A method for obtaining, using and device for health prediction model of instrument and control card

By building a functional failure tree of nuclear power instrument card control parts and machine supervision learning algorithm, and using historical data to train a health prediction model, the problem that traditional maintenance methods cannot prevent sudden failures is solved, and the safety and reliability of nuclear power plants are improved.

CN118520377BActive Publication Date: 2024-11-12CHINA TECHENERGY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410980142.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-12
Estimated Expiration
2044-07-22

AI Technical Summary

Technical Problem

The traditional maintenance methods of nuclear power instrument card control relies on regular inspections and post-malfunction repairs, and cannot effectively prevent sudden failures, resulting in passive and inefficient.

Method used

Build a functional failure tree of nuclear power instrument card control, use machine supervised learning algorithm to build an initial health prediction model, extract target feature vectors through historical health correlation information for model training, obtain a trained health prediction model, and input the current feature vector into the model To predict the operating status.

Benefits of technology

Preventive maintenance of nuclear power instrument card control parts has been achieved, the safety and reliability of nuclear power plants have been improved, and the risk of unplanned downtime has been reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118520377B_ABST
    Figure CN118520377B_ABST
Patent Text Reader

Abstract

The present invention provides a method for obtaining, a method for using, and a device for an instrument and control card health prediction model, which are applied to the technical field of nuclear power instrument and control card data processing. The present invention constructs a functional fault tree of nuclear power instrument and control cards, uses the functional fault tree to construct an instrument and control card health prediction model, and uses the target feature vector extracted from the health-related information of the nuclear power instrument and control cards at historical moments and annotated with the operating status result identifier to perform model training. The target feature vector of the nuclear power instrument and control card at the current moment is input into the trained instrument and control card health prediction model to obtain the instrument and control card operating status prediction result output by the model. The present invention constructs a target instrument and control card health prediction model by effectively utilizing the historical health-related information of the nuclear power instrument and control cards. The target instrument and control card health prediction model can predict the future health status of the nuclear power instrument and control cards, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of nuclear power instrumentation and control card component data processing, and in particular to a method for obtaining, a method for using and a device for an instrumentation and control card component health prediction model. Background Art

[0002] The application of digital control system (DCS) in nuclear power plants is becoming more and more popular, and has become a key technology for second-generation and third-generation nuclear power plants. DCS is not only used to monitor the instruments of nuclear power plants, but also directly affects the safe operation of nuclear power plants. As the core component of DCS, the stability and reliability of nuclear power instrumentation and control cards are crucial to the safe operation of the entire nuclear power plant.

[0003] However, the traditional maintenance method of nuclear power instrumentation and control components mainly relies on regular inspection and post-fault repair, which is passive and inefficient and cannot effectively prevent sudden failures.

[0004] Therefore, how to predict the health status of nuclear power instrumentation and control components in advance has become the key to achieving preventive maintenance of nuclear power instrumentation and control components. Summary of the invention

[0005] In view of the above problems, the present invention provides a method for obtaining, using and device of an instrumentation and control card health prediction model to overcome the above problems or at least partially solve the above problems. The technical solution is as follows:

[0006] A method for constructing an instrumentation and control card health prediction model, comprising:

[0007] Construct the functional fault tree of nuclear power instrumentation and control components;

[0008] Using the functional fault tree, constructing an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card;

[0009] Obtaining a key feature information training set of the nuclear power instrumentation and control card component, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrumentation and control card component at at least one historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrumentation and control card component;

[0010] The initial instrumentation and control card component health prediction model is trained using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model.

[0011] Optionally, the construction of a functional fault tree of a nuclear power instrumentation and control card component includes:

[0012] Obtaining a reliability function block diagram of the nuclear power instrumentation and control card component;

[0013] Using the reliability function block diagram to identify key components and key circuits of the nuclear power instrumentation and control card;

[0014] Performing functional development on the key components and the key circuits to obtain hardware functional development results of the nuclear power instrumentation and control card components;

[0015] The hardware function expansion result is used to construct a functional fault tree of the nuclear power instrumentation and control card from the component level to the circuit level.

[0016] Optionally, using the functional fault tree to construct an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card includes:

[0017] The functional fault tree is substituted into a pre-compiled machine supervised learning algorithm program according to the fault category, and the machine supervised learning algorithm program is executed to construct an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card.

[0018] Optionally, the using the target feature vector in the key feature information training set to train the initial instrumentation and control card component health prediction model to obtain a trained target instrumentation and control card component health prediction model includes:

[0019] When the machine supervised learning algorithm is a random forest algorithm, the target feature vector in the key feature information training set is used to train multiple decision trees in the initial instrumentation and control card health prediction model, so that each decision tree is split based on feature importance to obtain a trained target instrumentation and control card health prediction model.

[0020] Optionally, after the initial instrumentation and control card component health prediction model is trained using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model, the method further includes:

[0021] The model parameters of the target instrumentation and control card component health prediction model are tuned using a cross-validation algorithm and a network search algorithm to obtain an optimized target instrumentation and control card component health prediction model.

[0022] Optionally, after the cross-validation algorithm and the network search algorithm are used to tune the model parameters of the target I&C card component health prediction model to obtain the optimized target I&C card component health prediction model, the method further includes:

[0023] Obtaining a key feature information verification set of the nuclear power instrumentation and control card component, wherein the key feature information verification set includes at least one target feature vector extracted from the historical health-related information of the nuclear power instrumentation and control card component, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrumentation and control card component;

[0024] The target feature vector in the key feature information verification set is used to verify the optimized target instrumentation and control card component health prediction model to obtain a prediction accuracy evaluation result of the target instrumentation and control card component health prediction model.

[0025] Optionally, the health-related information includes process data, self-diagnosis status data and operating environment data of the nuclear power instrumentation and control card component.

[0026] A method for using an instrumentation and control card health prediction model includes:

[0027] Obtain the target feature vector of the nuclear power instrumentation and control card at the current moment;

[0028] The target feature vector is input into the target instrumentation and control card component health prediction model obtained by any of the above-mentioned instrumentation and control card component health prediction model construction methods to obtain the operating status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model.

[0029] A device for constructing an instrumentation and control card health prediction model comprises: a functional fault tree construction unit, an initial model construction unit, a training set acquisition unit and a target model acquisition unit.

[0030] The functional fault tree construction unit is used to construct a functional fault tree of a nuclear power instrumentation and control card component;

[0031] The initial model building unit is used to build an initial instrumentation and control card component health prediction model of the nuclear power instrumentation and control card component by using the functional fault tree;

[0032] The training set acquisition unit is used to obtain a key feature information training set of the nuclear power instrumentation and control card component, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrumentation and control card component at at least one historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrumentation and control card component;

[0033] The target model acquisition unit is used to train the initial instrumentation and control card component health prediction model using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model.

[0034] A device for using an instrumentation and control card health prediction model comprises: a current feature vector obtaining unit and an operation status prediction result obtaining unit,

[0035] The current feature vector obtaining unit is used to obtain the target feature vector of the nuclear power instrumentation and control card at the current moment;

[0036] The operating status prediction result obtaining unit is used to input the target feature vector into the target instrumentation and control card component health prediction model obtained by the construction device of the above-mentioned instrumentation and control card component health prediction model, and obtain the operating status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model.

[0037] By means of the above technical scheme, the present invention provides a method for obtaining, a method for using and a device for an instrument and control card health prediction model, which are applied to the technical field of nuclear power instrument and control card data processing. The present invention constructs a functional fault tree of nuclear power instrument and control cards, uses the functional fault tree to construct an instrument and control card health prediction model, and uses the target feature vector extracted from the health-related information of nuclear power instrument and control cards at historical moments and annotated with the operating status result identifier to train the model. The target feature vector of the nuclear power instrument and control card at the current moment is input into the trained instrument and control card health prediction model to obtain the instrument and control card operating status prediction result output by the model. The present invention constructs a target instrument and control card health prediction model by effectively utilizing the historical health-related information of nuclear power instrument and control cards. The target instrument and control card health prediction model can predict the future health status of nuclear power instrument and control cards, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0038] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0040] Figure 1 A flow chart of an implementation method of a method for constructing an instrumentation and control card health prediction model provided by an embodiment of the present invention is shown;

[0041] Figure 2 A schematic diagram illustrating data collection of a key chip provided by an embodiment of the present invention is shown;

[0042] Figure 3 A flowchart of a specific implementation of step S100 in the method for constructing an instrumentation and control card health prediction model provided in an embodiment of the present invention is shown;

[0043] Figure 4 A flow chart showing another implementation of a method for constructing an instrumentation and control card component health prediction model provided in an embodiment of the present invention;

[0044] Figure 5 A flow chart showing another implementation of a method for constructing an instrumentation and control card component health prediction model provided in an embodiment of the present invention;

[0045] Figure 6 A flow chart showing an implementation method of a method for using an instrumentation and control card component health prediction model provided in an embodiment of the present invention;

[0046] Figure 7 A schematic diagram showing the structure of a device for constructing an instrumentation and control card component health prediction model provided by an embodiment of the present invention;

[0047] Figure 8 A schematic diagram showing the structure of a device for using the instrumentation and control card health prediction model provided by an embodiment of the present invention;

[0048] Fig. 9 A schematic structural diagram of an electronic device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0049] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to enable the scope of the present invention to be fully communicated to those skilled in the art.

[0050] Nuclear power instrumentation and control components are the core components to ensure the safe operation of nuclear power plants. Their stability and reliability are crucial to the safe operation of the entire nuclear power plant. However, traditional maintenance methods mainly rely on regular inspections and repairs after failures. This method is not only passive, but also inefficient and difficult to effectively prevent sudden failures.

[0051] In the digital control system (DCS) used in nuclear power plants, nuclear power instrumentation and control cards can realize alarms through self-diagnosis, including collecting diagnostic signals, comparing with set limits, comparing fault codes when exceeding limits, and alarming through human-machine interfaces. However, this type of fault alarm is mostly a post-alarm and maintenance mode, and it is impossible to formulate a maintenance plan in advance. In addition, the diagnostic software is usually integrated with the nuclear power instrumentation and control cards on the same circuit board, and hardware failure or failure may cause the diagnostic software to fail.

[0052] At present, with the development of big data and artificial intelligence technology, predictive maintenance strategies based on machine learning have gradually become a new way to improve the efficiency of nuclear power operation and maintenance. Therefore, developing a method that can use big data and machine learning technology to predict the health status of nuclear power instrumentation and control components in advance and formulate effective maintenance strategies based on this is of great significance to improving the safety and operational efficiency of nuclear power plants.

[0053] Based on this, in an embodiment of the present invention, a method for constructing a health prediction model for an instrument and control card is provided. First, a functional fault tree of a nuclear power instrument and control card is constructed, and an initial health prediction model is created based on this. Subsequently, key features are extracted from the historical data of the nuclear power instrument and control card to form a training set. These key features constitute feature vectors annotated with the operating status identifier of the nuclear power instrument and control card. The health prediction model is trained using these feature vectors, and finally a trained health prediction model that can predict the health status of the instrument and control card is obtained. It can be seen that by effectively utilizing the historical data of the multi-reactor operating experience of the nuclear power instrument and control card, a health prediction model that can predict the future health status of the nuclear power instrument and control card is generated, and then the current data of the nuclear power instrument and control card is used for prediction, which helps to provide early warning and maintenance decision support for potential failures of the nuclear power instrument and control card.

[0054] like Figure 1 As shown, a method for constructing an instrumentation and control card health prediction model provided by an embodiment of the present invention is a flow chart of an implementation method, and the construction method may include:

[0055] S100. Construct a functional fault tree for nuclear power instrumentation and control components.

[0056] Among them, nuclear power instrumentation and control cards are important components used in nuclear power plants to monitor, control and protect nuclear reactors and their auxiliary systems. Nuclear power instrumentation and control cards are responsible for performing specific functions. For example: data acquisition, signal processing, logical operations and control output, etc. Nuclear power instrumentation and control cards can include analog input cards, digital input / output cards, control cards, communication cards and special function cards according to hardware types.

[0057] Among them, the functional fault tree (FFT) is a logical model used to analyze the causes of nuclear power instrumentation and control card failures. The functional fault tree graphically displays the possible failures of nuclear power instrumentation and control card failures and their impact on system functions.

[0058] Specifically, the embodiment of the present invention can perform functional fault tree analysis (FTA) on the nuclear power instrumentation and control card component according to the hardware principle of the nuclear power instrumentation and control card component, and construct a functional fault tree of the nuclear power instrumentation and control card component.

[0059] S110. Use the functional fault tree to construct an initial instrumentation and control card health prediction model for nuclear power instrumentation and control cards.

[0060] Among them, the I&C card health prediction model is a model built based on a machine supervised learning algorithm for predicting the operating status of nuclear power I&C cards. The machine supervised learning algorithm may include a decision tree algorithm, a random forest algorithm, and a K-nearest neighbor (K-NN) algorithm.

[0061] The embodiment of the present invention can combine the functional fault tree and the machine supervised learning algorithm to construct an initial instrument and control card health prediction model for nuclear power instrument and control cards. The embodiment of the present invention constructs a health prediction model for nuclear power instrument and control cards based on a machine learning supervised learning algorithm, with the core goal of improving the safety and economic benefits of nuclear power plant operation, and implements preventive maintenance measures by pre-identifying potential faults of instrument and control cards, thereby reducing the risk of unplanned downtime and ensuring the continuous and stable operation of nuclear power plants.

[0062] S120. Obtain a key feature information training set of nuclear power instrumentation and control card components, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrumentation and control card components at at least one historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrumentation and control card components.

[0063] The health-related information is information data related to the operating status of the nuclear power instrumentation and control card components. The health-related information contains characteristic data that can reflect the operating status and health status of the nuclear power instrumentation and control card components.

[0064] Optionally, the health-related information may include process data, self-diagnosis status data and operating environment data of nuclear power instrumentation and control card components.

[0065] Among them, process data refers to the input or output data collected and processed by nuclear power instrumentation and control cards during the operation of the control system. For example: voltage and current data of nuclear power instrumentation and control cards.

[0066] Among them, self-diagnosis status data refers to monitoring data that reflects the health status of the nuclear power instrumentation and control card itself. For example, the status data of whether the memory, processor and other components of the nuclear power instrumentation and control card are working normally.

[0067] The operating environment data refers to various stress environment data of nuclear power instrumentation and control components, such as temperature and humidity environment and mechanical stress environment data.

[0068] It should be noted that for key chips in nuclear power instrumentation and control cards, the key chips can be tested according to relevant accelerated aging test standards to obtain the operating environment data of the key chips. Figure 2 As shown, a DC power supply and a wave recording instrument are used to test the nuclear power instrument and control card components to obtain the voltage data of key chips in the nuclear power instrument and control card components.

[0069] Optionally, after collecting the health-related information of the nuclear power instrumentation and control card at at least one historical moment, the embodiment of the present invention can perform data preprocessing on the health-related information to clean, standardize and select features in the health-related information to remove outliers and redundant information. Feature selection refers to selecting features that have a significant impact on the operating status target of the nuclear power instrumentation and control card predicted by the model from a large number of input parameters in the health-related information during the data preprocessing stage, that is, key feature parameters.

[0070] Furthermore, embodiments of the present invention can use key feature parameters to train machine learning models, such as decision trees, random forests, support vector machines, and neural networks. After the model training is completed, a parameter importance algorithm is used to calculate the relative importance of each key feature parameter to the model prediction result, wherein the parameter importance algorithm can be the Gini importance algorithm or information gain algorithm of the decision tree, or a permutation importance algorithm based on statistical methods. By analyzing the importance calculation results, the target feature parameters that are most important to the model prediction can be evaluated among the key feature parameters (for example, the higher the importance score, the more important it is), so that the target feature parameters can be collected in the subsequent process to generate the corresponding target feature vector.

[0071] The embodiments of the present invention can improve the interpretability and reliability of the model through data preprocessing and parameter importance analysis, by identifying and focusing on important features of the operating status targets of nuclear power instrumentation and control components predicted by the model, thereby improving the prediction accuracy of the model, reducing unnecessary consumption of computing resources, and enhancing the credibility and acceptability of the model.

[0072] The embodiment of the present invention can mark the target feature vector with a corresponding operation status result identifier according to the known operation status result corresponding to the target feature vector. The operation status result identifier can be "normal operation", "imminent failure" and "failed", or "low voltage", "unstable current", "processor failure", "communication interruption" and "overheating".

[0073] In the operation status monitoring and prediction tasks of the instrumentation and control card components of a nuclear power plant, the embodiment of the present invention provides a corresponding operation status result identification for each target feature vector sample. These identifications will serve as the "correct answers" in supervised learning, helping the training model to learn how to accurately predict the operation status of the instrumentation and control card components based on the input feature vectors, thereby facilitating real-time monitoring of the status of the instrumentation and control card components, timely detection of potential faults or abnormalities, and taking preventive measures to ensure the safe operation of the nuclear power plant.

[0074] The embodiment of the present invention makes full use of the historical data of nuclear power instrumentation and control cards, and through data mining and big data training technology, the historical data is effectively utilized, thereby facilitating the high-precision construction of a health prediction model for nuclear power instrumentation and control cards.

[0075] S130. Train the initial instrumentation and control component health prediction model using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control component health prediction model.

[0076] In the embodiment of the present invention, when the health prediction model of the instrument and control card is a model built based on the decision tree algorithm, starting from the root node, a feature and a threshold value of the feature are selected from the key feature information training set for splitting, so that the samples in the child nodes after the split belong to the same category as much as possible (or have the smallest error). This process is performed recursively until a certain stop condition is met, such as the number of samples in the node is less than a certain threshold, or the samples in the node have completely belonged to the same category, and a trained target instrument and control card health prediction model is obtained.

[0077] In the embodiment of the present invention, when the instrument and control card health prediction model is a model built based on the random forest algorithm, the target feature vector in the key feature information training set can be sampled with replacement (bootstrap sampling) to generate multiple sub-data sets. Each sub-data set is used to train a decision tree. After the training is completed, a trained target instrument and control card health prediction model is obtained. The target instrument and control card health prediction model obtains the final prediction result by voting or averaging the prediction results of all decision trees.

[0078] In the embodiment of the present invention, when the instrumentation and control card health prediction model is a model built based on the K-nearest neighbors (K-NN) algorithm, all target feature vectors in the key feature information training set can be directly stored to obtain a trained target instrumentation and control card health prediction model. When a prediction is required, the target instrumentation and control card health prediction model calculates the distance (such as the Euclidean distance) between the new sample and all training samples, and selects the K training samples with the closest distance. The category of the new sample is determined by the majority category of the K nearest neighbor samples, or by the average value of the output values ​​of the K nearest neighbor samples.

[0079] The embodiment of the present invention uses a machine supervised learning algorithm and utilizes historical process data, self-diagnosis status data, and operating environment data for training, so that the model has multiple years of operating experience at the initial stage of construction.

[0080] The present invention provides a method for obtaining an instrument and control card component health prediction model, which is applied to the technical field of nuclear power instrument and control card component data processing. The present invention constructs a functional fault tree of nuclear power instrument and control card components; uses the functional fault tree to construct an initial instrument and control card component health prediction model of nuclear power instrument and control card components; obtains a key feature information training set of nuclear power instrument and control card components, wherein the key feature information training set includes at least one target feature vector extracted from the health-related information of the nuclear power instrument and control card components at at least one historical moment, and the target feature vector is annotated with the operating status result identifier of the nuclear power instrument and control card components; uses the target feature vector in the key feature information training set to train the initial instrument and control card component health prediction model to obtain a trained target instrument and control card component health prediction model. The present invention effectively utilizes the historical health-related information of nuclear power instrument and control card components to construct a target instrument and control card component health prediction model, so that the target instrument and control card component health prediction model can predict the future health status of the nuclear power instrument and control card components, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0081] Optional, based on Figure 1 The method shown, such as Figure 3 As shown, a flowchart of a specific implementation of step S100 in the method for constructing an instrumentation and control card health prediction model provided by an embodiment of the present invention is shown, and step S100 may include:

[0082] S300, obtain the reliability function block diagram of nuclear power instrumentation and control components.

[0083] Among them, the Reliability Block Diagram (RBD) is a graphical tool that represents the reliability of each component in the nuclear power instrumentation and control card and their relationship with each other. The reliability block diagram is usually composed of a series of blocks or nodes, each block represents a component or subsystem in the system. These blocks are connected by lines to represent the logical relationship between components, such as series, parallel or more complex combinations.

[0084] Specifically, the embodiment of the present invention can obtain technical documents, design drawings, historical maintenance records and fault data of nuclear power instrumentation and control cards, and identify all key components and subsystems in nuclear power instrumentation and control cards based on these data. Each component can be represented in the form of a block or node in the reliability function block diagram, and the logical relationship between the components can be determined, including series, parallel or other more complex combinations. These relationships determine the behavior of the system when the component fails. The reliability of each component is evaluated, for example: evaluating indicators such as failure rate, mean time between failures (MTBF), and repair time (MTTR). Then use professional reliability analysis software or drawing tools to draw a reliability function block diagram based on the above information.

[0085] S310. Use reliability function block diagrams to identify key components and key circuits of nuclear power instrumentation and control components.

[0086] Specifically, the embodiments of the present invention can analyze the structure of the reliability function block diagram and identify series, parallel and more complex combination relationships. Each component in the series structure has a direct impact on the system reliability, while the components in the parallel structure provide redundancy. Perform reliability assessment on each component in the reliability function block diagram, including failure rate, impact range and repair time. The greater the impact of the failure of a component on the overall reliability of the system, the higher its importance. In a series structure, the failure of any component may cause the failure of the entire system, that is, a single point of failure (SPOF), and these components are key components. Analyze the failures of circuits or subsystems in the reliability function box that may have a significant impact on the system, and identify these circuits as key circuits.

[0087] S320. Perform functional development on key components and key circuits to obtain hardware functional development results of nuclear power instrumentation and control cards.

[0088] Specifically, the embodiment of the present invention can identify the specific functions of each key component and key circuit to obtain the hardware function expansion results of the nuclear power instrumentation and control card. The hardware function expansion results can include the function definition of each key component and key circuit (for example: specific role, input and output interface, interaction with other components, etc.), technical specifications (for example: performance parameters, working conditions, reliability indicators), design details (for example: circuit diagram, layout diagram, bill of materials and manufacturing process) and interaction with other system components (for example: signal transmission, data exchange and energy transfer).

[0089] S330. Use the hardware function expansion results to construct a functional fault tree of nuclear power instrumentation and control cards from the component level to the circuit level.

[0090] Among them, the functional fault tree can include fault data of nuclear power instrumentation and control cards at the component level, assembly level, subsystem level, module level, board level and circuit level.

[0091] Specifically, the embodiment of the present invention can determine the top event of the functional fault tree, that is, the undesirable event in the nuclear power instrumentation and control card, such as system failure and safety accident. Based on the hardware function expansion results, the functions of each key component and key circuit, as well as their role in the system, including input function, processing function and output function, are analyzed. Starting from the component level, basic events that may cause the top event to occur are identified. These basic events include component failure, circuit failure, signal transmission error, etc. Logic gates (such as AND, OR, NOT, etc.) are used to represent the logical relationship between each basic event. For example, if the failure of multiple components occurs at the same time to cause the top event, an AND gate is used, and if the failure of any component causes the top event, an OR gate is used. Analyze the functions of each component and circuit and the interactions between them, starting from the top event, and expand downward step by step until all basic events are identified and connected to the logic gates, thereby constructing a functional fault tree from the component level to the circuit level.

[0092] The embodiment of the present invention analyzes the reliability function block diagram of nuclear power instrumentation and control components and applies machine supervised learning algorithms to build an intelligent health prediction model layer by layer from the key component level to the board equipment level. The model can automatically identify the operation mode and health status of nuclear power instrumentation and control components, predict potential failures, and achieve early warning and maintenance decision support.

[0093] The embodiment of the present invention can more accurately identify key components and key circuits through the reliability function block diagram of nuclear power instrumentation and control cards. The failure of these components often has a significant impact on the performance of nuclear power instrumentation and control cards, and the function expansion and fault tree analysis further refine the functions and potential failure modes of these components, ensuring the accuracy of the constructed functional fault tree of nuclear power instrumentation and control cards, and also providing reliable data support for the subsequent construction of the instrumentation and control card health prediction model, thereby improving the accuracy of the instrumentation and control card health prediction model.

[0094] Optional, in the above Figure 1 On the basis of one or more corresponding embodiments, another optional embodiment provided by the embodiment of the present invention uses a functional fault tree to construct an initial instrumentation and control card health prediction model for a nuclear power instrumentation and control card, including:

[0095] The functional fault tree is substituted into the pre-compiled machine supervised learning algorithm program according to the fault category, and the machine supervised learning algorithm program is executed to build the initial instrumentation and control card health prediction model of nuclear power instrumentation and control cards.

[0096] In a functional fault tree, fault events are usually classified, for example, by the nature of the fault (such as electrical fault, mechanical fault, software fault, etc.) or by the impact of the fault (such as minor fault, serious fault, safety fault, etc.). These classifications help machine learning algorithms better understand and process different types of fault data.

[0097] The embodiment of the present invention can convert the fault events and logical relationships in the functional fault tree into a fault data format that can be processed by a machine learning algorithm, and add these fault data to a pre-compiled machine supervised learning algorithm program according to the fault category for model training. During the training process, the machine supervised learning algorithm will try to find patterns in the fault data and establish an initial instrumentation and control card health prediction model. The initial instrumentation and control card health prediction model can predict future fault events of nuclear power instrumentation and control cards based on the input features.

[0098] The embodiment of the present invention combines the fault data of the functional fault tree of nuclear power instrumentation and control cards and the machine supervised learning algorithm program to construct an initial instrumentation and control card health prediction model suitable for nuclear power instrumentation and control cards, thereby helping to identify and prevent potential failures in advance, improve the reliability and safety of nuclear power instrumentation and control cards, reduce operation and maintenance costs, and support more effective operation and maintenance decision-making.

[0099] Optional, in the above Figure 1 On the basis of one or more corresponding embodiments, in another optional embodiment provided by the embodiment of the present invention, the target feature vector in the key feature information training set is used to train the initial instrumentation and control card component health prediction model to obtain the trained target instrumentation and control card component health prediction model, which may specifically include:

[0100] When the machine supervised learning algorithm is a random forest algorithm, the target feature vector in the key feature information training set is used to train multiple decision trees in the initial instrumentation and control card health prediction model, so that each decision tree is split based on the feature importance to obtain a trained target instrumentation and control card health prediction model.

[0101] Specifically, an embodiment of the present invention can perform a sampling with replacement on the target feature vector in the key feature information training set to generate multiple sub-datasets. Each sub-dataset is used to train a decision tree separately. In the process of building the decision tree, the node splitting is based on the randomly selected sub-dataset. When the node is split, the information gain or Gini impurity of each feature is calculated, and the feature that can best improve the prediction accuracy is selected for splitting to improve the generalization ability of the model and avoid overfitting. After all decision trees have been trained independently, the final prediction result is obtained by summarizing the prediction results of all decision trees. The trained random forest model, that is, the one containing multiple trained decision trees, can be used as a target instrumentation and control card health prediction model to predict the current health status of nuclear power instrumentation and control cards.

[0102] The embodiment of the present invention uses a random forest algorithm to train the initial instrumentation and control card health prediction model, which can automatically evaluate the importance of each feature, thereby identifying the most critical features for predicting the health status of nuclear power instrumentation and control cards, and improve the overall accuracy of the model by integrating the prediction results of multiple decision trees, thereby improving the prediction accuracy of the target instrumentation and control card health prediction model.

[0103] Optional, based on Figure 1 The method shown, such as Figure 4 As shown, a flow chart of another implementation of a method for constructing an instrumentation and control card health prediction model provided by an embodiment of the present invention is shown. After step S130, the construction method may further include:

[0104] S400, using a cross-validation algorithm and a network search algorithm to tune the model parameters of the target instrumentation and control card component health prediction model to obtain an optimized target instrumentation and control card component health prediction model.

[0105] Specifically, the embodiment of the present invention can use a grid search algorithm to define the search space of the parameter combination of the model parameters. Use a cross-validation algorithm to cross-validate each parameter combination to evaluate the model performance. Select the parameter combination with the best cross-validation result as the optimal model parameter. Apply the optimal model parameters to the target instrumentation and control card health prediction model to obtain an optimized target instrumentation and control card health prediction model.

[0106] The embodiment of the present invention combines the use of a cross-validation algorithm and a network search algorithm to accurately find the model parameters with the best performance results, and uses the optimal model parameters to optimize the target instrumentation and control card health prediction model, thereby improving the prediction accuracy and generalization ability of the target instrumentation and control card health prediction model.

[0107] Optional, based on Figure 4 The method shown, such as Figure 5 As shown, a flow chart of another implementation of a method for constructing an instrumentation and control card health prediction model provided by an embodiment of the present invention is shown. After step S400, the construction method may further include:

[0108] S500. Obtain a key feature information verification set of a nuclear power instrument and control card component, wherein the key feature information verification set includes at least one target feature vector extracted from historical health-related information of the nuclear power instrument and control card component, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrument and control card component.

[0109] It should be noted that the target feature vector in the key feature information verification set is the feature vector that did not participate in the training of the initial I&C card health prediction model.

[0110] The embodiment of the present invention can collect a large number of target feature parameters and generate corresponding target feature vectors after evaluating the target feature parameters that are most important for model prediction from the key feature parameters, and divide these target feature vectors into a key feature information training set and a key feature information verification set according to a certain ratio. The key feature information training set is used to train the initial instrumentation and control card health prediction model, and the key feature information verification set is used to verify the target instrumentation and control card health prediction model.

[0111] The embodiment of the present invention performs model verification on a verification set of key feature information extracted from historical health-related information of nuclear power instrumentation and control card components, thereby ensuring a high degree of relevance of the verification process and a degree of matching with actual application scenarios. Model verification is performed using a target feature vector annotated with an operating status result identifier, which can more directly and effectively evaluate the accuracy of the model's prediction of the actual operating status.

[0112] S510, using the target feature vector in the key feature information verification set, verify the optimized target instrumentation and control card health prediction model to obtain a prediction accuracy evaluation result of the target instrumentation and control card health prediction model.

[0113] Specifically, the embodiment of the present invention can input the target feature vector in the key feature information verification set into the optimized target instrument and control card health prediction model, so that the model can predict the health status of the nuclear power instrument and control card corresponding to each target feature vector, and calculate a series of evaluation indicators to measure the prediction accuracy of the model based on the prediction results of the model and the operating status result identification marked by the feature vector, wherein the evaluation indicators can include accuracy, precision, recall rate and F1 score. Finally, according to the calculated evaluation indicators, the prediction performance of the model is analyzed to obtain the prediction accuracy evaluation result of the target instrument and control card health prediction model.

[0114] The embodiment of the present invention can quickly iterate and optimize the instrumentation and control card component health prediction model according to the prediction accuracy evaluation result until a target instrumentation and control card component health prediction model that meets the prediction accuracy requirement is obtained.

[0115] The embodiment of the present invention verifies the optimized target I&C card health prediction model by utilizing the target feature vector in the key feature information verification set, which can not only obtain the prediction accuracy evaluation result of the model, but also provide direction for the continuous optimization of the model, thereby helping to build a more robust and reliable I&C card health prediction model and improve the performance of the I&C card health prediction model in predicting the operating status of nuclear power instrumentation and control cards.

[0116] like Figure 6 As shown, a flowchart of an implementation method of a method for using an instrumentation and control card health prediction model provided by an embodiment of the present invention is provided. The method for obtaining the model may include:

[0117] S600, obtaining the target feature vector of the nuclear power instrumentation and control card at the current moment.

[0118] Specifically, the embodiment of the present invention can collect the target characteristic parameters of the nuclear power instrumentation and control card components at the current moment in real time, or collect the target characteristic parameters of the nuclear power instrumentation and control card components at the current moment according to a scheduled task, and organize the target characteristic parameters into corresponding target characteristic vectors.

[0119] S610. Input the target feature vector into the target instrumentation and control card component health prediction model obtained by the method for constructing the instrumentation and control card component health prediction model provided in the embodiment of the present disclosure, and obtain the operation status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model.

[0120] The embodiment of the present invention inputs the target feature vector into the target instrumentation and control card health prediction model, and the target instrumentation and control card health prediction model processes the target feature vector. In the case where the target instrumentation and control card health prediction model is a model built based on the random forest algorithm, each decision tree in the target instrumentation and control card health prediction model will independently evaluate the target feature vector to obtain the prediction result output by each decision tree. The prediction process of the decision tree is based on the branch conditions of the tree. By comparing the feature value with the threshold on the tree node, it moves down along the path of the tree until it reaches a leaf node. Each leaf node corresponds to a prediction category or prediction value. Finally, the prediction results of each decision tree are integrated to generate the final prediction result of the operating status of the nuclear power instrumentation and control card.

[0121] The target instrumentation and control card health prediction model provided in the embodiment of the present invention can run on the monitoring layer of the digital control system, and does not run directly on the circuit board of the nuclear power instrumentation and control card, so as to decouple the target instrumentation and control card health prediction model from the card hardware, ensuring that the operating status prediction of the nuclear power instrumentation and control card is not affected by the hardware faults and failures of the card.

[0122] The embodiment of the present invention realizes the automation of the entire chain from the characteristic data collection of nuclear power instrumentation and control card components to the operation status prediction, thereby improving the intelligence level of nuclear power plant operation and maintenance, and can provide strong support for the implementation of preventive maintenance strategies.

[0123] The present invention provides a method for using an instrument and control card health prediction model, which is applied to the technical field of nuclear power instrument and control card data processing. The present invention obtains the target feature vector of the nuclear power instrument and control card at the current moment. The target feature vector is input into the target instrument and control card health prediction model obtained by the construction method of the instrument and control card health prediction model provided in the embodiment of the present disclosure, and the operating status prediction result of the nuclear power instrument and control card output by the target instrument and control card health prediction model is obtained. The present invention predicts the future health status of the nuclear power instrument and control card through the target feature vector of the nuclear power instrument and control card at the current moment and through the trained target instrument and control card health prediction model, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0124] Although operations are depicted in a particular order, this should not be understood as requiring that the operations be performed in the particular order shown or in a sequential order.Multitasking and parallel processing may be advantageous under certain circumstances.

[0125] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0126] Corresponding to the method embodiment of the method for constructing the health prediction model of the instrument and control card component, the embodiment of the present invention also provides a device for constructing the health prediction model of the instrument and control card component, and its structure is as follows: Figure 7 As shown, it may include: a functional fault tree building unit 10, an initial model building unit 11, a training set obtaining unit 12 and a target model obtaining unit 13.

[0127] The functional fault tree construction unit 10 is used to construct the functional fault tree of the nuclear power instrumentation and control card component.

[0128] The initial model building unit 11 is used to build an initial instrumentation and control card health prediction model for nuclear power instrumentation and control cards by using a functional fault tree.

[0129] The training set acquisition unit 12 is used to obtain a key feature information training set of a nuclear power instrumentation and control card component, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrumentation and control card component at at least one historical moment, and the target feature vector is marked with an operating status result identifier of the nuclear power instrumentation and control card component.

[0130] The target model obtaining unit 13 is used to train the initial instrumentation and control component health prediction model using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control component health prediction model.

[0131] Optionally, the functional fault tree construction unit 10 can be specifically used to obtain a reliability function block diagram of a nuclear power instrument and control card; use the reliability function block diagram to identify key components and key circuits of the nuclear power instrument and control card; perform functional expansion of key components and key circuits to obtain hardware function expansion results of the nuclear power instrument and control card; use the hardware function expansion results to construct a functional fault tree of the nuclear power instrument and control card from the component level to the circuit level.

[0132] Optionally, the initial model building unit 11 can be specifically used to substitute the functional fault tree into a pre-compiled machine supervised learning algorithm program according to the fault category, execute the machine supervised learning algorithm program, and build an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card.

[0133] Optional. The target model acquisition unit 13 can be specifically used to train multiple decision trees in the initial instrumentation and control card health prediction model using the target feature vector in the key feature information training set when the machine supervised learning algorithm program is a random forest algorithm, so that each decision tree is split based on feature importance to obtain a trained target instrumentation and control card health prediction model.

[0134] Optionally, the construction device may further include: a model optimization unit.

[0135] The model optimization unit is used for the target model acquisition unit 13 to train the initial instrumentation and control card health prediction model using the target feature vector in the key feature information training set. After obtaining the trained target instrumentation and control card health prediction model, the model parameters of the target instrumentation and control card health prediction model are tuned using a cross-validation algorithm and a network search algorithm to obtain an optimized target instrumentation and control card health prediction model.

[0136] Optionally, the construction device may further include: a model verification unit.

[0137] The model verification unit is used for the model optimization unit to use the cross-validation algorithm and the network search algorithm to tune the model parameters of the target instrumentation and control card component health prediction model. After obtaining the optimized target instrumentation and control card component health prediction model, the target feature vector in the key feature information verification set is used to verify the optimized target instrumentation and control card component health prediction model to obtain the prediction accuracy evaluation result of the target instrumentation and control card component health prediction model.

[0138] Optionally, the health-related information includes process data, self-diagnosis status data and operating environment data of nuclear power instrumentation and control card components.

[0139] The present invention provides a device for obtaining a health prediction model of an instrument and control card, which is applied to the technical field of data processing of nuclear power instrument and control card. The present invention constructs a functional fault tree of a nuclear power instrument and control card; uses the functional fault tree to construct an initial instrument and control card health prediction model of the nuclear power instrument and control card; obtains a key feature information training set of the nuclear power instrument and control card, wherein the key feature information training set includes at least one target feature vector extracted from the health-related information of the nuclear power instrument and control card at at least one historical moment, and the target feature vector is marked with the operating status result identifier of the nuclear power instrument and control card; uses the target feature vector in the key feature information training set to train the initial instrument and control card health prediction model to obtain a trained target instrument and control card health prediction model. The present invention effectively utilizes the historical health-related information of the nuclear power instrument and control card to construct a target instrument and control card health prediction model, so that the target instrument and control card health prediction model can predict the future health status of the nuclear power instrument and control card, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0140] Corresponding to the method embodiment of the method for using the instrument and control card health prediction model, the embodiment of the present invention also provides a device for using the instrument and control card health prediction model, and its structure is as follows: Figure 8 As shown, it may include: a current feature vector obtaining unit 20 and an operation status prediction result obtaining unit 21.

[0141] The current feature vector obtaining unit 20 is used to obtain the target feature vector of the nuclear power instrumentation and control card component at the current moment.

[0142] The operation status prediction result obtaining unit 21 is used to input the target feature vector into the target instrumentation and control card component health prediction model obtained by the construction device of the instrumentation and control card component health prediction model in the embodiment of the present invention, and obtain the operation status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model.

[0143] The present invention provides a device for using an instrument and control card health prediction model, which is applied to the technical field of nuclear power instrument and control card data processing. The present invention obtains the target feature vector of the nuclear power instrument and control card at the current moment. The target feature vector is input into the target instrument and control card health prediction model obtained by the construction method of the instrument and control card health prediction model provided by the embodiment of the present disclosure, and the operating status prediction result of the nuclear power instrument and control card output by the target instrument and control card health prediction model is obtained. The present invention predicts the future health status of the nuclear power instrument and control card through the target feature vector of the nuclear power instrument and control card at the current moment and through the trained target instrument and control card health prediction model, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0144] Regarding the device in the above embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0145] The device for constructing the health prediction model of the instrumentation and control card component includes a processor and a memory. The above-mentioned functional fault tree construction unit 10, initial model construction unit 11, training set acquisition unit 12 and target model acquisition unit 13 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

[0146] The device for using the instrumentation and control card health prediction model includes a processor and a memory. The current feature vector acquisition unit 20 and the operation status prediction result acquisition unit 21 are stored in the memory as program units, and the processor executes the program units stored in the memory to realize the corresponding functions.

[0147] The processor includes a kernel, which calls the corresponding program unit from the memory. One or more kernels can be set, and the historical health association information of nuclear power instrumentation and control cards can be effectively utilized by adjusting the kernel parameters to build a target instrumentation and control card health prediction model. The target instrumentation and control card health prediction model can be used to predict the future health status of nuclear power instrumentation and control cards, thereby achieving preventive maintenance and improving the safety and reliability of nuclear power plants.

[0148] An embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the method for constructing the instrumentation and control card component health prediction model and / or the method for using the instrumentation and control card component health prediction model.

[0149] An embodiment of the present invention provides a processor, which is used to run a program, wherein when the program is running, the method for constructing the instrumentation and control card component health prediction model and / or the method for using the instrumentation and control card component health prediction model are executed.

[0150] like Fig. 9 As shown, an embodiment of the present invention provides an electronic device 1000, which includes at least one processor 1001, and at least one memory 1002 and a bus 1003 connected to the processor 1001; wherein the processor 1001 and the memory 1002 communicate with each other through the bus 1003; the processor 1001 is used to call the program instructions in the memory 1002 to execute the above-mentioned method for constructing the health prediction model of the instrument and control card component, and / or the method for using the health prediction model of the instrument and control card component. The electronic device in this article can be a server, a PC, a PAD, a mobile phone, etc.

[0151] The present invention also provides a computer program product, which, when executed on an electronic device, is suitable for executing a program that initializes a method for constructing a health prediction model for an instrumentation and control card component, and / or a method for using the health prediction model for an instrumentation and control card component.

[0152] The present invention is described with reference to flowcharts and / or block diagrams of methods, apparatuses, electronic devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable device to generate a machine, so that the instructions executed by the processor of the computer or other programmable device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0153] In a typical configuration, an electronic device includes one or more processors (CPU), a memory, and a bus. The electronic device may also include an input / output interface, a network interface, and the like.

[0154] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip. The memory is an example of a computer-readable medium.

[0155] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0156] In the description of the present invention, it should be understood that the terms "up", "down", "front", "back", "left" and "right" etc. indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the positions or elements referred to must have specific directions, be constructed and operate in specific directions. Therefore, they should not be understood as limitations of the present invention.

[0157] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. It should also be noted that the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0158] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0159] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for constructing an instrumentation and control card health prediction model, characterized in that: include: Obtain the reliability function block diagram of nuclear power instrumentation and control components; Using the reliability function block diagram to identify key components and key circuits of the nuclear power instrumentation and control card; Performing functional expansion on the key components and the key circuits to obtain hardware functional expansion results of the nuclear power instrumentation and control card; using the hardware functional expansion results, constructing a functional fault tree of the nuclear power instrumentation and control card from the component level to the circuit level; the functional fault tree at least includes fault data of the nuclear power instrumentation and control card at the component level, assembly level, subsystem level, module level, board level and circuit level; Using the functional fault tree, constructing an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card; A key feature information training set of the nuclear power instrument and control card component is obtained, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrument and control card component at at least one historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrument and control card component; the health-related information contains feature data that can reflect the operating status and health status of the nuclear power instrument and control card component; The initial instrumentation and control card component health prediction model is trained using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model.

2. The method according to claim 1, characterized in that The method of using the functional fault tree to construct an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card comprises: The functional fault tree is substituted into a pre-compiled machine supervised learning algorithm program according to the fault category, and the machine supervised learning algorithm program is executed to construct an initial instrumentation and control card health prediction model for the nuclear power instrumentation and control card.

3. The method according to claim 2, characterized in that The using the target feature vector in the key feature information training set to train the initial instrumentation and control card component health prediction model to obtain a trained target instrumentation and control card component health prediction model includes: When the machine supervised learning algorithm is a random forest algorithm, the target feature vector in the key feature information training set is used to train multiple decision trees in the initial instrumentation and control card health prediction model, so that each decision tree is split based on feature importance to obtain a trained target instrumentation and control card health prediction model.

4. The method according to claim 1, characterized in that: After the initial instrumentation and control card component health prediction model is trained using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model, the method further includes: The model parameters of the target instrumentation and control card component health prediction model are tuned using a cross-validation algorithm and a network search algorithm to obtain an optimized target instrumentation and control card component health prediction model.

5. The method according to claim 4, characterized in that After the cross-validation algorithm and the network search algorithm are used to tune the model parameters of the target instrumentation and control card component health prediction model to obtain the optimized target instrumentation and control card component health prediction model, the method further includes: Obtaining a key feature information verification set of the nuclear power instrument and control card component, wherein the key feature information verification set includes at least one target feature vector extracted from the health-related information of the nuclear power instrument and control card component at the historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrument and control card component; The target feature vector in the key feature information verification set is used to verify the optimized target instrumentation and control card component health prediction model to obtain a prediction accuracy evaluation result of the target instrumentation and control card component health prediction model.

6. The method according to any one of claims 1 to 5, characterized in that The health-related information includes process data, self-diagnosis status data and operating environment data of the nuclear power instrumentation and control card component.

7. A method for using an instrumentation and control card health prediction model, characterized in that: include: Obtain the target feature vector of the nuclear power instrumentation and control card at the current moment; The target feature vector is input into the target instrumentation and control card component health prediction model obtained by the method for constructing the instrumentation and control card component health prediction model according to any one of claims 1 to 6, and the operating status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model is obtained.

8. A device for constructing an instrumentation and control card health prediction model, characterized in that: include: Functional fault tree building unit, initial model building unit, training set acquisition unit and target model acquisition unit, The functional fault tree construction unit is used to obtain a reliability function block diagram of a nuclear power instrument and control card; use the reliability function block diagram to identify key components and key circuits of the nuclear power instrument and control card; perform functional expansion on the key components and key circuits to obtain hardware function expansion results of the nuclear power instrument and control card; use the hardware function expansion results to construct a functional fault tree of the nuclear power instrument and control card from the component level to the circuit level; the functional fault tree at least contains fault data of the nuclear power instrument and control card at the component level, assembly level, subsystem level, module level, board level and circuit level; The initial model building unit is used to build an initial instrumentation and control card component health prediction model of the nuclear power instrumentation and control card component by using the functional fault tree; The training set acquisition unit is used to obtain a key feature information training set of the nuclear power instrumentation and control card component, wherein the key feature information training set includes at least one target feature vector extracted from health-related information of the nuclear power instrumentation and control card component at at least one historical moment, and the target feature vector is annotated with an operating status result identifier of the nuclear power instrumentation and control card component; the health-related information contains feature data that can reflect the operating status and health status of the nuclear power instrumentation and control card component; The target model acquisition unit is used to train the initial instrumentation and control card component health prediction model using the target feature vector in the key feature information training set to obtain a trained target instrumentation and control card component health prediction model.

9. A device for using an instrumentation and control card health prediction model, characterized in that: include: Current feature vector acquisition unit and operation status prediction result acquisition unit, The current feature vector obtaining unit is used to obtain the target feature vector of the nuclear power instrumentation and control card at the current moment; The operating status prediction result obtaining unit is used to input the target feature vector into the target instrumentation and control card component health prediction model obtained by the construction device of the instrumentation and control card component health prediction model in claim 8, and obtain the operating status prediction result of the nuclear power instrumentation and control card component output by the target instrumentation and control card component health prediction model.