Method for constructing a steady-state intelligent monitoring model for system equipment components of nuclear power units

Abstract

This invention relates to a method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, comprising: acquiring the fault modes of the monitored object; determining the measurement points of the steady-state intelligent monitoring model based on the fault modes of the monitored object; acquiring training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model; calculating the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data; setting the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted values; and setting the sampling period of the steady-state intelligent monitoring model based on the data change period of the training data, thereby completing the construction of the steady-state intelligent monitoring model. This invention associates the selection of model measurement points with the fault modes of the monitored object, making the model's representation of the monitored object in steady-state operation accurate, comprehensive, and realistic; it also reduces the volume of training data; and it allows all monitored state deviations to be interpreted to specific fault modes, improving the interpretability of the model.

Description

Technical Field

[0001] This invention relates to the technical field of condition monitoring, and more specifically, to a method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit. Background Technology

[0002] Nuclear power plants have numerous systems and equipment. According to incomplete statistics, a single unit can have over 200 systems and more than 85,000 devices, with approximately 15,000 measurement points and a massive amount of data. During normal, stable operation, traditional monitoring methods are insufficient to effectively detect early deviations in the condition of systems, equipment, and components. Therefore, the industry has introduced intelligent monitoring technology that utilizes data analytics for condition monitoring. This technology analyzes online measurement data through computer systems and identifies early deviations by setting dynamic alarm thresholds.

[0003] However, on the one hand, since intelligent monitoring technology uses data analysis technology for status monitoring, its model is a data model and does not consider the selection methods and principles of measurement points to describe the status of systems, equipment, and components, or whether the data collected by the measurement points can comprehensively, accurately, and truthfully reflect the status of systems, equipment, and components. This results in the need for a large amount of historical data for training after the model is established, slow model growth, and low accuracy. The introduction of redundant measurement points in the model will cause a large number of false alarms during the model's operation. The lack of measurement points in the model will cause missed alarms during the model's operation.

[0004] On the other hand, even if nuclear power plant personnel are involved in the model development, the selection of measurement points to form the model is based on personal experience, resulting in the development of models that vary from person to person. When the model runs and triggers an alarm, it is difficult to explain what equipment or component has experienced what kind of state deviation.

[0005] The main shortcomings of current intelligent monitoring model development methods include:

[0006] (1) The selection methods and principles for the measurement points of the model are not fully considered. The intelligent monitoring model that relies solely on the data model is difficult to comprehensively, accurately and realistically describe the monitoring model of the system, equipment and components.

[0007] (2) The training data required for model development is large. When the model measurement points are not selected properly, the information carried by the measurement points needs to be compensated by the data volume.

[0008] (3) The interpretability of the intelligent monitoring model is poor. The deviation information is not associated with the fault mode during the model development stage, which makes it difficult to implement the state deviation monitored during the model operation stage into specific equipment, components and their fault modes. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide a method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, in order to address the deficiencies of the prior art.

[0010] The technical solution adopted by this invention to solve its technical problem is: a method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, comprising the following steps:

[0011] Obtain the fault mode of the monitored object, and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object;

[0012] Based on the measurement points of the steady-state intelligent monitoring model, the training data of the steady-state intelligent monitoring model is obtained;

[0013] Calculate the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data;

[0014] The dynamic alarm threshold of the steady-state intelligent monitoring model is set based on the predicted value;

[0015] The sampling period of the steady-state intelligent monitoring model is set according to the data change period of the training data to complete the construction of the steady-state intelligent monitoring model.

[0016] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, the step of acquiring the fault mode of the monitored object and determining the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object includes:

[0017] Obtain the failure mode of the monitored object;

[0018] The failure modes of the monitored object are analyzed to obtain the failure phenomena of the monitored object; the failure phenomena include: characteristic parameter groups;

[0019] The feature parameter set is preprocessed to obtain preprocessed data;

[0020] All online measurement points of the monitored object are obtained, and the online measurement points are analyzed to obtain online measurement points that represent characteristic parameters;

[0021] The measurement points of the steady-state intelligent monitoring model are determined based on the preprocessed data and the online measurement points representing the characteristic parameters.

[0022] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, the step of analyzing the fault modes of the monitored object to obtain the fault phenomena of the monitored object includes:

[0023] The abnormal physical structure, energy and mass transfer, and medium characteristics of the monitored object are analyzed to obtain the fault phenomena of the monitored object.

[0024] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit as described in this invention, after obtaining online measurement points representing characteristic parameters, the method further includes:

[0025] Collect historical data from online measurement points representing the characteristic parameters.

[0026] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, the step of preprocessing the feature parameter set to obtain preprocessed data includes:

[0027] The feature parameters and their changes in the feature parameter group are classified and summarized.

[0028] After completing the classification and summarization, all feature parameters in the feature parameter group are deduplicated to obtain all unique feature parameters of the monitored object.

[0029] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, determining the measurement points of the steady-state intelligent monitoring model based on the preprocessed data and the online measurement points representing characteristic parameters includes:

[0030] Based on the online measurement points representing the characteristic parameters, obtain the measurement content of the online measurement points representing the characteristic parameters;

[0031] Determine whether the preprocessed data contains feature parameters that match the measurement content;

[0032] If so, the online measurement points of representative feature parameters that match the measurement content with the feature parameters in the preprocessed data are determined as candidate points;

[0033] The measurement points of the steady-state intelligent monitoring model are determined based on the candidate points.

[0034] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, determining the measurement points of the steady-state intelligent monitoring model based on the candidate points includes:

[0035] All candidate points are filtered according to preset principles to obtain the filtered candidate points;

[0036] The selected candidate points are the measurement points of the steady-state intelligent monitoring model.

[0037] In the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to the present invention, the step of obtaining training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model includes:

[0038] Based on the measurement points of the steady-state intelligent monitoring model, historical data of the measurement points of the steady-state intelligent monitoring model are extracted from the historical data collected by the online measurement points representing the characteristic parameters.

[0039] The historical data of the measurement points of the steady-state intelligent monitoring model are set as the training data of the steady-state intelligent monitoring model.

[0040] This invention also provides a system for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, comprising:

[0041] The measurement point determination unit is used to acquire the fault mode of the monitored object and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object.

[0042] The training data acquisition unit is used to acquire the training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model.

[0043] The prediction value calculation unit is used to calculate the predicted value of each measurement point of the steady-state intelligent monitoring model based on the training data.

[0044] A threshold setting unit is used to set the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted value;

[0045] The sampling period setting unit is used to set the sampling period of the steady-state intelligent monitoring model according to the data change period of the training data, so as to complete the construction of the steady-state intelligent monitoring model.

[0046] The present invention also provides a computer-readable storage medium storing a computer program adapted for loading by a processor to execute the steps of the method for constructing a steady-state intelligent monitoring model of system equipment components of a nuclear power unit as described above.

[0047] The present invention also provides a computer, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the steps of the method for constructing a steady-state intelligent monitoring model of system equipment components of a nuclear power unit as described above by calling the computer program stored in the memory.

[0048] The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, implementing the present invention, has the following beneficial effects: It includes: acquiring the fault modes of the monitored object; determining the measurement points of the steady-state intelligent monitoring model based on the fault modes of the monitored object; acquiring training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model; calculating the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data; setting the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted values; and setting the sampling period of the steady-state intelligent monitoring model based on the data change period of the training data, thus completing the construction of the steady-state intelligent monitoring model. This invention associates the selection of model measurement points with the fault modes of the monitored object, making the model's description of the monitored object in steady-state operation accurate, comprehensive, and realistic; it also reduces the volume of training data; and it allows all monitored state deviations to be interpreted to specific fault modes, improving the interpretability of the model. Attached Figure Description

[0049] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0050] Figure 1 This is a flowchart illustrating the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, as provided in an embodiment of the present invention.

[0051] Figure 2 This is a schematic diagram of the measurement point selection process of the steady-state intelligent monitoring model provided in this embodiment of the invention;

[0052] Figure 3 This is a schematic diagram of the structure of the system for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, as provided in an embodiment of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.

[0054] To address the problems of incomplete selection methods and principles for measurement points, large volume of required training data, and poor model interpretability in the development of steady-state intelligent monitoring models for system and equipment component state deviations in nuclear power plant units, this invention provides a method for constructing a steady-state intelligent monitoring model for system and equipment components of nuclear power plants. The steady-state intelligent monitoring model constructed by this method can achieve accurate, comprehensive, and realistic intelligent monitoring of state deviations of nuclear power plant unit systems, equipment, and components.

[0055] Specifically, such as Figure 1 As shown, the method for constructing the steady-state intelligent monitoring model of the system equipment components of this nuclear power unit includes the following steps:

[0056] Step S101: Obtain the fault mode of the monitored object, and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object.

[0057] Specifically, such as Figure 2 As shown, obtaining the fault modes of the monitored object and determining the measurement points of the steady-state intelligent monitoring model based on the fault modes of the monitored object includes:

[0058] Step S111: Obtain the fault mode of the monitored object.

[0059] Optionally, in this embodiment of the invention, the monitoring objects include: systems, equipment, and components of the nuclear power unit. The failure mode of the monitoring object refers to the form of failure of the monitoring system, equipment, and components. For example, bearing wear.

[0060] Step S112: Analyze the fault modes of the monitored object to obtain the fault phenomena of the monitored object. Among them, the fault phenomena include: characteristic parameter groups.

[0061] Optionally, in this embodiment of the invention, the analysis of the failure mode of the monitored object can be performed from multiple dimensions, such as abnormal physical structure of the monitored object, abnormal energy and mass transfer of the monitored object, and abnormal medium characteristics of the monitored object, to obtain various phenomena of the monitored object's failure, i.e., failure phenomena of the monitored object (here, failure phenomena of the monitored object include: system failure phenomena, equipment failure phenomena, and component failure phenomena). These failure phenomena are expressed in the form of characteristic parameter sets, i.e., these failure phenomena include: characteristic parameter sets. Optionally, in this embodiment of the invention, the content described by the characteristic parameter set includes: the name of the characteristic parameter corresponding to the failure phenomenon, the change process of the characteristic parameter, etc.

[0062] Among them, the analysis of abnormal physical structure of the monitored object, abnormal energy and mass transfer of the monitored object, and abnormal medium characteristics of the monitored object is as follows: analyze the physical meaning described by the characteristic parameter, and clarify whether the characteristic parameter represents the state of the physical structure of the monitored object (such as size, wear, corrosion, etc.), the state of energy / mass transfer (such as temperature, pressure, flow rate, etc.), or the state of medium characteristics (such as lubricating oil particle size, water content, viscosity, steam dryness, etc.).

[0063] In this embodiment of the invention, characteristic parameters refer to physical parameters that characterize the fault phenomena of a fault mode, such as vibration, bearing inner diameter, and bearing surface smoothness as characteristic parameters of bearing wear. The change process of characteristic parameters refers to the changes in characteristic parameters over time, i.e., the characteristics of time series data.

[0064] Step S113: Perform data preprocessing on the feature parameter group to obtain preprocessed data.

[0065] In some embodiments, data preprocessing of the feature parameter group to obtain preprocessed data includes: classifying and summarizing the feature parameters and their changes in the feature parameter group; and after completing the classification and summarization, deduplicating all feature parameters in the feature parameter group to obtain all unique feature parameters of the monitored object.

[0066] Specifically, firstly, the characteristic parameters and their changes in the characteristic parameter group are classified and summarized according to the combination form of the characteristic parameter and the characteristics of the change process of the characteristic parameter group; secondly, all characteristic parameters after classification and summarization are deduplicated to identify all unique characteristic parameters of the monitored object.

[0067] In this context, the combination of characteristic parameters refers to a combination of characteristics that describe the monitored object. For example, characteristic parameters describing equipment vibration include: online X-vibration, online Y-vibration, online Z-vibration, and offline vibration spectrum measurements. These parameters are collectively referred to as a combination of characteristic parameters.

[0068] In this embodiment of the invention, the characteristics of the feature parameter group change process include: the feature parameters rising or falling synchronously, one feature parameter being positively correlated with another feature parameter, and the feature parameters changing periodically.

[0069] Step S114: Obtain all online measurement points of the monitored object and analyze the online measurement points to obtain online measurement points that represent characteristic parameters.

[0070] Specifically, first, all online measurement points related to the state of the monitored object are collected. Then, the measurement content of these online measurement points is analyzed to determine whether the measurements are of characteristic parameters related to the physical structure, energy and mass transfer, and medium properties of the monitored object. These online measurement points are the representative characteristic parameters. After obtaining the representative characteristic parameters, historical data collected from these online measurement points are collected.

[0071] The state of the monitored object refers to its operational status, such as the pump's outlet pressure, speed, and current, as well as the vibration and temperature of the pump bearings. These states refer to characteristic parameters related to the physical structure, energy and mass transfer, and the state of the medium. State-related online measurement points refer to sensors that can directly or indirectly reflect these characteristic parameters.

[0072] Online measurement points refer to sensors that can continuously monitor the status of equipment and send data to external devices.

[0073] Step S115: Determine the measurement points of the steady-state intelligent monitoring model based on the preprocessed data and the online measurement points representing the characteristic parameters.

[0074] In some embodiments, determining the measurement points of the steady-state intelligent monitoring model based on preprocessed data and online measurement points representing characteristic parameters includes: obtaining the measurement content of the online measurement points representing characteristic parameters based on the online measurement points representing characteristic parameters; determining whether there are characteristic parameters in the preprocessed data that match the measurement content; if so, determining the online measurement points of the representative characteristic parameters that match the measurement content with the characteristic parameters in the preprocessed data as candidate points; and determining the measurement points of the steady-state intelligent monitoring model based on the candidate points.

[0075] Specifically, firstly, the feature parameters are associated with online measurement points that represent the feature parameters. That is, if there is an online measurement point that matches the feature parameter in the preprocessed data (feature parameters summarized after deduplication), then that online measurement point is recorded as a candidate point for the steady-state intelligent monitoring model.

[0076] In addition, if there is no online measurement point in the preprocessed data that matches the characteristic parameter, the fault mode is recorded as a fault mode without intelligent monitoring model management. At the same time, if there is no monitoring object fault mode that matches the online measurement point that matches the characteristic parameter, the online measurement point is recorded as an observation point.

[0077] In some embodiments, determining the measurement points of the steady-state intelligent monitoring model based on the candidate points includes: screening all candidate points according to preset principles to obtain the selected candidate points; the selected candidate points are the measurement points of the steady-state intelligent monitoring model.

[0078] Optionally, in this embodiment of the invention, the preset principle is as follows: for all determined candidate points, for the same feature parameter, no more than two (i.e., one or two) candidate points corresponding to it are selected, and the remaining candidate points are deleted. The one or two online measurement candidate points selected are used as measurement points of the steady-state intelligent monitoring model. The measurement points of the steady-state intelligent monitoring model are the points involved in the calculations of the steady-state intelligent monitoring model, including: measurement point combination calculations, measurement point correlation calculations, etc.

[0079] In addition, all recorded observation points need to be processed; that is, only one measurement point with the same properties is selected, and the rest are filtered out. These observation points serve as state observation points for the steady-state intelligent monitoring model and do not participate in any calculations.

[0080] Step S102: Obtain the training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model.

[0081] In some embodiments, obtaining training data for the steady-state intelligent monitoring model based on its measurement points includes: extracting historical data of the measurement points of the steady-state intelligent monitoring model from historical data collected from online measurement points representing feature parameters; and setting the historical data of the measurement points of the steady-state intelligent monitoring model as the training data of the steady-state intelligent monitoring model.

[0082] In this embodiment of the invention, after obtaining the training data of the steady-state intelligent monitoring model, the training data with corresponding features are further filtered out using the feature parameters and their change process characteristics in the preprocessed data, so that the remaining training data can be used as normal state data for training the steady-state intelligent monitoring model.

[0083] Step S103: Calculate the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data.

[0084] Specifically, after obtaining the training data of the steady-state intelligent monitoring model, the predicted values ​​for each measurement point are calculated. These predicted values ​​can be obtained using various algorithms and hyperparameter combinations. For example, one or more of the following algorithms can be used to calculate the predicted values ​​for each measurement point: neural network algorithm, linear regression algorithm, nonlinear regression algorithm, classification algorithm, and clustering algorithm. In this embodiment of the invention, the predicted value refers to the value of each measurement point in the historical state most similar to the current state.

[0085] Among them, the algorithm and hyperparameter values ​​that minimize the deviation between the predicted value and the actual value (i.e., the historical data of the measurement point) can be selected.

[0086] Step S104: Set the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted value.

[0087] Specifically, after the predicted value is calculated, the calculated predicted value is used as a reference, and the allowed (such as user-allowed) absolute or relative positive or negative offset is used as the dynamic alarm threshold for each level, thereby completing the setting of the dynamic alarm threshold of the steady-state intelligent monitoring model.

[0088] The positive and negative offsets here can be determined based on expert experience or design documents. These positive and negative offsets represent the threshold values. For example, if the measured value is 10 and the predicted value is 11, my experience suggests that the fluctuation range of this parameter is generally within ±3, then the deviation of the current measured value from the predicted value is acceptable. However, if my experience suggests that the parameter fluctuation range is within ±0.5, then the deviation of the current measured value from the predicted value is unacceptable.

[0089] Step S105: Set the sampling period of the steady-state intelligent monitoring model according to the data change period of the training data to complete the construction of the steady-state intelligent monitoring model.

[0090] Optionally, in this embodiment of the invention, the sampling period is less than or equal to 1 / 10 of the data change period presented by the training data.

[0091] Furthermore, in this embodiment of the invention, in order to verify that historical data used as training data will not generate false alarms or miss alarms, the method for constructing the steady-state intelligent monitoring model of the system equipment components of the nuclear power unit of the present invention further performs the following steps before putting the steady-state intelligent monitoring model into use:

[0092] Test data is determined based on historical data; the intelligent monitoring model is then validated using the test data.

[0093] Specifically, the collected historical data is input into the steady-state intelligent monitoring model as test data for calculation. If the input historical data does not trigger an alarm but the other historical data does, the accuracy of the steady-state intelligent monitoring model is considered to meet the requirements. Otherwise, steps S101 to S105 are iteratively repeated and verified until the accuracy requirements are met.

[0094] refer to Figure 3 This invention provides a system for constructing a transient intelligent monitoring model for system equipment components of a nuclear power unit.

[0095] The construction system may include:

[0096] The measurement point determination unit 301 is used to acquire the fault mode of the monitored object and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object.

[0097] The training data acquisition unit 302 is used to acquire the training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model.

[0098] The prediction calculation unit 303 is used to calculate the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data.

[0099] The threshold setting unit 304 is used to set the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted value.

[0100] The sampling period setting unit 305 is used to set the sampling period of the steady-state intelligent monitoring model according to the data change period of the training data, so as to complete the construction of the steady-state intelligent monitoring model.

[0101] The present invention also provides a computer-readable storage medium storing a computer program adapted for loading by a processor to execute the steps of the method for constructing a steady-state intelligent monitoring model of system equipment components of a nuclear power unit disclosed in the embodiments of the present invention.

[0102] The present invention also provides a computer, including a memory and a processor. The memory stores a computer program, and the processor executes the steps of the method for constructing a steady-state intelligent monitoring model of system equipment components of a nuclear power unit disclosed in the embodiments of the present invention by calling the computer program stored in the memory.

[0103] The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit disclosed in this invention associates the selection of measurement points for the steady-state intelligent monitoring model with the fault modes of the monitored object, and takes into account principles such as mutual verification between measurement points and minimizing the number of measurement points, so that the intelligent monitoring model's description of the monitored object in steady-state operation is accurate, comprehensive, and true.

[0104] Because this method achieves the most comprehensive and minimal selection of measurement points for the steady-state intelligent monitoring model, a small amount of training data from steady-state operation can fully describe the required state change features, thus greatly reducing the volume of training data.

[0105] Since this method realizes that the measurement points of the steady-state intelligent monitoring model are the embodiment of the fault mode of the monitored object, any state deviation detected can be interpreted as the fault mode of the specific monitored object. Users can quickly and accurately understand the fault meaning of the state deviation in the alarm feedback.

[0106] This steady-state intelligent monitoring model development method solidifies fault mode analysis into three aspects: the physical structure of the monitoring object, energy and mass transfer, and medium characteristics, making the selection of measurement points more direct, fast, and accurate.

[0107] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0108] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0109] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0110] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They do not limit the scope of protection of the present invention. All equivalent changes and modifications made within the scope of the claims of the present invention should fall within the scope of the claims of the present invention.

Claims

1. A method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, characterized in that, Includes the following steps: Obtain the fault mode of the monitored object, and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object; Based on the measurement points of the steady-state intelligent monitoring model, the training data of the steady-state intelligent monitoring model is obtained; Calculate the predicted values ​​of each measurement point of the steady-state intelligent monitoring model based on the training data; The dynamic alarm threshold of the steady-state intelligent monitoring model is set based on the predicted value; The sampling period of the steady-state intelligent monitoring model is set according to the data change period of the training data to complete the construction of the steady-state intelligent monitoring model; The process of acquiring the fault mode of the monitored object and determining the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object includes: Obtain the failure mode of the monitored object; The failure modes of the monitored object are analyzed to obtain the failure phenomena of the monitored object; the failure phenomena include: characteristic parameter groups; The feature parameter set is preprocessed to obtain preprocessed data; All online measurement points of the monitored object are obtained, and the online measurement points are analyzed to obtain online measurement points that represent characteristic parameters; Based on the preprocessed data and the online measurement points representing the characteristic parameters, the measurement points of the steady-state intelligent monitoring model are determined; The step of determining the measurement points of the steady-state intelligent monitoring model based on the preprocessed data and the online measurement points representing the characteristic parameters includes: Based on the online measurement points representing the characteristic parameters, obtain the measurement content of the online measurement points representing the characteristic parameters; Determine whether the preprocessed data contains feature parameters that match the measurement content; If so, the online measurement points of representative feature parameters that match the measurement content with the feature parameters in the preprocessed data are determined as candidate points; The measurement points of the steady-state intelligent monitoring model are determined based on the candidate points.

2. The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to claim 1, characterized in that, The analysis of the failure modes of the monitored object, and the resulting failure phenomena, include: The abnormal physical structure, energy and mass transfer, and medium characteristics of the monitored object are analyzed to obtain the fault phenomena of the monitored object.

3. The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to claim 1, characterized in that, After obtaining online measurement points representing characteristic parameters, the following steps are also included: Collect historical data from online measurement points representing the characteristic parameters.

4. The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to claim 2, characterized in that, The step of preprocessing the feature parameter set to obtain preprocessed data includes: The feature parameters and their changes in the feature parameter group are classified and summarized. After completing the classification and summarization, all feature parameters in the feature parameter group are deduplicated to obtain all unique feature parameters of the monitored object.

5. The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to claim 1, characterized in that, The determination of measurement points for the steady-state intelligent monitoring model based on the candidate points includes: All candidate points are filtered according to preset principles to obtain the filtered candidate points; The selected candidate points are the measurement points of the steady-state intelligent monitoring model.

6. The method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit according to claim 3, characterized in that, The step of obtaining the training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model includes: Based on the measurement points of the steady-state intelligent monitoring model, historical data of the measurement points of the steady-state intelligent monitoring model are extracted from the historical data collected by the online measurement points representing the characteristic parameters. The historical data of the measurement points of the steady-state intelligent monitoring model are set as the training data of the steady-state intelligent monitoring model.

7. A system for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit, constructed using the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit as described in claim 1, characterized in that... include: The measurement point determination unit is used to acquire the fault mode of the monitored object and determine the measurement points of the steady-state intelligent monitoring model based on the fault mode of the monitored object. The training data acquisition unit is used to acquire the training data of the steady-state intelligent monitoring model based on the measurement points of the steady-state intelligent monitoring model. The prediction value calculation unit is used to calculate the predicted value of each measurement point of the steady-state intelligent monitoring model based on the training data. A threshold setting unit is used to set the dynamic alarm threshold of the steady-state intelligent monitoring model based on the predicted value; The sampling period setting unit is used to set the sampling period of the steady-state intelligent monitoring model according to the data change period of the training data, so as to complete the construction of the steady-state intelligent monitoring model.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted for loading by a processor to perform the steps of the method for constructing a steady-state intelligent monitoring model for system equipment components of a nuclear power unit as described in any one of claims 1 to 6.

9. A computer, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the steps of the method for constructing a steady-state intelligent monitoring model of system equipment components of a nuclear power unit as described in any one of claims 1 to 6 by calling the computer program stored in the memory.

Images