grids, guide tubes, upper tube seats, and lower tube seats. The fuel assembly can include fuel assemblies at various stages of their entire life cycle, including: the loading stage after delivery, the in-core stable operation stage, the power adjustment stage, the shutdown maintenance stage, and the decommissioning disposal stage. Fuel assemblies can also encompass different types and specifications, such as pressurized water reactor fuel assemblies and boiling water reactor fuel assemblies. The data management subsystem 10 can collect, analyze, and store the operating parameters and key parameters of the fuel assembly throughout its entire lifecycle, ultimately achieving real-time monitoring, risk warning, and safety management of its operating status. The current sampling period can refer to the data interval in which the data acquisition module 101 completes one full data acquisition. The duration of the current sampling period can be dynamically configured according to the operating stage of the fuel assembly or can be manually customized. The current sampling period can serve as the basic time unit for data processing, condition prediction, and encrypted storage by the data management subsystem 10. Three adjacent sampling periods can be sequentially denoted as: previous sampling period, current sampling period, and next sampling period. Current operating data can refer to the operating parameters of the fuel assembly itself collected by the data acquisition module 101 within the current sampling period. Alternatively, current operating data can be understood as the operating data generated by the fuel assembly within the current sampling period. Current operating data can include, but is not limited to, physical quantities that directly reflect the operating status of the assembly, such as assembly surface temperature, internal pressure, vibration amplitude, and power output value. Current operational data is a core basis for determining whether the fuel assembly is in a safe operating state. Current environmental data refers to the parameters of the surrounding environment of the fuel assembly collected by the data acquisition module 101 during the current sampling period. Current environmental data may include, but is not limited to, external environmental physical quantities that affect the operating state of the fuel assembly, such as reactor coolant temperature, coolant flow rate, ambient radiation dose, and in-core humidity.

[0028] In this embodiment, a data management subsystem 10 can be used to manage the full-cycle data of the fuel assembly. During the operation of the fuel assembly, the data acquisition module 101 collects at least one operating parameter of the fuel assembly and environmental parameters of the environment in which the fuel assembly is located according to a preset sampling period. This allows the acquisition of current operating data and current environmental data of the fuel assembly within the current sampling period. Furthermore, the collected current operating data and current environmental data can be input into the operating condition prediction module 102.

[0029] The operating condition prediction module 102 can refer to the core functional unit in the data management subsystem 10 responsible for operating condition analysis and prediction. The operating condition prediction module 102 can integrate three sub-functions: feature extraction, model inference, and strategy optimization. Based on a pre-trained operating condition prediction model, the operating condition prediction module 102 can analyze current operating data and current environmental data, output the current operating status and predicted operating data for the next sampling period, and adjust the data management strategy of the data management subsystem 10 accordingly. The operating condition prediction module 102 can be the core module for realizing intelligent decision-making in the data management subsystem 10. The operating condition prediction model can refer to a deep learning model pre-trained based on historical operating data of fuel components, environmental data, and operating condition labels. The model structure of the operating condition prediction model can include long short-term memory networks, support vector regression, or ensemble tree models, etc. The input to the operating condition prediction model can be the current operating data and current environmental data for the current sampling period. The output consists of two types of results: first, the current operating state of the fuel assembly in the current sampling period (including the current operating stage and / or operating risk level); and second, the predicted operating data for the next sampling period (such as predicted temperature, predicted pressure, predicted vibration data, etc.). In this embodiment, the operating condition prediction model can support online incremental learning and dynamically optimize model parameters based on changes in operating conditions. The current operating state can refer to the life cycle stage and / or operating risk level of the fuel assembly within the current sampling period, as output by the operating condition prediction model. The current operating state can be a qualitative and quantitative description of the real-time operating condition of the fuel assembly. The classification of the current operating state can be based on the design standards and safety thresholds of the fuel assembly. The current operating state can include, but is not limited to: initial loading state, stable operating state, abnormal operating state, power adjustment state, shutdown preparation state, fault warning state, etc. The predicted operating data can refer to the predicted operating parameters of the fuel assembly in the next sampling period, obtained by the operating condition prediction model based on the current operating data and current environmental data. Optionally, the predicted operating data may include predicted temperature, predicted pressure, predicted vibration amplitude, etc. Predicted operating data can anticipate changes in fuel assembly operating conditions, providing a forward-looking basis for adjusting the data management strategy of the data management subsystem 10. The data management strategy can refer to the data processing and storage rules dynamically adjusted by the data management subsystem 10 based on the current operating status and predicted operating data. Data management strategies may include, but are not limited to: sensor sampling frequency adjustment strategies (e.g., increasing the sampling frequency during fault warnings), data encryption level switching strategies (e.g., enabling advanced homomorphic encryption under high-risk conditions), and data upload frequency strategies to the blockchain (e.g., delayed upload under stable conditions, real-time upload under abnormal conditions). The core objective of adjusting the data management strategy is to optimize system resource consumption and improve anomaly response efficiency while ensuring data security.

[0030] In this embodiment, when the operating condition prediction module 102 receives the collected current operating data and current environmental data, it can input the received current operating data and current environmental data into a pre-trained operating condition prediction model. Furthermore, the operating condition prediction model processes the current operating data and current environmental data, and outputs the current operating state of the fuel assembly in the current sampling period and the predicted operating data of the fuel assembly in the next sampling period. Furthermore, the data management strategy of the data management subsystem 10 can be adjusted based on the obtained current operating state and predicted operating data.

[0031] In this embodiment, to reduce data redundancy, improve model computational efficiency, and enable the operating condition prediction model to learn and identify operating condition change patterns more efficiently, feature extraction is performed on the current operating data and current environmental data. From the original multi-dimensional operating data and environmental data, key features that can accurately characterize the operating conditions of the fuel assembly are screened and refined to form a structured state feature vector. This state feature vector is then input into the operating condition prediction model to obtain the current operating state and predicted operating data.

[0032] Optionally, the operating condition prediction module 102 includes a feature extraction unit and a strategy update unit; wherein, the feature extraction unit is used to preprocess the collected current operating data and current environmental data, and extract features from the preprocessed current operating data and current environmental data to obtain a state feature vector corresponding to the current sampling period; the strategy update unit is used to input the state feature vector corresponding to the current sampling period into the pre-trained operating condition prediction model to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data of the next sampling period, and adjust the data management strategy of the data management subsystem 10 based on the current operating state and the predicted operating data.

[0033] The feature extraction unit can refer to the functional subunit in the operating condition prediction module 102 that undertakes the preprocessing of raw data and feature mining. The feature extraction unit can preprocess the received current operating data and current environment data, and extract features from the preprocessed current operating data and current environment data to integrate them into a state feature vector with fixed dimensions that can be directly input into the operating condition prediction model. Preprocessing can refer to the data purification and standardization operations performed by the feature extraction unit on the received raw data. Optionally, preprocessing includes at least one of denoising, normalization, and dimension alignment. For example, the steps for preprocessing the current operating data and current environment data can be: using the sliding window averaging method or the Laida criterion (3... The criteria for removing outliers caused by radiation interference from the current operating data and current environmental data are as follows: Furthermore, missing values ​​from the data acquisition process are filled in using linear interpolation algorithms. Additionally, parameters with different physical dimensions are scaled to a unified numerical range using min-max standardization or Z-score standardization, while the current operating data and current environmental data are time-series aligned to ensure consistency in the time dimension. Feature extraction from the preprocessed current operating data and current environmental data can be performed using at least one of the following methods: principal component analysis, autoencoder, or convolutional feature extraction. The state feature vector refers to the multi-dimensional data set output by the feature extraction unit, used to quantitatively characterize the current operating condition of the fuel assembly, and is the core input of the operating condition prediction model. Each dimension of the state feature vector corresponds to a key feature reflecting the operating state of the fuel assembly, for example, it can be represented as [mean temperature, pressure variance, dominant vibration frequency, coolant flow trend slope, peak environmental radiation dose]. The number of dimensions included in the state feature vector can be flexibly set according to the fuel assembly detection requirements and model complexity; changes in vector values ​​directly correspond to fluctuations in the operating condition of the fuel assembly.

[0034] The strategy update unit, within the operating condition prediction module 102, is a functional subunit responsible for model inference calculation and dynamic optimization of management strategies. It serves as the core carrier for realizing intelligent decision-making in the data management subsystem 10. The strategy update unit inputs state feature vectors into a pre-trained operating condition prediction model, obtains the current operating state of the fuel assembly and the predicted operating data for the next sampling period through model inference, and automatically adjusts the data management strategy of the data management subsystem 10 based on the model output.

[0035] In one implementation, when the operating condition prediction module 102 receives the collected current operating data and current environmental data, it can input the received current operating data and current environmental data into the feature extraction unit. Further, the feature extraction unit can preprocess the current operating data and current environmental data to obtain preprocessed current operating data and preprocessed current environmental data. Further, principal component analysis (PCA) can be used to extract features from the preprocessed current operating data and preprocessed current environmental data to obtain at least one extracted key feature. Integrating the at least one extracted key feature into a vector yields a state feature vector corresponding to the current sampling period. Further, the state feature vector can be input into the strategy update unit. Upon receiving the state feature vector, the strategy update unit can input the state feature vector corresponding to the current sampling period into a pre-trained operating condition prediction model to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period. Based on the current operating state and the predicted operating data, the data management strategy of the data management subsystem 10 is adjusted.

[0036] In this embodiment, before inputting the state feature vector corresponding to the current sampling period into the pre-trained operating condition prediction model, the operating condition prediction model can be trained first through the operating condition prediction module 102. The operating condition prediction model can include two training stages: an offline training stage and an online training stage. Furthermore, the operating condition prediction module 102 can also include an offline training unit and an online training unit.

[0037] Optionally, the operating condition prediction module 102 further includes an offline training unit and an online training unit. The offline training unit is used to train a pre-built deep learning model offline based on pre-stored historical operating data and historical environment data to obtain an offline-trained operating condition prediction model. The online training unit is used to train the offline-trained operating condition prediction model based on the operating data and environment data corresponding to the current sampling period and the previous operating state and data of the previous sampling period to obtain a trained operating condition prediction model.

[0038] The offline training unit refers to the functional subunit within the operating condition prediction module 102 that undertakes initial training. It is the core carrier for generating the basic operating condition prediction model (i.e., the operating condition prediction model completed offline). The offline training unit can be used during the initial deployment of the data management subsystem 10 or during periodic maintenance to batch train a pre-built deep learning model using pre-stored historical operating data and historical environmental data from the entire lifecycle of the fuel assembly. By iteratively optimizing parameters such as model weights and biases, the model gains the ability to initially identify the operating status of the fuel assembly and predict operating condition trends. The final output is an offline-trained operating condition prediction model that can be directly applied online. Historical operating data refers to the set of operating parameters of the fuel assembly accumulated and stored before or during the deployment of the data management subsystem 10, covering various historical sampling periods. Optionally, historical operating data includes physical quantities that directly reflect the operating status of the fuel assembly, such as historical temperature, historical pressure, historical vibration amplitude, and historical power output. Historical operating data can cover the entire lifecycle stages of the fuel assembly, including factory loading, stable operation, shutdown maintenance, and decommissioning, and has undergone preprocessing such as noise reduction and normalization. Historical environmental data refers to the set of parameters accumulated and stored in the data management subsystem 10 before deployment or during operation, representing the historical environment in which the fuel assembly was located. Optionally, historical environmental data includes external physical quantities that affect the operating status of the fuel assembly, such as historical coolant temperature, historical coolant flow rate, historical environmental radiation dose, and historical in-reactor humidity. Historical environmental data can be used in conjunction with historical operating data as input for offline training to improve the accuracy of the model's judgment of operating conditions.

[0039] Deep learning models refer to pre-built machine learning models based on multi-layer neural network structures, which are the core algorithms for predicting operating conditions. Considering the characteristics of fuel assembly time-series operating condition data, the model structure of deep learning models can include temporal neural networks such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU), or combine them with Convolutional Neural Networks (CNN) to extract spatiotemporal features of the data, or support vector regression, tree-based regression models, etc. The constructed deep learning models can capture the coupling and correlation of multiple parameters and the nonlinear laws governing operating condition changes, and are used to learn the operating condition change trends of fuel assemblies at different life cycle stages. Offline-trained operating condition prediction models refer to basic models that meet preset performance indicators after batch training by offline training units. Offline-trained operating condition prediction models can include optimized weights, biases, and hyperparameter configurations, and can be directly used as the initial model for online training units, supporting incremental iterative optimization during system operation.

[0040] In one implementation, the offline training process of the operating condition prediction model can be as follows: The offline training unit retrieves pre-stored historical operating data and historical environmental data from the data management subsystem 10. The retrieved historical operating data and historical environmental data are preprocessed and feature extracted to obtain structured historical feature vectors, which are then divided into training, validation, and test sets according to a preset ratio. Next, a pre-built deep learning model is loaded, and the training set is input into the deep learning model for batch iterative training. The forward propagation algorithm determines the loss value between the model's output operating condition prediction result and the actual operating condition label. Then, the backpropagation algorithm updates the model parameters such as weights and biases along the loss gradient descent direction. Furthermore, a validation set is used to test model performance. When the validation set loss value increases continuously for multiple rounds, a stopping mechanism is triggered to prevent model overfitting. Furthermore, the test set is used to evaluate the performance of the trained deep learning model. If the accuracy of the working condition identification and the parameter prediction error of the trained deep learning model meet the preset threshold, the trained deep learning model can be stored in the local model library as an offline trained working condition prediction model for subsequent use and iterative optimization by the online training unit.

[0041] The online training unit, specifically within the operating condition prediction module 102, is a functional sub-unit responsible for incremental iterative optimization of the model. It serves as the core carrier for dynamically adapting the model to changes in fuel assembly operating conditions. During real-time system operation, the online training unit continuously fine-tunes the offline-trained operating condition prediction model using current operating data and environmental data from the current sampling period, along with the previous operating state and data from the previous sampling period, as incremental training data. By dynamically updating model parameters, the model accurately adapts to real-time fluctuations in fuel assembly operating conditions (such as power adjustments and changes in environmental radiation), ultimately outputting a trained operating condition prediction model that highly matches the current operating conditions. The previous operating state refers to the specific operating condition category of the fuel assembly in the previous sampling period, including stable operation, operating condition fluctuations, and fault warning states. The previous operating state serves as supervisory information (real label) for online training, assisting the model in judging the rationality of current operating condition changes and improving the efficiency and accuracy of incremental training. The previous operating data refers to the set of operating parameters of the fuel assembly in the previous sampling period, which is key data connecting the operating conditions trends of adjacent sampling periods. The previous operating data can be combined with the current operating data and the current environmental data to construct short-cycle time-series features, assisting the online training unit in optimizing the model's ability to capture gradual changes in operating conditions.

[0042] In one implementation, an offline-trained operational condition prediction model is retrieved as a baseline model via an online training unit. The system also acquires current operational data and current environmental data corresponding to the current sampling period, as well as previous operational data and previous operational state from the previous sampling period. Preprocessing and feature extraction are performed on the current operational data and current environmental data to obtain a state feature vector corresponding to the current sampling period. This state feature vector, the previous operational data, and the previous operational state are then integrated into an incremental training dataset. Further, the incremental training dataset can be input into the baseline model, and the model's prediction results are compared with the previous operational data and previous operational state to determine the loss value between the model's output and the true label. Furthermore, an adaptive learning rate strategy (such as the Adam optimizer) is employed to fine-tune the model's weights, biases, and other parameters along the loss gradient descent direction to optimize the model's ability to capture the associated features of operational conditions in adjacent sampling periods. Furthermore, by verifying in real time whether the current model outputs the accuracy of operating condition identification and the parameter prediction residuals meet the preset requirements, if they do, the updated model is determined as the trained operating condition prediction model and is directly used for operating condition analysis in the current sampling period. If they do not meet the requirements, the model continues to be iteratively optimized based on incremental data in subsequent sampling periods to ensure that the model always adapts to the real-time operating condition changes of the fuel assembly.

[0043] In this embodiment, the data management subsystem 10 includes: a data encryption module 103; the data encryption module 103 is used to encrypt the collected current operating data and current environmental data to obtain encrypted data corresponding to the current collection cycle, and to associate and store the encrypted data and the current sampling cycle in a blockchain platform associated with the fuel component full life cycle data management system.

[0044] The data encryption module 103 can be understood as the core functional unit of the data management subsystem 10, responsible for protecting the privacy of original data, verifying its integrity, and ensuring its trusted storage. It is a key carrier for achieving data security management. The data encryption module 103 can integrate encryption algorithms, checksum generation algorithms, etc., and can encrypt the operational and environmental data collected by the data acquisition module 101. It then binds the encrypted data to the sampling period and stores it synchronously on the blockchain platform, ensuring the immutability and privacy of the data throughout its entire lifecycle. Encrypted data refers to the ciphertext data generated by the data encryption module 103 after processing the current operational and environmental data using a preset encryption algorithm. Encrypted data can directly participate in operational condition analysis and calculations in its undecrypted state, ensuring the privacy of the original data without affecting subsequent model predictions and trend analysis, possessing the characteristics of uniqueness and immutability. The blockchain platform can refer to a distributed ledger platform using a consortium blockchain architecture, with nodes covering authorized entities such as various data management subsystems, the nuclear power plant operation and maintenance center, and regulatory agencies. Blockchain platforms can leverage the decentralized and immutable characteristics of blockchain to store encrypted data and information related to sampling cycles, providing a reliable basis for compliance auditing and traceability of fuel component lifecycle data.

[0045] In this embodiment, after the data acquisition module 101 acquires the current operating data and current environmental data of the fuel assembly in the current sampling period, the acquired current operating data and current environmental data can be input to the data encryption module 103. Furthermore, the data encryption module 103 can encrypt the current operating data and current environmental data using a preset encryption algorithm to obtain encrypted data. Furthermore, the encrypted data and the current sampling period can be associated and stored on a blockchain platform linked to the fuel assembly full-lifecycle data management system.

[0046] Optionally, the data encryption module 103 includes: a data encryption unit, a checksum generation unit, and a context-aware unit. The data encryption unit generates a current data vector corresponding to the current sampling period based on the current running data and current environment data, and encrypts the current data vector to obtain a current ciphertext vector corresponding to the current sampling period. The checksum generation unit generates a target checksum based on the current running data, current environment data, and the current sampling period. The context-aware unit performs data trend anomaly detection based on the current data vector and the previous data vector from the previous sampling period during the generation of the current ciphertext vector and the target checksum. If no data trend anomaly is detected, the current ciphertext vector and the target checksum are used as encrypted data corresponding to the current sampling period, and the encrypted data and the current sampling period are associated and stored on a blockchain platform associated with the fuel component full-cycle data management system.

[0047] The data encryption unit can be understood as the core functional subunit of the data encryption module 103, responsible for converting raw data into ciphertext. The data encryption unit integrates the current operating data and current environmental data of the current sampling period into a structured data vector, encrypts this data vector, and finally outputs a current ciphertext vector with privacy protection characteristics. The data encryption unit ensures the security of data during storage and transmission. The verification code generation unit can be understood as the functional subunit of the data encryption module 103 responsible for generating data integrity verification credentials. The verification code generation unit generates a unique target verification code based on the current data vector and the current sampling period. The context-aware unit can be understood as the functional subunit of the data encryption module 103 that implements data trend anomaly detection and encryption process control. The context-aware unit is used to perform time-series trend analysis by linking the current data vector with the previous data vector, determining whether data changes meet the normal operating conditions of the fuel component, and is a pre-verification step to ensure the trustworthiness of data stored in the blockchain.

[0048] The current data vector refers to a structured multidimensional data set formed by the data encryption unit integrating the current operating data and current environmental data of the current sampling period according to a preset dimensional order. It serves as the unified input carrier for encryption processing. For example, the current data vector can be represented as [temperature value, pressure value, vibration amplitude, coolant flow rate, environmental radiation dose], with fixed dimensions and data types, facilitating standardized processing by the encryption algorithm. The current ciphertext vector refers to a multidimensional data set in ciphertext form generated after the current data vector is processed by the preset encryption algorithm of the data encryption unit. The dimensions of the current ciphertext vector completely correspond to the current data vector, but the data content is presented in ciphertext form. The preset encryption algorithm can refer to a cryptographic algorithm pre-configured by the data encryption module 103, suitable for encrypting multidimensional numerical data. Preferably, the preset encryption algorithm can include homomorphic encryption algorithms (such as the CKKS algorithm and the BFV algorithm). It can be understood that homomorphic encryption algorithms can support addition, subtraction, multiplication, and division operations in the ciphertext state, meeting the needs of subsequent operational trend analysis without decryption. Alternatively, a symmetric encryption algorithm can be selected as the preset encryption algorithm according to the scenario to achieve efficient privacy protection.

[0049] The target checksum refers to the data integrity verification identifier output by the checksum generation unit. The target checksum is typically generated using a hash algorithm or segmented verification rules, and can be bound to the current ciphertext vector. By comparing the checksums, it can be quickly determined whether the encrypted data has been tampered with during transmission or storage. Optionally, the target checksum can be generated using segmented verification rules, meaning it can be composed of multiple checksum segments. The target checksum may include a first checksum segment, a second checksum segment, and a third checksum segment. The first checksum segment can be used to describe the characteristic entropy of the data collected in real time by the data acquisition module 101; the characteristic entropy can be used to quantify the volatility and complexity of the batch of data. The second checksum segment can be used to mark the acquisition time of the data collected in real time by the data acquisition module 101. The third checksum segment can be used to reflect the trust level of the data management subsystem 10. The first, second, and third checksum segments are concatenated in a preset data structure order, and the resulting multi-segment checksum is used as the target checksum. In the multi-segment checksum system, each checksum segment can dynamically adjust its encoding length or redundancy mechanism according to the sampled data or risk level, achieving a flexible configuration with "variable structure."

[0050] The previous data vector can refer to the set of plaintext multidimensional data generated by the data encryption unit in the previous sampling period, which has the same dimensions as the current data vector. It serves as the historical reference benchmark for the context-aware unit to detect trend anomalies. Data trend anomaly prediction refers to the time-series data consistency verification operation performed by the context-aware unit. Specifically, it determines the rate of change, deviation magnitude, and fitting residual of each dimension of the current data vector and the previous data vector, compares them with preset thresholds, and judges whether the data change belongs to non-operating condition abnormal fluctuations (such as sensor failure or data transmission interference).

[0051] In this embodiment, upon receiving the collected current operating data and current environment data through the data encryption module 103, the data encryption unit within the data encryption module 103 can format and standardize the current operating data and current environment data, converting them into a structured multidimensional vector to obtain the current data vector. Furthermore, the current data vector is encrypted using a homomorphic encryption algorithm within the data encryption unit to obtain the current ciphertext vector.

[0052] In one implementation, when the target checksum includes a multi-segment checksum, the checksum generation unit can perform information entropy analysis on the current running data and current environment data. Using an information entropy determination method (such as Shannon entropy or an improved weighted entropy model), the distribution characteristics of the current running data and current environment data are extracted to obtain a set of parameters characterizing their data entropy values. This parameter segment is used as the first checksum segment. Further, using the start time of the current sampling period as a reference, a timestamp is generated using a high-precision clock. This generated timestamp is used as the second checksum segment for consistency verification in the time dimension. Further, a trust level calculation is performed on the data management subsystem 10. The trust level can be comprehensively derived based on indicators such as the historical stability, anomaly rate, and consensus participation frequency of the data management subsystem 10. The calculated trust level is used as the third checksum segment. Further, the first, second, and third checksum segments can be concatenated according to a preset data structure order, and the resulting multi-segment checksum is used as the target checksum.

[0053] In this embodiment, data trend anomaly detection is performed using a context-aware unit, which can determine the indicator value of at least one data trend indicator based on the current data vector and the previous data vector. Furthermore, the indicator value can be compared with a preset threshold to determine whether a data trend anomaly is stored.

[0054] Optionally, the context-aware unit includes: a context modeling subunit, a fitting deviation determination subunit, and a deviation judgment subunit; wherein, the context modeling subunit is used to determine the data trend vector based on the current data vector and the previous data vector; the fitting deviation determination subunit is used to fit the data trend vector, construct a theoretical expected value prediction function, input the current data vector into the theoretical expected value prediction function to obtain the theoretical expected value corresponding to the current sampling period, and determine the fitting deviation corresponding to the current sampling period based on the current data vector and the theoretical expected value; the deviation judgment subunit is used to compare the fitting deviation with a preset fitting deviation threshold, and if the fitting deviation is less than or equal to the fitting deviation threshold, it is determined that no data trend anomaly has been detected, the current ciphertext vector and the multi-segment check code are used as encrypted data corresponding to the current sampling period, and the encrypted data and the current sampling period are associated and stored in a blockchain platform associated with the fuel component full-lifecycle data management system.

[0055] The context modeling subunit refers to the functional subunit within the context awareness unit responsible for constructing a correlation model of data trends between adjacent periods. This subunit receives the current and previous data vectors and generates a data trend vector characterizing the continuous changes in fuel assembly operating conditions by calculating the difference, rate of change, and temporal correlation of corresponding dimensions between the two vectors. This provides a trend benchmark for subsequent calculations of theoretical expected values. The fitting deviation determination subunit refers to the functional subunit within the context awareness unit responsible for trend fitting and deviation quantification calculations. The deviation determination subunit refers to the functional subunit within the context awareness unit that performs the final determination of data trend anomalies and manages the encryption process. The data trend vector refers to the temporal change feature vector calculated by the context modeling subunit based on the current and previous data vectors. Each dimension of the data trend vector corresponds to the rate of change or difference of the same dimension in the original data vector, visually reflecting the changing trend of fuel assembly operating conditions within adjacent sampling periods. The theoretical expected value prediction function refers to the data prediction model obtained by the fitting deviation determination subunit based on the data trend vector. The theoretical expected value prediction function can predict the theoretical expected values ​​of each dimension of the current data vector based on historical data trends. The output dimensions completely correspond to the current data vector, serving as an ideal reference value for the current operating conditions. The expression of the theoretical expected value prediction function can be a linear function, a polynomial function, or an LSTM time series prediction model, depending on the characteristics of the operating conditions. The theoretical expected value prediction function can be determined by linear or nonlinear fitting of the data trend vector. Linear fitting can include least squares fitting. Nonlinear fitting can include support vector regression, polynomial fitting, or locally weighted fitting. The theoretical expected value refers to the reasonable reference values ​​of each dimension of the current data vector output after inputting the current data vector of the current sampling period into the theoretical expected value prediction function. It is a benchmark value for judging whether the current data vector conforms to the operating condition trend and completely corresponds to the dimensions of the current data vector. Fitting deviation refers to a numerical index that quantifies the degree of deviation between the current data vector and the theoretical expected value. The larger the fitting deviation, the more the current data vector deviates from the normal operating condition trend. The fitting deviation can be determined by using absolute deviation (…). ) or relative deviation ( The fitting deviation threshold refers to the critical deviation value pre-set between historical operating data of the fuel assembly throughout its entire life cycle and safe operating standards. It serves as the basis for distinguishing between normal and abnormal data trends. The fitting deviation threshold can be dynamically adjusted according to the operating stage of the fuel assembly, or it can be manually defined.

[0056] In one implementation, a context modeling subunit normalizes the current data vector and the previous data vector to obtain normalized current and previous data vectors. Further, the differences between the normalized current and previous data vectors in each dimension are determined, and this difference vector is defined as the data trend vector. This difference vector can be used to intuitively represent the direction and magnitude of changes in the characteristics of each operating condition within adjacent sampling periods. Further, the data trend vector is input to a fitting deviation determination subunit, which uses linear or polynomial fitting to fit the data trend vector, constructing a theoretical expected value prediction function with the data trend vector as the independent variable and the theoretical expected values ​​of each dimension of the current data vector as the dependent variable. Further, the current data vector can be input into the theoretical expected value prediction function, outputting the theoretical expected values ​​corresponding to each dimension of the data in the current sampling period. Further, a deviation determination model can be used to determine the deviation between the current data vector and the theoretical expected value, obtaining the fitting deviation corresponding to the current sampling period. Furthermore, the deviation determination subunit calls a preset fitting deviation threshold and compares the fitting deviation with the fitting deviation threshold dimension by dimension. If the fitting deviation of all dimensions is less than or equal to its corresponding fitting deviation threshold, it is determined that no abnormal data trend has been detected. At this time, the current ciphertext vector generated by the data encryption unit and the target checksum generated by the checksum generation unit are integrated into the encrypted data of the current sampling period. Finally, the encrypted data is bound with the timestamp, period number, and other associated information of the current sampling period and uploaded to the associated blockchain platform to complete trusted storage. Furthermore, if the fitting deviation of any dimension is greater than its corresponding fitting deviation threshold, it can be determined that an abnormal data trend has been detected. At this time, a blocking command is sent to the write control interface of the data encryption module 103 to suspend the writing process of the current encrypted data to the blockchain platform to prevent potentially abnormal data from entering the chain.

[0057] In this embodiment, the deviation determination subunit can also record information such as abnormal data trend events, the compared fitting deviation, the sensor identifier and sampling time corresponding to the sampled data with abnormal fitting deviation to the abnormal log table of the data management subsystem 10 for subsequent analysis and traceability. Through the above mechanism, the deviation determination subunit can realize the perception and precise blocking of dynamic trend deviations in sampled data, effectively preventing abnormal data from entering the trusted encrypted storage channel, thereby improving the stability, security and anti-attack capability of the nuclear power plant fuel assembly full-cycle data management system.

[0058] In this embodiment, when storing encrypted data and the current sampling period together on a blockchain platform associated with the fuel component full-lifecycle data management system, a smart contract can be used to associate the encrypted data and the current sampling period on the blockchain platform. Optionally, the smart contract may include the following steps: verifying the legality of the target verification code; recording the write time of the encrypted data, the identifier of the data management subsystem 10 corresponding to the encrypted data, and the current operation type; and generating a traceable operation record chain to ensure the immutability and privacy protection of the encrypted data during transmission and storage.

[0059] The smart contract parses the target verification code and verifies whether each verification segment matches the feature entropy, timestamp, and trust level, ensuring that the data has not been tampered with or forged. Through a predefined data structure, it automatically records the write time of the encrypted data, the identifier of the data management subsystem 10, and the operation type (such as write, update, rollback, etc.), generating metadata. This metadata is appended to the blockchain's evidence record as part of a chain structure, allowing subsequent operations on the encrypted data to be fully traceable and auditable. This mechanism introduces an immutable verification and recording process during the encrypted data writing process, significantly improving the trustworthiness, integrity, and security of the fuel component's full-lifecycle data management system in terms of data storage and tracking.

[0060] In this embodiment, the data management subsystem 10 can also perform reliable rollback of abnormal operating data to ensure the operational stability and data reliability of the fuel assembly throughout its entire life cycle.

[0061] Optionally, the data management subsystem 10 further includes a data self-recovery module, wherein the data self-recovery module is used to, when it is detected that the running data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal running data, call the target running snapshot corresponding to the sampling period from the pre-generated stable running snapshot for the multiple abnormal sampling periods, replace the running data corresponding to the sampling period according to the target running snapshot, and synchronously update the replaced running data to the running condition prediction module 102 and the data encryption module 103.

[0062] The data self-recovery module, within the data management subsystem 10, is a core functional module responsible for abnormal data repair and system data link self-healing. It serves as a safety net to ensure the continuity and reliability of system data. Integrated within the data management subsystem 10, the data self-recovery module enables reliable rollback of critical data when abnormalities or state deviations occur in the operational data collected by the data acquisition module 101, ensuring the operational stability and data reliability of the nuclear power plant's fuel assemblies throughout their entire lifecycle. The sliding window refers to a preset time-series data detection interval, a fixed-length time window that dynamically shifts with the sampling cycle (e.g., a window length of 5 sampling cycles). The sliding window contains operational data from multiple consecutive sampling cycles. The data self-recovery module can batch-judge abnormal states within the sliding window to avoid accidental interference from a single sampling triggering erroneous repairs. Multiple consecutive sampling cycles refer to a continuous time-cycle sequence within the sliding window that meets the abnormality judgment conditions; these are the threshold conditions for triggering the data self-recovery process (e.g., operational data from 3 consecutive sampling cycles are all abnormal). The threshold condition is set to distinguish between "transient sensor interference" and "persistent data anomalies," and only performs repair operations on persistent anomalies. Abnormal operating data refers to operating data that, after verification by the data self-recovery module, is determined to deviate from the normal operating trend or exhibit acquisition failure. Abnormal operating data can include invalid data caused by sensor failure, erroneous data caused by transmission interference, and abnormal data caused by non-operating condition mutations. A stable operating snapshot refers to a standardized set of operating data pre-generated and stored according to the sampling period when the fuel assembly is in a stable operating state. A stable operating snapshot can contain operating data and environmental data for the corresponding sampling period, and has been verified to be anomaly-free, serving as a baseline template for replacing abnormal operating data. The generation frequency of stable operating snapshots can be dynamically adjusted according to operating condition stability (e.g., one stable operating snapshot is generated every 10 sampling periods under stable operating conditions). A target operating snapshot refers to snapshot data retrieved from the stable operating snapshot library that matches the operating condition characteristics of the abnormal sampling period. Matching logic can be determined based on similarity of operating conditions (such as consistent power levels and environmental parameters) or temporal correlation (such as stable snapshots of historical periods) to ensure that the replaced data conforms to the operating patterns of the fuel assembly. Operational data replacement can refer to the core operation of the data self-recovery module, which uses standardized data from the target operational snapshot to cover the abnormal operational data corresponding to the abnormal sampling period within the sliding window, generating repaired data that conforms to normal operating conditions. Replaced operational data can refer to compliant operational data obtained after replacing the target operational snapshot. This data possesses characteristics consistent with normal operating conditions and can effectively replace abnormal operational data in subsequent processes.Synchronous updates can refer to the coordinated operation of the data self-recovery module pushing the replaced running data to the running condition prediction module 102 and the data encryption module 103 simultaneously, so as to ensure that the running condition prediction module 102 and the data encryption module 103 use consistent and compliant data, and avoid model training deviations or encryption storage errors due to data inconsistency.

[0063] In this embodiment, during the process of collecting operating data of the fuel assembly and environmental data of the environment in which the fuel assembly is located according to the sampling period by the data acquisition module 101, the collected operating data and environmental data can be input to the data self-recovery module. Furthermore, if the data self-recovery module detects that the operating data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal operating data, for the multiple abnormal sampling periods, a target operating snapshot corresponding to the sampling period can be retrieved from a pre-generated stable operating snapshot, and the operating data corresponding to the sampling period can be replaced according to the target operating snapshot. The replaced operating data is then synchronously updated to the operating condition prediction module 102 and the data encryption module 103.

[0064] Optionally, the data self-recovery module includes: a snapshot generation unit, an anomaly detection unit, and a rollback triggering unit. The snapshot generation unit, when detecting that the current sampling period is a stable operating period, extracts features from the current operating data and current environment data corresponding to the current sampling period to obtain a state feature vector. Based on the state feature vector and the timestamp corresponding to the current sampling period, it generates a stable operating snapshot and stores it in association with its corresponding operating condition stage. The anomaly detection unit, when detecting that the operating data corresponding to multiple consecutive sampling periods within a sliding window are all abnormal operating data, generates an anomaly determination signal and sends the anomaly determination signal to the rollback triggering unit. The rollback triggering unit, upon receiving the anomaly determination signal, for multiple sampling periods with anomalies, calls the target operating snapshot corresponding to the sampling period from the pre-generated stable operating snapshots according to a preset calling rule, replaces the operating data corresponding to the sampling period according to the target operating snapshot, and synchronously updates the replaced operating data to the operating condition prediction module 102 and the data encryption module 103.

[0065] The snapshot generation unit can refer to the functional subunit within the data self-recovery module responsible for generating and storing stable operation snapshots. A stable operation cycle refers to the sampling period during which the fuel assembly's operating state meets a preset safety threshold, and the rate of change and fitting deviation of both the current operating data and the current environmental data are within the normal range; this is a prerequisite for generating a stable operation snapshot. The state feature vector refers to the structured multidimensional data set obtained after feature extraction from the current operating data and current environmental data within the stable operation cycle; it is the core data carrier of the stable operation snapshot and can accurately characterize the operating conditions of that sampling period. The anomaly detection unit can refer to the functional subunit within the data self-recovery module responsible for determining persistent abnormal data. The anomaly detection unit can continuously detect the operating data within a window based on a sliding window mechanism. If it detects that the operating data for multiple consecutive sampling periods within the sliding window are all abnormal operating data, it generates an anomaly determination signal and sends it to the rollback trigger unit, triggering the data rollback process. The anomaly determination signal can refer to the trigger instruction generated by the anomaly detection unit when it determines that persistent data anomalies exist; it is used to notify the rollback trigger unit to initiate the data rollback repair process. The rollback trigger unit can refer to the functional subunit in the data self-recovery module responsible for replacing and synchronizing abnormal data. The preset calling rules can refer to pre-configured snapshot retrieval criteria, which may include ensuring that the target running snapshot matches the fuel component's operating status during the abnormal sampling period, and that the snapshot is closest to the abnormal sampling period in time, thus guaranteeing that the replaced data conforms to actual operating conditions.

[0066] In one implementation, the snapshot generation unit in the data self-recovery module can determine whether the current sampling period is a stable operating period. If it is determined to be a stable operating period, feature extraction is performed on the current operating data and current environmental data of that sampling period to generate a state feature vector characterizing the operating state of the fuel assembly. Further, this state feature vector is bound to the timestamp of the current sampling period to generate a standardized stable operating snapshot, and the stable operating snapshot is associated with its corresponding fuel assembly operating state and stored in a local snapshot library. Further, the anomaly detection unit continuously monitors the operating data within the sliding window based on a sliding window mechanism to determine the rate of change, offset magnitude, and / or fitting deviation between the operating data of multiple consecutive sampling periods. Further, if the determined rate of change, offset magnitude, and / or fitting deviation are greater than their corresponding thresholds, the operating data corresponding to multiple consecutive sampling periods within the sliding window are determined to be abnormal operating data. Further, if it is detected that the operating data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal operating data, an anomaly determination signal is immediately generated and sent to the rollback trigger unit. Furthermore, after receiving the anomaly determination signal, the rollback trigger unit retrieves the target running snapshot that matches the abnormal sampling period from the local snapshot library according to the preset calling rules (running status matching, time proximity priority) for multiple sampling periods with anomalies within the sliding window. Then, it uses the standard data of the target running snapshot to perform a replacement operation on the abnormal running data corresponding to the abnormal sampling period. Finally, it synchronously updates the replaced compliant running data to the running condition prediction module 102 and the data encryption module 103 to ensure that the data used by downstream modules has consistency and reliability.

[0067] Optionally, the data management subsystem 10 further includes a disturbance sensitivity assessment module, which is used to determine the disturbance sensitivity corresponding to the current sampling period based on the collected current operating data, current environmental data, and historical data trend vector of the previous stable period of the current sampling period, and to perform a disturbance recovery operation when the disturbance sensitivity exceeds the preset disturbance threshold corresponding to the current sampling period.

[0068] The disturbance sensitivity assessment module refers to the functional module in the data management subsystem 10 responsible for quantifying the anti-interference capability of the fuel assembly's operating status. The core task of this module is to combine real-time data from the current sampling period with trend data from historical stable periods to calculate the disturbance sensitivity of the current sampling period, determine whether the current sampling period has deviated from the stable range due to external or internal disturbances, and perform recovery operations when the risk threshold is exceeded, further improving the system's operational stability and anti-interference capability. The previous stable period refers to a historical sampling period immediately adjacent to the current sampling period that has been determined to be a stable operating period. The operating data and environmental data corresponding to the previous stable period are both within the normal fluctuation range, with no abnormal deviations. The historical data trend vector refers to a short-cycle trend vector calculated based on the operating data vector of the previous stable period and the data vector of the previous sampling period preceding that stable period. The calculation method of the historical data trend vector can be consistent with the logic of the context modeling subunit in generating the data trend vector. The historical data trend vector can characterize the normal change trend of the operating conditions within the previous stable period and serves as a benchmark trend reference for determining whether the current sampling period is affected by disturbances. Disturbance sensitivity is a numerical indicator that quantifies the degree to which the current sampling period is affected by internal and external disturbances. The calculation logic for disturbance sensitivity can be as follows: compare the deviation of the current data vector from the historical data trend vector of the previous stable period, and weight it using parameters such as the fluctuation amplitude of the current data vector and the fitting residual. A higher disturbance sensitivity value indicates a greater impact of disturbances on the current sampling period and a greater likelihood of deviation from the stable operating range. The preset disturbance threshold can be a critical value for disturbance sensitivity pre-set based on different operating conditions of the fuel assembly. The preset disturbance threshold can be set by combining historical stable operating condition data and safe operating standards, or it can be manually defined. The preset disturbance threshold can serve as a basis for distinguishing between "disturbance-resistant stability" and "disturbance risk." Disturbance recovery operation refers to the automatic operation performed to maintain operational stability when the disturbance sensitivity exceeds the preset disturbance threshold. Optionally, the disturbance recovery operation includes at least one of the following: calling the data self-recovery module to perform a stable operation snapshot rollback; adjusting the model parameters of the operation condition prediction model in the operation condition prediction module 102; and updating the encryption level and / or check code generation rules of the data encryption module 103.

[0069] In one implementation, the disturbance sensitivity assessment module retrieves the current operating data and current environmental data collected by the data acquisition module 101 for the current sampling period, and integrates them to generate a current data vector. Simultaneously, it extracts the historical data trend vector corresponding to the previous stable period from local storage, using it as a benchmark for stable operating conditions. Further, it calculates indicators such as the deviation between the current data vector and the historical data trend vector, and the fluctuation coefficients of each dimension of the current data, and uses a weighted algorithm to quantify the disturbance sensitivity corresponding to the current sampling period. Finally, it compares this disturbance sensitivity with a preset disturbance threshold corresponding to the operating condition stage of the current sampling period. If the disturbance sensitivity exceeds the preset disturbance threshold, a disturbance recovery operation is immediately performed to ensure that the fuel assembly's operating state returns to a stable range.

[0070] For example, the disturbance sensitivity can be determined using the following formula: in, This indicates the perturbation sensitivity of the current sampling period; This indicates the total number of sensors included in the data acquisition module 101; Indicates the first The weight of each sensor in the disturbance assessment; This indicates that the current data vector and the historical data trend vector are at the 1st... The deviation amplitude corresponding to each sensor; Indicates the first The average value of each sensor over a stable period; It is a constant and can be set to... To avoid the constant introduced by a denominator of zero; This indicates the weighting of the adjustment trend term in the overall sensitivity. Indicates the first The rate of change of the standardization trend of individual sensors; Indicates multiplication.

[0071] Assume the current data vector corresponding to the current sampling period is: The historical data trend vector corresponding to the previous stable period is: First, calculate the perturbation amplitude in each dimension (sensor): Furthermore, calculate the... The rate of change of the standardization trend of each sensor.

[0072] No. The rate of change of the standardized trend of an individual sensor can be determined by the following formula: in, Indicates the first The instantaneous rate of change of each sensor; The time interval representing the sampling period; Indicates the first A sensor in the past The variance of sampled values ​​within a sampling period.

[0073] It should be noted that the first The weight of each sensor in the disturbance assessment can be assigned an initial weight to each channel by domain experts based on historical experience, according to the importance of the variables monitored by the sensor. For example, if a certain type of temperature sensor has a decisive impact on the thermal stability of the fuel assembly, its corresponding weight can be appropriately increased. The weight ratio of the adjustment trend term in the overall sensitivity is set to 0.2, as this setting in actual calculations shows that a disturbance sensitivity index curve is stable and the error rate is low.

[0074] The nuclear power plant fuel assembly full-lifecycle data management system provided by this invention adopts a distributed architecture and is composed of multiple data management subsystems working together. Each data management subsystem integrates a data acquisition module, an operating condition prediction module, a data self-recovery module, and a data encryption module. Through the coordinated linkage of these modules, the system can monitor the status of nuclear fuel assemblies throughout their entire lifecycle, from manufacturing, transportation, loading, in-core operation, unloading to storage, prevent anomalies, and ensure reliable data storage, effectively guaranteeing the operational safety of fuel assemblies and the reliability of data management.

[0075] Implementation principle: During the operation phase of the fuel assembly full life cycle data management system, the data acquisition modules of each data management subsystem first perform real-time sensing and acquisition of the operating status of the fuel assembly throughout its entire life cycle. The acquired sampling data includes, but is not limited to, the current operating data of the fuel assembly body (temperature data, internal pressure data, vibration amplitude data) and the current environmental data (such as coolant temperature and environmental radiation dose). The collected multi-source heterogeneous data is transmitted as input data to the operating condition prediction module, which performs local learning and decision optimization processing: First, the feature extraction unit performs standardization and normalization preprocessing on the input data to eliminate dimensional differences and extract real-time state feature vectors representing the operating state of the fuel assembly; Second, the offline training unit trains a pre-built deep learning model offline based on pre-stored historical operating data and historical environmental data to obtain an offline-trained operating condition prediction model; Third, the online training unit trains the offline-trained operating condition prediction model based on the operating data corresponding to the current sampling period, the environmental data, and the previous operating state and data of the previous sampling period to obtain a trained operating condition prediction model; Fourth, the strategy update unit optimizes the local decision logic of the data management subsystem based on the model tuning results, enabling the data management subsystem to adapt to changes in the operating state of the fuel assembly at different life cycle stages and under different operating conditions.

[0076] Furthermore, when the system detects abnormal operating data or deviations in the operating status of the fuel assembly, the data self-recovery module performs anomaly diagnosis and data repair: the snapshot generation unit determines in real time whether the current sampling period is a stable operating period. If it is determined to be a stable operating period, it performs feature extraction operations on the current operating data and current environmental data of that period to generate a state feature vector that can characterize the operating status of the fuel assembly. Then, it binds the state feature vector with the timestamp of the current sampling period to generate a standardized stable operating snapshot, and stores the stable operating snapshot and its corresponding fuel assembly operating status in the local snapshot library; the anomaly detection unit continuously monitors the real-time operating data based on a sliding window mechanism. If the system detects that the running data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal running data, it immediately generates an anomaly judgment signal and sends it to the rollback trigger unit. After receiving the anomaly judgment signal, the rollback trigger unit retrieves the target running snapshot that matches the abnormal sampling period from the local snapshot library according to the preset calling rules (running status matching, time proximity priority) for the multiple abnormal sampling periods within the sliding window. Then, it performs a replacement operation on the abnormal running data corresponding to the abnormal sampling period using the standard data of the target running snapshot. Finally, it synchronously updates the replaced compliant running data to the running condition prediction module and the data encryption module to ensure that the data used by downstream modules has consistency and reliability.

[0077] Furthermore, to ensure the security and integrity of data transmission and storage throughout the entire lifecycle, the raw data collected by the data acquisition module (current operating data, current environmental data) must be encrypted by the data encryption module: First, the data encryption unit uses a preset homomorphic encryption algorithm (such as the CKKS algorithm) to encrypt the current data vector formed by integrating the raw data, generating a current ciphertext vector to ensure that the data can directly participate in subsequent working condition analysis and calculation in a encrypted state, balancing privacy protection and data application value; Second, the check code generation unit generates a multi-segment check code containing a data feature check segment, a periodic identifier segment, and a hash digest segment based on the Shannon entropy of the sensor-collected data, the data acquisition timestamp, and the trust level of the distributed intelligent agent, for data integrity verification and security traceability; Third, the context awareness unit constructs a short-period trend vector (data trend vector) by combining the preprocessed data of the current sampling period (current data vector) with the preprocessed data of the previous sampling period (previous data vector), and performs linear or nonlinear fitting calculations on the trend vector. If the fitting residual exceeds a preset fitting deviation threshold, the check code update process is triggered or the current data writing link is blocked to prevent abnormal data from entering the subsequent evidence storage stage.

[0078] Furthermore, after the data encryption module completes its processing, the current ciphertext vector and multi-segment checksum are written into the decentralized consortium blockchain platform via a pre-defined smart contract. During the execution of the smart contract, the validity of the multi-segment checksum is automatically verified, timestamps are written, distributed intelligent agent identification is added, and operation type (such as data collection or data repair) is registered. This generates an immutable, full-lifecycle data operation traceability chain, ensuring the immutability and privacy protection of the entire process of data transmission and storage related to fuel components.

[0079] This system achieves refined monitoring of the entire life cycle operation status of nuclear power plant fuel assemblies, real-time anomaly prevention and control, and highly reliable data management through a closed-loop collaborative mechanism of "sensing and acquisition - learning and optimization - self-healing and repair - encrypted storage". It significantly improves the operational safety of fuel assemblies and the controllability of nuclear power plant operations, and provides reliable technical support for the entire life cycle management of fuel assemblies.

[0080] This disclosure provides a fuel assembly full-cycle data management system, including multiple data management subsystems. Each data management subsystem includes: a data acquisition module, an operating condition prediction module, and a data encryption module. The data acquisition module is used to acquire current operating data of the fuel assembly and current environmental data of the environment in which the fuel assembly is located within the current sampling period. The operating condition prediction module is used to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period based on the acquired current operating data, current environmental data, and a pre-trained operating condition prediction model. It also adjusts the data management strategy of the data management subsystem based on the current operating state and the predicted operating data. The data encryption module is used to... The collected current operating data and current environmental data are encrypted to obtain encrypted data corresponding to the current collection cycle. This encrypted data and the current sampling cycle are then associated and stored on a blockchain platform linked to the fuel assembly full-lifecycle data management system. This solves the problems of poor timeliness in fuel assembly condition detection in related technologies, which prevents timely capture of sudden changes in operating conditions during fuel assembly operation, resulting in low accuracy and efficiency in fuel assembly condition detection. It enables real-time and accurate monitoring of the operating status of nuclear power plant fuel assemblies throughout their entire lifecycle and rapid capture of sudden changes in operating conditions, improving the timeliness and accuracy of condition detection. At the same time, relying on data encryption and evidence storage capabilities, it ensures the continuity, credibility, and security of data, ultimately significantly enhancing the operational safety of fuel assemblies and the controllability of nuclear power plant operations.

[0081] Figure 2 This is a flowchart illustrating the fuel assembly lifecycle data management method provided in this embodiment. This method can be applied to any data management subsystem within the fuel assembly lifecycle data management system provided in the above embodiments. Figure 2 As shown, the method in this embodiment may specifically include: S210. Collect the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located within the current sampling period through the data acquisition module.

[0082] S220. The operating condition prediction module obtains the current operating status of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period based on the collected current operating data, current environmental data, and the pre-trained operating condition prediction model. The data management strategy of the data management subsystem is then adjusted based on the current operating status and the predicted operating data.

[0083] S230. The collected current running data and current environment data are encrypted through the data encryption module to obtain encrypted data corresponding to the current collection period, and the encrypted data and the current sampling period are associated and stored in the blockchain platform.

[0084] The technical solution of this disclosure includes: acquiring current operating data of the fuel assembly and current environmental data of the environment in which the fuel assembly is located within the current sampling period through a data acquisition module; further, obtaining the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period through an operating condition prediction module based on the acquired current operating data, current environmental data, and a pre-trained operating condition prediction model, and adjusting the data management strategy of the data management subsystem based on the current operating state and predicted operating data; and further, encrypting the acquired current operating data and current environmental data through a data encryption module. This method obtains encrypted data corresponding to the current sampling period and stores the encrypted data and the current sampling period on the blockchain platform. This solves the problems of poor timeliness in fuel assembly condition detection in related technologies, which makes it impossible to capture sudden changes in the operating conditions of fuel assemblies in a timely manner, resulting in low accuracy and efficiency in fuel assembly condition detection. It enables real-time and accurate monitoring of the operating status of nuclear power plant fuel assemblies throughout their entire life cycle and rapid capture of sudden changes in operating conditions, improving the timeliness and accuracy of condition detection. At the same time, relying on data encryption and evidence storage capabilities, it ensures the continuity, credibility and security of data, ultimately significantly enhancing the operational safety of fuel assemblies and the controllability of nuclear power plant operations.

[0085] Optionally, the data management subsystem also includes: a data self-recovery module; when the data self-recovery module detects that the running data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal running data, for the multiple sampling periods with abnormalities, it calls the target running snapshot corresponding to the sampling period from the pre-generated stable running snapshot, replaces the running data corresponding to the sampling period according to the target running snapshot, and synchronously updates the replaced running data to the running condition prediction module and the data encryption module.

[0086] Optionally, the data self-recovery module includes: a snapshot generation unit, an anomaly detection unit, and a rollback triggering unit. The snapshot generation unit, upon detecting that the current sampling period is a stable operating period, extracts features from the current operating data and current environmental data corresponding to the current sampling period to obtain a state feature vector. Based on the state feature vector and the timestamp corresponding to the current sampling period, a stable operating snapshot is generated, and the stable operating snapshot is associated with and stored with its corresponding operating state. The anomaly detection unit, upon detecting that the operating data corresponding to multiple consecutive sampling periods within a sliding window are all abnormal operating data, generates an anomaly determination signal and sends the anomaly determination signal to the rollback triggering unit. The rollback triggering unit, upon receiving the anomaly determination signal, for multiple sampling periods with anomalies, calls the target operating snapshot corresponding to the sampling period from the pre-generated stable operating snapshots according to preset calling rules, replaces the operating data corresponding to the sampling period based on the target operating snapshot, and synchronously updates the replaced operating data to the operating condition prediction module and the data encryption module.

[0087] Optionally, the operating condition prediction module includes a feature extraction unit and a strategy update unit. The feature extraction unit preprocesses the collected current operating data and current environmental data, and extracts features from the preprocessed current operating data and current environmental data to obtain a state feature vector corresponding to the current sampling period. The strategy update unit inputs the state feature vector corresponding to the current sampling period into a pre-trained operating condition prediction model to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period. Based on the current operating state and the predicted operating data, the data management strategy of the data management subsystem is adjusted.

[0088] Optionally, the operating condition prediction module further includes: an offline training unit and an online training unit; the offline training unit trains a pre-built deep learning model offline based on pre-stored historical operating data and historical environmental data to obtain an offline-trained operating condition prediction model; the online training unit trains the offline-trained operating condition prediction model based on the operating data and environmental data corresponding to the current sampling period and the previous operating state and data of the previous sampling period to obtain a trained operating condition prediction model.

[0089] Optionally, the data encryption module includes: a data encryption unit, a checksum generation unit, and a context-aware unit. The data encryption unit generates a current data vector corresponding to the current sampling period based on the current running data and current environment data, and encrypts the current data vector to obtain a current ciphertext vector corresponding to the current sampling period. The checksum generation unit generates a target checksum based on the current running data, current environment data, and the current sampling period. During the generation of the current ciphertext vector and the target checksum, the context-aware unit performs data trend anomaly detection based on the current data vector and the previous data vector from the previous sampling period. If no data trend anomaly is detected, the current ciphertext vector and the target checksum are used as encrypted data corresponding to the current sampling period, and the encrypted data and the current sampling period are associated and stored on a blockchain platform associated with the fuel component full-lifecycle data management system.

[0090] Optionally, the context-aware unit includes: a context modeling subunit, a fitting deviation determination subunit, and a deviation judgment subunit. The context modeling subunit determines the data trend vector based on the current data vector and the previous data vector. The fitting deviation determination subunit fits the data trend vector to construct a theoretical expected value prediction function, inputs the current data vector into the theoretical expected value prediction function to obtain the theoretical expected value corresponding to the current sampling period, and determines the fitting deviation corresponding to the current sampling period based on the current data vector and the theoretical expected value. The deviation judgment subunit compares the fitting deviation with a preset fitting deviation threshold. If the fitting deviation is less than or equal to the fitting deviation threshold, it determines that no data trend anomaly has been detected, uses the current ciphertext vector and the target checksum as encrypted data corresponding to the current sampling period, and stores the encrypted data and the current sampling period together on the blockchain platform.

[0091] Optionally, the sampling deviation determination subunit writes the encrypted data and the current sampling period into the blockchain platform via a smart contract.

[0092] Optionally, the data management subsystem further includes: a disturbance sensitivity assessment module; this module determines the disturbance sensitivity corresponding to the current sampling period based on the collected current operating data, current environmental data, and the historical data trend vector of the previous stable period of the current sampling period, and performs a disturbance recovery operation if the disturbance sensitivity exceeds a preset disturbance threshold corresponding to the current sampling period; wherein the disturbance recovery operation includes at least one of the following: calling the data self-recovery module to roll back the stable operation snapshot; adjusting the model parameters of the operating condition prediction model in the operating condition prediction module; updating the encryption level and / or verification code generation rules of the data encryption module.

[0093] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0094] like Figure 3 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0095] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0096] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the fuel component full-cycle data management method.

[0097] In some embodiments, the fuel assembly lifecycle data management method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via read-only memory (ROM) 12 and / or communication unit 19. When the computer program is loaded into random access memory (RAM) 13 and executed by processor 11, one or more steps of the fuel assembly lifecycle data management method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the fuel assembly lifecycle data management method by any other suitable means (e.g., by means of firmware).

[0098] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0099] Computer programs used to implement the fuel assembly lifecycle data management method of this disclosure can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0100] This disclosure provides a computer-readable storage medium storing computer instructions for instructing a processor to execute a fuel assembly lifecycle data management method, comprising: acquiring current operating data of the fuel assembly and current environmental data of the environment in which the fuel assembly is located within the current sampling period through a data acquisition module; obtaining the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period through an operating condition prediction module based on the acquired current operating data, current environmental data, and a pre-trained operating condition prediction model, and adjusting the data management strategy of the data management subsystem based on the current operating state and the predicted operating data; encrypting the acquired current operating data and current environmental data through a data encryption module to obtain encrypted data corresponding to the current sampling period, and storing the encrypted data and the current sampling period together in a blockchain platform associated with the fuel assembly lifecycle data management system.

[0101] In the context of this disclosure, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0102] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0103] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0104] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0105] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from ROM 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of embodiments of this disclosure.

[0106] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements a fuel component full-cycle data management method according to any embodiment of this disclosure.

[0107] In implementing a computer program product, computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0108] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this disclosure can be achieved, and this is not limited herein.

[0109] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A fuel assembly full-lifecycle data management system, characterized in that, It includes multiple data management subsystems, each of which includes: a data acquisition module, an operational condition prediction module, and a data encryption module. The data acquisition module is used to acquire the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located during the current sampling period. The operating condition prediction module is used to obtain the current operating state of the fuel assembly in the current sampling period and the predicted operating data of the next sampling period based on the collected current operating data, the current environmental data and the pre-trained operating condition prediction model, and to adjust the data management strategy of the data management subsystem based on the current operating state and the predicted operating data. The data encryption module is used to encrypt the collected current running data and the current environment data to obtain encrypted data corresponding to the current collection period, and to store the encrypted data and the current sampling period together in the blockchain platform.

2. The fuel assembly full-lifecycle data management system according to claim 1, characterized in that, The data management subsystem further includes: a data self-recovery module, wherein, The data self-recovery module is used to, when it is detected that the running data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal running data, call the target running snapshot corresponding to the sampling period from the pre-generated stable running snapshot for the multiple abnormal sampling periods, replace the running data corresponding to the sampling period according to the target running snapshot, and synchronously update the replaced running data to the running condition prediction module and the data encryption module.

3. The fuel assembly full-lifecycle data management system according to claim 2, characterized in that, The data self-recovery module includes: a snapshot generation unit, an anomaly detection unit, and a rollback triggering unit, wherein, The snapshot generation unit is used to extract features from the current running data and current environment data corresponding to the current sampling period when the current sampling period is detected to be a stable running period, to obtain a state feature vector, to generate a stable running snapshot based on the state feature vector and the timestamp corresponding to the current sampling period, and to associate and store the stable running snapshot with its corresponding running state. The anomaly detection unit is used to generate an anomaly determination signal when it detects that the running data corresponding to multiple consecutive sampling periods within the sliding window are all abnormal running data, and to send the anomaly determination signal to the rollback trigger unit. The rollback trigger unit is used to, upon receiving the anomaly determination signal, call the target running snapshot corresponding to the sampling period from the pre-generated stable running snapshot according to the preset calling rules for the multiple sampling periods with anomalies, replace the running data corresponding to the sampling period according to the target running snapshot, and synchronously update the replaced running data to the running condition prediction module and the data encryption module.

4. The fuel assembly full-lifecycle data management system according to claim 1, characterized in that, The operating condition prediction module includes: a feature extraction unit and a policy update unit; wherein... The feature extraction unit is used to preprocess the collected current running data and current environment data, and to extract features from the preprocessed current running data and current environment data to obtain a state feature vector corresponding to the current sampling period. The strategy update unit is used to input the state feature vector corresponding to the current sampling period into the pre-trained operating condition prediction model to obtain the current operating state of the fuel component in the current sampling period and the predicted operating data of the next sampling period of the current sampling period, and to adjust the data management strategy of the data management subsystem based on the current operating state and the predicted operating data.

5. The fuel assembly full-lifecycle data management system according to claim 4, characterized in that, The operating condition prediction module further includes: an offline training unit and an online training unit, wherein... The offline training unit is used to train a pre-built deep learning model offline based on pre-stored historical operating data and historical environment data, so as to obtain an offline-trained operating condition prediction model. The online training unit is used to train the offline-trained operating condition prediction model based on the operating data corresponding to the current sampling period, the environmental data, the previous operating state and the previous operating data of the previous sampling period, so as to obtain the trained operating condition prediction model.

6. The fuel assembly full-lifecycle data management system according to claim 1, characterized in that, The data encryption module includes: a data encryption unit, a checksum generation unit, and a context-aware unit, wherein, The data encryption unit is used to generate a current data vector corresponding to the current sampling period based on the current running data and current environment data corresponding to the current sampling period, and to encrypt the current data vector to obtain a current ciphertext vector corresponding to the current sampling period. The check code generation unit is used to generate a target check code based on the current running data, the current environment data, and the current sampling period; The context-aware unit is used to perform data trend anomaly detection based on the current data vector and the previous data vector of the previous sampling period during the generation of the current ciphertext vector and the target check code. If no data trend anomaly is detected, the current ciphertext vector and the target check code are used as encrypted data corresponding to the current sampling period, and the encrypted data and the current sampling period are associated and stored in the blockchain platform.

7. The fuel assembly full-lifecycle data management system according to claim 6, characterized in that, The context-aware unit includes: a context modeling subunit, a fitting deviation determination subunit, and a deviation judgment subunit; wherein... The context modeling subunit is used to determine the data trend vector based on the current data vector and the previous data vector; The fitting deviation determination subunit is used to fit the data trend vector, construct a theoretical expected value prediction function, input the current data vector into the theoretical expected value prediction function to obtain the theoretical expected value corresponding to the current sampling period, and determine the fitting deviation corresponding to the current sampling period based on the current data vector and the theoretical expected value. The deviation determination subunit is used to compare the fitting deviation with a preset fitting deviation threshold. If the fitting deviation is less than or equal to the fitting deviation threshold, it is determined that no abnormal data trend has been detected. The current ciphertext vector and the target check code are used as encrypted data corresponding to the current sampling period, and the encrypted data and the current sampling period are associated and stored in the blockchain platform.

8. The fuel assembly full-lifecycle data management system according to claim 7, characterized in that, The deviation determination subunit is specifically used to write the encrypted data and the current sampling period into the blockchain platform through a smart contract.

9. The fuel assembly full-lifecycle data management system according to claim 1, characterized in that, The data management subsystem further includes a disturbance sensitivity assessment module, wherein... The disturbance sensitivity assessment module is used to determine the disturbance sensitivity corresponding to the current sampling period based on the collected current operating data, the current environmental data, and the historical data trend vector of the previous stable period of the current sampling period. If the disturbance sensitivity exceeds a preset disturbance threshold corresponding to the current sampling period, a disturbance recovery operation is performed. The disturbance recovery operation includes at least one of the following: calling the data self-recovery module to roll back the stable operating snapshot; adjusting the model parameters of the operating condition prediction model in the operating condition prediction module; and updating the encryption level and / or checksum generation rules of the data encryption module.

10. A method for full-lifecycle data management of fuel assemblies, characterized in that, include: The data acquisition module collects the current operating data of the fuel assembly and the current environmental data of the environment in which the fuel assembly is located within the current sampling period. The operating condition prediction module obtains the current operating state of the fuel assembly in the current sampling period and the predicted operating data for the next sampling period based on the collected current operating data, the current environmental data, and the pre-trained operating condition prediction model. The data management strategy of the data management subsystem is then adjusted based on the current operating state and the predicted operating data. The data encryption module encrypts the collected current running data and current environment data to obtain encrypted data corresponding to the current sampling period, and then stores the encrypted data and the current sampling period together on the blockchain platform.

Images