Nuclear power plant unit state abnormity detection and diagnosis method and system, storage medium and electronic equipment

Abstract

The invention relates to a nuclear power plant unit state abnormity detection and diagnosis method and system, a storage medium and electronic equipment. The method comprises the following steps: acquiring historical global parameters of a nuclear power plant; constructing a PCA model in a normal operation state based on the historical global parameters; monitoring real-time global parameters of the nuclear power plant based on the PCA model, and judging whether the real-time global parameters deviate from normal values or not; if yes, performing anomaly detection and diagnosis on the nuclear power plant unit by adopting a global-local parameter combined sign directed graph; and if not, continuing to monitor the real-time global parameters of the nuclear power plant based on the PCA model. According to the method, the situation that an envelope alarm value is set through a large amount of analysis work can be avoided, unit state deviation can be found earlier, large occurrence is prevented, faults are diagnosed in time, and shutdown is avoided. And operation and maintenance personnel can quickly understand, trace and confirm.

Description

Technical Field

[0001] The present invention relates to the technical field of abnormality detection of nuclear power units, and more specifically, to a method, system, storage medium and electronic equipment for detecting and diagnosing abnormality of a nuclear power plant unit state. Background Art

[0002] In order to monitor the operating status of nuclear power units and find operating deviations in time, it is necessary to monitor the important operating parameters of nuclear power plants and find abnormal parameters in time. When the important parameters deviate from normal operation to a certain extent, the unit is considered to be in an abnormal state, and it is necessary to find the cause of the fault and then troubleshoot and correct the deviation. For the determination of parameter abnormalities, nuclear power plants currently use a fixed threshold method to set alarm values ​​based on fixed thresholds. The alarm values ​​are usually based on the response of the unit under various assumed fault conditions, or based on the analysis of functions exceeding the allowable operating range. When an alarm occurs, the operator performs operations such as finding the cause of the fault and controlling the unit according to the corresponding alarm procedures. For faults that cause transient states in the unit, it is usually difficult to diagnose the source equipment of the fault. The operator initially focuses on the control of the unit status, and later conducts diagnosis and troubleshooting of the source equipment of the fault, mainly based on a large number of changing parameters to trace back the transient process and conduct expert analysis.

[0003] In order to achieve abnormal detection and diagnosis of nuclear power units, there are currently two commonly used methods. The first method: Under the assumed design benchmark conditions, analyze the range and trend of various parameters, consider the consequences of the operator's non-intervention, set the corresponding threshold, and inform the operator of the occurrence of a fault in the form of an alarm or remind the operator to intervene as soon as possible, such as the alarm threshold of the leakage rate of the first and second circuits, the high threshold of the radioactivity of the second circuit steam generator, etc.; in addition, the threshold can be set according to the normal operating range of various functional requirements. Once the threshold is exceeded, it means that the function cannot be implemented normally, and the function is declared lost or deteriorated, such as the low alarm threshold of the upper filling flow, the low alarm threshold of the lower discharge flow, etc. The second method: Use statistical methods to set variable thresholds for parameter abnormalities.

[0004] However, the threshold determined by the first method can usually only be exceeded after a transient or functional failure occurs, and cannot detect the deviation of the unit's early operation from normal; at the same time, the fixed threshold cannot adapt to the situation where the normal range of parameters of the unit changes under different operating conditions. There may be a situation where the threshold is exceeded and an alarm is generated even when the unit is started and stopped normally. The second method can only be used for parameter deviation warning, and cannot realize the abnormal diagnosis function and cannot realize the overall status judgment of the unit. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method, system, storage medium and electronic equipment for detecting and diagnosing abnormal status of nuclear power plant units in view of the problems existing in the prior art.

[0006] The technical solution adopted by the present invention to solve the technical problem is: constructing a method for detecting and diagnosing abnormal state of a nuclear power plant unit, comprising the following steps:

[0007] Obtain historical global parameters of nuclear power plants;

[0008] Building a PCA model under normal operating conditions based on the historical global parameters;

[0009] Monitoring the real-time global parameters of the nuclear power plant based on the PCA model, and determining whether the real-time global parameters deviate from normal values;

[0010] If so, a global-local parameter combined signed directed graph is used to detect and diagnose anomalies in nuclear power plant units;

[0011] If not, continue to monitor the real-time global parameters of the nuclear power plant based on the PCA model.

[0012] In the method for detecting and diagnosing abnormal state of a nuclear power plant unit according to the present invention, the acquisition of historical global parameters of the nuclear power plant includes:

[0013] Determine the parameters reflecting the reactivity control of nuclear power plants;

[0014] Determine the parameters reflecting the removal of waste heat from nuclear power plants;

[0015] Determine the parameters reflecting the radioactivity of nuclear power plants;

[0016] Determine the influencing parameters of the unit transient caused by abnormalities;

[0017] The historical global parameter is acquired according to the parameter reflecting the reactivity control of the nuclear power plant, the parameter reflecting the residual heat extraction of the nuclear power plant, the parameter reflecting the radioactivity of the nuclear power plant and the influencing parameter.

[0018] In the method for detecting and diagnosing abnormal conditions of a nuclear power plant unit according to the present invention, the PCA model under normal operating conditions is constructed based on the historical global parameters, including:

[0019] Based on the historical global parameters, determining global parameter samples of multiple units under normal operating conditions;

[0020] The PCA model is constructed based on the global parameter sample.

[0021] In the method for detecting and diagnosing abnormal conditions of nuclear power plant units according to the present invention, constructing the PCA model based on the global parameter sample comprises:

[0022] Constructing a sample matrix based on the global parameter samples;

[0023] Preprocessing the data of the sample matrix to obtain preprocessed data;

[0024] Calculation is performed based on the preprocessed data to obtain the PCA model.

[0025] In the method for detecting and diagnosing abnormal conditions of nuclear power plant units according to the present invention, preprocessing the data of the sample matrix to obtain preprocessed data includes:

[0026] The data of the sample matrix is ​​subjected to mean standardization and normalization processing to obtain the preprocessed data.

[0027] In the method for detecting and diagnosing abnormal conditions of nuclear power plant units according to the present invention, the calculation based on the preprocessed data to obtain the PCA model includes:

[0028] Perform calculation based on the preprocessed data to obtain a covariance matrix;

[0029] Calculating eigenvalues ​​and eigenvectors of the covariance matrix to obtain eigenvalues ​​and eigenvectors of the covariance matrix;

[0030] Processing is performed according to the eigenvalues ​​and eigenvectors of the covariance matrix to obtain a coding matrix;

[0031] Reconstructing based on the sample matrix and the encoding matrix to obtain a reconstructed data matrix;

[0032] Perform residual calculation based on the reconstructed data matrix to obtain a residual matrix;

[0033] Calculate statistics according to the residual matrix to obtain statistics;

[0034] The statistical threshold is set to complete the construction of the PCA model.

[0035] In the method for detecting and diagnosing abnormal state of a nuclear power plant unit according to the present invention, the monitoring of the real-time global parameters of the nuclear power plant based on the PCA model and judging whether the real-time global parameters deviate from normal values ​​include:

[0036] Performing real-time monitoring on the nuclear power plant units to obtain real-time global parameters of the nuclear power plant;

[0037] Normalizing the real-time global parameters based on the mean and variance used in the normalization calculation of the historical global parameters to obtain a normalized data matrix;

[0038] Reconstructing the normalized data matrix based on the encoding matrix to obtain a reconstructed matrix;

[0039] Calculating residual statistics between vectors of the normalized data matrix and vectors of the reconstruction matrix;

[0040] Whether the real-time global parameter deviates from a normal value is determined according to the residual statistic and the statistic threshold.

[0041] In the method for detecting and diagnosing abnormal conditions of a nuclear power plant unit according to the present invention, judging whether the real-time global parameter deviates from a normal value according to the residual statistic and the statistic threshold comprises:

[0042] comparing the residual statistic with the statistic threshold;

[0043] If the residual statistic is greater than the statistic threshold, it is determined that the currently acquired real-time global parameter is abnormal data;

[0044] Calculating the residual contribution rate of each real-time global parameter in the real-time global parameters;

[0045] According to whether the residual contribution rate of each real-time global parameter is greater than a set threshold;

[0046] If so, it is determined that the real-time global parameter deviates from the normal value;

[0047] If not, it is determined that the real-time global parameter has no abnormality.

[0048] In the method for detecting and diagnosing abnormal conditions of nuclear power plant units according to the present invention, setting the statistical threshold value includes:

[0049] The percentile method is used to set the statistical threshold.

[0050] In the method for detecting and diagnosing abnormal conditions of nuclear power plant units of the present invention, the abnormality detection and diagnosis of nuclear power plant units using a global-local parameter combined signed directed graph includes:

[0051] Based on the real-time global parameters that deviate from normal values, a global-local parameter symbolic directed graph is drawn;

[0052] Abnormal detection and diagnosis of nuclear power plant units are performed according to the global-local parameter symbol directed graph.

[0053] The present invention also provides a nuclear power plant unit state abnormality detection and diagnosis system, comprising:

[0054] A historical parameter acquisition unit, used to acquire historical global parameters of the nuclear power plant;

[0055] A PCA model building unit, used for building a PCA model under normal operating conditions based on the historical global parameters;

[0056] A global parameter monitoring unit, used to monitor the real-time global parameters of the nuclear power plant based on the PCA model, and determine whether the real-time global parameters deviate from normal values;

[0057] The anomaly detection and diagnosis unit is used to detect and diagnose anomalies of nuclear power plant units using a symbolic directed graph combining global and local parameters.

[0058] The present invention also provides a storage medium storing a computer program, wherein the computer program is suitable for being loaded by a processor to execute the steps of the method for detecting and diagnosing abnormal status of a nuclear power plant unit as described above.

[0059] The present invention also provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the steps of the above-mentioned method for detecting and diagnosing abnormal status of nuclear power plant units by calling the computer program stored in the memory.

[0060] The method, device, storage medium and electronic device for detecting and diagnosing abnormal conditions of nuclear power plant units implemented in the present invention have the following beneficial effects: including the following steps: obtaining historical global parameters of the nuclear power plant; constructing a PCA model under normal operating conditions based on historical global parameters; monitoring the real-time global parameters of the nuclear power plant based on the PCA model, and judging whether the real-time global parameters deviate from normal values; if so, using a symbolic directed graph combining global and local parameters to detect and diagnose abnormalities of the nuclear power plant units; if not, continuing to monitor the real-time global parameters of the nuclear power plant based on the PCA model. The present invention can avoid setting the alarm value of the envelope through a large amount of analysis work, can detect the deviation of the unit state earlier, prevent a large one from occurring, diagnose the fault in time, and avoid shutdown and reactor shutdown. It can also facilitate the operation and maintenance personnel to quickly understand, trace and confirm. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0062] Figure 1 It is a flow chart of a method for detecting and diagnosing abnormal state of a nuclear power plant unit provided by the present invention;

[0063] Figure 2 It is a local SDG diagram of a steam flow increase fault type combining global and local parameters provided by the present invention;

[0064] Figure 3 It is a logic block diagram of the abnormal state detection and diagnosis system of the nuclear power plant unit of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] In order to solve the problems existing in the prior art, the present invention provides a method for detecting and diagnosing abnormal status of a nuclear power plant unit, by which the abnormal status of the unit can be discovered in time, the cause of the abnormality can be analyzed in time, and the faulty equipment can be found and isolated in time.

[0067] refer to Figure 1 , Figure 1 A flowchart of a preferred embodiment of a method for detecting and diagnosing abnormal state of a nuclear power plant unit provided by the present invention.

[0068] Specifically, Figure 1 As shown, the method for detecting and diagnosing abnormal state of a nuclear power plant unit includes the following steps:

[0069] Step S101: Acquire historical global parameters of a nuclear power plant.

[0070] Optionally, in an embodiment of the present invention, obtaining the historical global parameters of a nuclear power plant includes: determining parameters reflecting the reactivity control of the nuclear power plant; determining parameters reflecting the residual heat extraction of the nuclear power plant; determining parameters reflecting the radioactivity of the nuclear power plant; determining the influencing parameters of the transient state of the unit caused by the abnormality; obtaining the historical global parameters according to the parameters reflecting the reactivity control of the nuclear power plant, the parameters reflecting the residual heat extraction of the nuclear power plant, the parameters reflecting the radioactivity of the nuclear power plant and the influencing parameters. Among them, the historical global parameters refer to the existing global parameters (i.e., the data stored in the database or system).

[0071] Specifically, the global parameters of a nuclear power plant refer to the parameters that characterize changes in the operating status of the unit. Among them, the historical global parameters can be determined by the following methods:

[0072] (1) Determine the parameters that can reflect the status of the three major safety functions of the nuclear power plant: reactivity control, residual heat removal, and radioactivity containment. Specific parameters may include but are not limited to: reactor coolant temperature, pressure, water level, flow rate, core power, steam generator liquid level, pressure, feedwater flow rate, containment pressure, radioactivity, etc.;

[0073] (2) Carry out qualitative or quantitative abnormal condition consequence analysis for different types of abnormal conditions, determine the unit transients caused by the assumed abnormalities, and select the parameters affected by these transients. Specifically, it may include but is not limited to: accidents such as increased steam flow in the secondary circuit may cause an increase in core power and a decrease in primary circuit temperature and pressure, so these parameters are used as global parameters. This process can be used to check whether the global parameters selected by method (1) are complete, ensuring that the selected parameter range can cover all target abnormal conditions.

[0074] Step S102: constructing a PCA model under normal operation based on historical global parameters.

[0075] PCA: Principal component analysis, which is used for dimensionality reduction and data compression. It converts the original variables into a set of new variables. These new variables are explicit combinations of the original variables and are independent of each other. Its main goal is to reduce the maintenance of the data set while retaining as much information as possible. Currently, related research has been used for nuclear power plant fault monitoring. PCA for parameter monitoring is mainly divided into the following steps: (1) Sample the m system parameters n times to form a training set; (2) Normalize the n vectors; (3) Calculate the covariance matrix; (4) Calculate the eigenvectors and eigenvalues ​​of the covariance matrix; (5) Take the eigenvector values ​​with a certain proportion of values, determine the principal components, and their corresponding eigenvectors to form a transformation matrix; (6) Use the transformation matrix to reduce the dimension of the original training data to obtain the reduced-dimensional data; (7) The reduced-dimensional data is then reconstructed into the principal component matrix through the decoding matrix; (8) Calculate T 2 and Q statistics, and set a threshold based on the distribution of the statistics; (9) If it exceeds the threshold, it is considered an anomaly, and the specific abnormal parameter is determined by the residual contribution of each parameter.

[0076] Optionally, in an embodiment of the present invention, constructing a PCA model under normal operating conditions based on historical global parameters includes: determining global parameter samples of multiple units under normal operating conditions based on historical global parameters; and constructing a PCA model based on the global parameter samples.

[0077] Furthermore, constructing a PCA model based on the global parameter samples includes: constructing a sample matrix based on the global parameter samples; preprocessing data of the sample matrix to obtain preprocessed data; and performing calculations based on the preprocessed data to obtain a PCA model.

[0078] Wherein, the data of the sample matrix is ​​preprocessed to obtain the preprocessed data, including: performing mean standardization and normalization processing on the data of the sample matrix to obtain the preprocessed data. Wherein, the mean standardization and normalization processing on the data of the sample matrix is ​​specifically: subtracting the mean of the parameter in all samples from each data and then dividing it by the standard deviation. In an embodiment of the present invention, the PCA model is obtained by calculating based on the preprocessed data, including: calculating based on the preprocessed data to obtain a covariance matrix; calculating eigenvalues ​​and eigenvectors of the covariance matrix to obtain eigenvalues ​​and eigenvectors of the covariance matrix; processing according to the eigenvalues ​​and eigenvectors of the covariance matrix to obtain a coding matrix; reconstructing based on the sample matrix and the coding matrix to obtain a reconstructed data matrix; performing residual calculation based on the reconstructed data matrix to obtain a residual matrix; calculating statistics according to the residual matrix to obtain statistics; setting a statistical threshold to complete the construction of the PCA model. Wherein, setting the statistical threshold includes: setting the statistical threshold using the percentile method.

[0079] Specifically, first, from the historical global parameters obtained, multiple global parameters of the units under normal operating conditions are selected as data samples, namely global parameter samples. Let m be the number of global parameter samples, n be the number of samples, and construct a sample matrix, set as X n×m Then, the data in the sample matrix is ​​normalized by the mean value to obtain the standardized data matrix. Then, the standardized data matrix is ​​calculated to obtain the covariance matrix, which is set as C = X T X / (n-1), perform eigenvalues ​​(λ1, λ2, λ3, ... λ) on the covariance matrix n ) and feature vectors (v1, v2, v3, ... v n ) is solved to obtain the eigenvalues ​​and eigenvectors of the agreement difference matrix. Then, the eigenvectors corresponding to the first k eigenvalues ​​are selected using the cumulative variance contribution rate method to form the encoding matrix P n×m . Transform the data matrix X n×m Projected to the principal component space, we get T = XP. Then reconstruct the data matrix to obtain the reconstructed data matrix The residual matrix is ​​calculated based on the reconstructed data matrix, that is Computing statistics using the residual matrix where e ij is the element in the i-th row and j-th column of E. Finally, the percentile method (such as 95% or 99%) is used to set the statistical threshold of Q. In other words, the data group exceeding this limit is an outlier. Among them, the constructed PCA model includes: the constructed encoding matrix P n×m And set the statistical threshold Q.

[0080] Step S103: monitor the real-time global parameters of the nuclear power plant based on the PCA model, and determine whether the real-time global parameters deviate from normal values.

[0081] Optionally, in an embodiment of the present invention, real-time global parameters of a nuclear power plant are monitored based on a PCA model, and it is determined whether the real-time global parameters deviate from normal values, including: real-time monitoring of nuclear power plant units to obtain real-time global parameters of the nuclear power plant; normalizing the real-time global parameters based on the mean and variance used in the normalization calculation of historical global parameters to obtain a normalized data matrix; reconstructing the normalized data matrix based on a coding matrix to obtain a reconstructed matrix; calculating the residual statistics of the vectors of the normalized data matrix and the vectors of the reconstructed matrix; and determining whether the real-time global parameters deviate from normal values ​​based on the residual statistics and a statistical threshold.

[0082] Optionally, in an embodiment of the present invention, judging whether a real-time global parameter deviates from a normal value based on a residual statistic and a statistic threshold includes: comparing the residual statistic with the statistic threshold; if the residual statistic is greater than the statistic threshold, judging that the currently acquired real-time global parameter is abnormal data; calculating the residual contribution rate of each real-time global parameter in the real-time global parameter; judging whether the residual contribution rate of each real-time global parameter is greater than a set threshold; if so, judging that the real-time global parameter deviates from the normal value; if not, judging that the real-time global parameter is normal.

[0083] Specifically, the real-time global parameters of the nuclear power plant at a certain moment are obtained in real time, wherein the real-time global parameters are the global parameters collected at the current moment, and are defined as S 1×m , for S 1×m Normalization is performed, and then the encoding matrix P constructed in step S102 is n×m After normalization, S 1×m Reconstruct and obtain the reconstruction matrix S′ 1×m Then, calculate the normalized S 1×m The vector and P n×m The residual statistics of the vector are: where e i is the coordinate value in the residual vector. Compare q with Q. If q is greater than Q, it means that the data set is abnormal. This monitoring method takes into account the coupling factors of various global parameters and detects the overall abnormal deviation of the unit. That is, this method can detect the overall abnormal deviation of the unit's operating status.

[0084] Among them, for abnormal data (i.e. abnormal parameters), the residual contribution rate of each real-time global parameter is calculated to further determine whether it is an abnormal parameter. Specifically, the residual contribution rate is: e iis the coordinate value in the residual vector, q is the residual statistic, c i is the residual contribution rate of the i-th real-time global parameter.

[0085] Furthermore, after the real-time global parameter is determined to be an abnormal parameter through the residual contribution rate, further judgment is required to ensure the accuracy of the judgment. Specifically, the global parameters of the nuclear power plant are configured with redundant instruments. If the measurement value of a single instrument is abnormal, it is considered that a single instrument failure rather than a global parameter abnormality is considered. If multiple instrument readings of the same measurement value are abnormal, the global parameter can be determined to be abnormal. That is, when it is determined to be a single instrument failure, step S104 does not need to be executed. When multiple instrument readings of the same measurement value are abnormal, step S104 needs to be executed.

[0086] Step S104: If yes, use the global-local parameter combined signed directed graph to detect and diagnose the abnormality of the nuclear power plant unit; if no, continue to monitor the real-time global parameters of the nuclear power plant based on the PCA model. If no, return to step S103 to continue monitoring.

[0087] Optionally, in an embodiment of the present invention, using a symbolic directed graph combining global and local parameters to perform abnormal detection and diagnosis on nuclear power plant units includes: drawing a symbolic directed graph of global and local parameters based on real-time global parameters that deviate from normal values; and performing abnormal detection and diagnosis on nuclear power plant units according to the global and local parameter symbolic directed graph.

[0088] Specifically, for the case of eliminating instrument faults, that is, when it is determined that the global parameters are abnormal, the abnormality diagnosis can be further performed by drawing a symbolic graph. The symbolic directed graph method (SDG) is a graphical method for system fault diagnosis and causal analysis. It helps identify and locate faults by constructing a causal relationship graph between system components, and is particularly suitable for complex industrial systems. It consists of nodes, symbols and edges. The nodes are key variables and components. The causal relationship between them is established. They are connected by directed edges, and the symbols of each edge are marked ("+", "-"). The complexity of the graph is simplified by merging paths and deleting redundant nodes. The process of fault diagnosis using SDG is as follows: (1) Use physical equations or control processes or expert knowledge methods to establish a symbolic directed graph; (2) When the system is abnormal, record the changes in the abnormal quantity; (3) Use the SDG graph to start from the abnormal variable and deduce along the edge of the symbolic directed graph until the fault source is obtained; (4) Confirm the fault location and cause through further testing or verification. The establishment of the SDG graph relies on expert knowledge, and reasoning can only be performed after the abnormal variable is determined.

[0089] Based on the above principle, after determining that the global parameter is abnormal, the present invention further uses a symbolic directed graph combining global and local parameters to perform abnormal diagnosis on the nuclear power plant unit. The local parameters of a nuclear power plant refer to those parameters that are directly related to the operating status of the equipment. For example, they may include but are not limited to: the speed and flow of the pump, the voltage and current of the motor, the opening of the valve, the inlet and outlet pressure difference of the heat exchanger, the local pressure of the pipeline, etc. Specifically, a symbolic directed graph can be drawn for each fault type. Among them, the graph contains global parameters related to the fault type and local parameters closely related to the initial fault. For a certain equipment fault, multiple local parameters can be selected for fault confirmation. For equipment faults, parameters that characterize the overall functional failure of the equipment are selected, such as the outlet flow and outlet pressure of the pump; the valve position and the control valve position change rate are selected for the valve; the heat exchanger selects the outlet fluid temperature and outlet flow of the heat exchanger. Parameters that only characterize the functional failure of equipment components, such as vibration, oil pressure, etc., are not selected because the diagnostic purpose of the present invention is to locate the equipment at the source of the fault after the transient state of the unit has occurred. For passive faults such as pipeline ruptures, local parameters select the pressure difference or flow at different positions of the pipeline. In addition, changes in system parameters caused by equipment failure, such as system flow and pressure, can also be used to confirm the occurrence of a fault and can also be regarded as local parameters.

[0090] The following is a specific example. Figure 2 As shown, it is a local example of the steam flow increase fault SDG provided by the present invention. Among them, Figure 2 In the figure, the rectangular box is the originating fault, the circular box is the local parameter, the ellipse is the global parameter, the actual value indicates that the parameter becomes larger after the fault occurs, and the dotted line indicates that the change of the previous parameter causes the next parameter to decrease. Figure 2 In the symbolic directed graph combining global and local parameters, the global parameters are obtained by monitoring in step S103, and the local parameters can be directly obtained through the human-machine interface of the nuclear power plant. Figure 2 As shown in the figure, after the symbol directed graph is drawn, the originating fault can be quickly and accurately inferred in reverse according to the direction of parameter fluctuation in the node and the symbol directed graph (i.e., SDG graph). In order to further confirm the inference result, it is also possible to start from the inferred originating fault and forwardly infer the state of each parameter of the nuclear power plant, and finally compare whether the forward inferred state of each parameter of the nuclear power plant is consistent with the actual state. If they are consistent, the inferred originating fault is confirmed.

[0091] For example, Figure 2As shown in the figure, if the pressure difference between other SG (steam generator) and SG3 is abnormal, it can be reversely inferred that the initial fault is the SG3 VDA isolation valve accidentally opened or the pipeline between the SG3 VVP valves is broken. Further, it can be inferred and confirmed from the local parameters. If the local parameter monitoring at this time is the VDA isolation valve state (open), it can be determined that the current initial fault is: the SG3 VDA isolation valve accidentally opened; if the local parameter monitoring at this time is the temperature increase of the VVP3 valve compartment, it can be determined that the current initial fault is: the pipeline between the SG3 VVP valves is broken.

[0092] In the present invention, the PCA model monitoring only acts on the global variables, the logical quantities are set as local variables, and the anomalies are directly read from the power plant control system. At the same time, the local variables are included in the SDG diagram, so that the purpose of directly diagnosing the faulty equipment is achieved.

[0093] refer to Figure 3 , Figure 3 This is a principle block diagram of the nuclear power plant unit state abnormality detection and diagnosis system provided by the present invention.

[0094] Specifically, Figure 3 As shown, the abnormal state detection and diagnosis system of the nuclear power plant unit includes:

[0095] The historical parameter acquisition unit 301 is used to acquire the historical global parameters of the nuclear power plant.

[0096] The PCA model building unit 302 is used to build a PCA model under normal operating conditions based on historical global parameters.

[0097] The global parameter monitoring unit 303 is used to monitor the real-time global parameters of the nuclear power plant based on the PCA model, and determine whether the real-time global parameters deviate from normal values.

[0098] The abnormality detection and diagnosis unit 304 is used to perform abnormality detection and diagnosis on the nuclear power plant units by using a global-local parameter combined signed directed graph.

[0099] Specifically, the specific coordination operation process between the various units in the nuclear power plant unit state abnormality detection and diagnosis system here can refer to the above-mentioned nuclear power plant unit state abnormality detection and diagnosis method, which will not be repeated here.

[0100] The present invention uses a more optimized unit abnormality monitoring and diagnosis method to monitor and analyze the unit status, avoiding setting the envelope alarm value through a large amount of analysis work, and can detect the unit status deviation earlier, prevent the occurrence of large transients, diagnose faults early, and avoid shutdowns. The fault diagnosis method proposed by the present invention is convenient for operation and maintenance personnel to quickly understand, trace and confirm.

[0101] In addition, an electronic device of the present invention includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program to implement any of the above nuclear power plant unit state abnormality detection and diagnosis methods. Specifically, according to an embodiment of the present invention, the process described with reference to the flowchart above can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed by an electronic device and executed to perform the above functions defined in the method of the embodiment of the present invention. The electronic device in the present invention can be a terminal such as a notebook, a desktop, a tablet computer, a smart phone, or a server.

[0102] In addition, a storage medium of the present invention stores a computer program thereon, and when the computer program is executed by a processor, any one of the above-mentioned methods for detecting and diagnosing abnormal conditions of a nuclear power plant unit is implemented. Specifically, it should be noted that the above-mentioned storage medium of the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0103] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0104] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0105] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0106] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0107] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot limit the scope of protection of the present invention. All equivalent changes and modifications made to 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 detecting and diagnosing abnormal conditions of a nuclear power plant unit, characterized in that: The following steps are involved: Obtain historical global parameters of nuclear power plants; Building a PCA model under normal operating conditions based on the historical global parameters; Monitoring the real-time global parameters of the nuclear power plant based on the PCA model, and determining whether the real-time global parameters deviate from normal values; If so, a global-local parameter combined signed directed graph is used to detect and diagnose anomalies in nuclear power plant units; If not, continue to monitor the real-time global parameters of the nuclear power plant based on the PCA model.

2. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 1, characterized in that: The acquisition of historical global parameters of the nuclear power plant includes: Determine the parameters reflecting the reactivity control of nuclear power plants; Determine the parameters reflecting the removal of waste heat from nuclear power plants; Determine the parameters reflecting the radioactivity of nuclear power plants; Determine the influencing parameters of the unit transient caused by abnormalities; The historical global parameter is acquired according to the parameter reflecting the reactivity control of the nuclear power plant, the parameter reflecting the residual heat extraction of the nuclear power plant, the parameter reflecting the radioactivity of the nuclear power plant and the influencing parameter.

3. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 1, characterized in that: The PCA model under normal operation state is constructed based on the historical global parameters, including: Based on the historical global parameters, determining global parameter samples of multiple units under normal operating conditions; The PCA model is constructed based on the global parameter sample.

4. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 3, characterized in that: The constructing the PCA model based on the global parameter sample comprises: Constructing a sample matrix based on the global parameter samples; Preprocessing the data of the sample matrix to obtain preprocessed data; Calculation is performed based on the preprocessed data to obtain the PCA model.

5. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 4, characterized in that: The preprocessing of the data of the sample matrix to obtain preprocessed data comprises: The data of the sample matrix is ​​subjected to mean standardization and normalization processing to obtain the preprocessed data.

6. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 5, characterized in that: The performing calculation based on the preprocessed data to obtain the PCA model comprises: Perform calculation based on the preprocessed data to obtain a covariance matrix; Calculating eigenvalues ​​and eigenvectors of the covariance matrix to obtain eigenvalues ​​and eigenvectors of the covariance matrix; Processing is performed according to the eigenvalues ​​and eigenvectors of the covariance matrix to obtain a coding matrix; Reconstructing based on the sample matrix and the encoding matrix to obtain a reconstructed data matrix; Perform residual calculation based on the reconstructed data matrix to obtain a residual matrix; Calculate statistics according to the residual matrix to obtain statistics; The statistical threshold is set to complete the construction of the PCA model.

7. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 6, characterized in that: The monitoring of the real-time global parameters of the nuclear power plant based on the PCA model and judging whether the real-time global parameters deviate from normal values ​​includes: Performing real-time monitoring on the nuclear power plant units to obtain real-time global parameters of the nuclear power plant; Normalizing the real-time global parameters based on the mean and variance used in the normalization calculation of the historical global parameters to obtain a normalized data matrix; Reconstructing the normalized data matrix based on the encoding matrix to obtain a reconstructed matrix; Calculating residual statistics between vectors of the normalized data matrix and vectors of the reconstruction matrix; Whether the real-time global parameter deviates from a normal value is determined according to the residual statistic and the statistic threshold.

8. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 7, characterized in that: The determining whether the real-time global parameter deviates from a normal value according to the residual statistic and the statistic threshold comprises: comparing the residual statistic with the statistic threshold; If the residual statistic is greater than the statistic threshold, it is determined that the currently acquired real-time global parameter is abnormal data; Calculating the residual contribution rate of each real-time global parameter in the real-time global parameters; According to whether the residual contribution rate of each real-time global parameter is greater than a set threshold; If so, it is determined that the real-time global parameter deviates from the normal value; If not, it is determined that the real-time global parameter has no abnormality.

9. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 6, characterized in that: The setting of the statistical threshold comprises: The percentile method is used to set the statistical threshold.

10. The method for detecting and diagnosing abnormal conditions of nuclear power plant units according to claim 1, characterized in that: The method of using a global-local parameter combined signed directed graph to detect and diagnose abnormalities in a nuclear power plant unit includes: Based on the real-time global parameters that deviate from normal values, a global-local parameter symbolic directed graph is drawn; Abnormal detection and diagnosis of nuclear power plant units are performed according to the global-local parameter symbol directed graph.

11. A nuclear power plant unit state abnormality detection and diagnosis system, characterized in that: include: A historical parameter acquisition unit, used to acquire historical global parameters of the nuclear power plant; A PCA model building unit, used for building a PCA model under normal operating conditions based on the historical global parameters; A global parameter monitoring unit, used to monitor the real-time global parameters of the nuclear power plant based on the PCA model, and determine whether the real-time global parameters deviate from normal values; The anomaly detection and diagnosis unit is used to detect and diagnose anomalies of nuclear power plant units using a symbolic directed graph combining global and local parameters.

12. A storage medium, characterized in that: The storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the steps of the method for detecting and diagnosing abnormal status of a nuclear power plant unit as described in any one of claims 1 to 10.

13. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the steps of the method for detecting and diagnosing abnormal status of a nuclear power plant unit as claimed in any one of claims 1 to 10 by calling the computer program stored in the memory.

Images