Method and device for identifying abnormal running trend of nuclear power equipment, equipment and medium

Abstract

The invention is suitable for the technical field of nuclear power equipment fault analysis, and particularly relates to a nuclear power equipment operation abnormal trend identification method and device, equipment and a medium. The method comprises the following steps: acquiring a first vibration spectrum in a current time period and a second vibration spectrum acquired in a first historical time period by using a vibration sensor arranged at a preset position of nuclear power equipment, determining a first energy difference between the first vibration spectrum and the second vibration spectrum, analyzing the second vibration spectrum, and determining a dynamic difference threshold value; when the first energy difference is greater than a dynamic difference threshold value, performing frequency division on the first vibration frequency spectrum to obtain a first frequency band frequency spectrum of M frequency bands, and calculating a second energy difference with a corresponding second frequency band frequency spectrum extracted from the second vibration frequency spectrum, and determining that the frequency spectrum of the frequency band of which the second energy difference is greater than the dynamic difference threshold value in the first vibration frequency spectrum is the abnormal frequency spectrum, and more accurately describing the trend change based on the comparison of the energy difference and the dynamic difference threshold value, thereby accurately determining the abnormal identification result.

Description

Technical Field

[0001] This application is applicable to the technical field of nuclear power equipment fault analysis, and particularly relates to a method, device, equipment and medium for identifying abnormal operation trends of nuclear power equipment. Background Art

[0002] Identifying abnormal trends in the vibration signal spectrum plays a crucial role in fault diagnosis, especially in the environment of nuclear power equipment. It can not only help engineers locate the fault source in a timely and accurate manner but also effectively predict the operating state of the equipment, thereby avoiding potential safety risks and production interruptions.

[0003] Generally, the normal vibration spectrum pattern of the equipment should be relatively stable, and the amplitudes of each frequency component fluctuate within a certain range. If significant changes occur in the spectrum pattern, such as sudden increases or decreases in the amplitudes of certain frequency components, it may indicate that the equipment is abnormal. Additionally, in the early stage of equipment faults, the fault signals generated during equipment operation are very weak, and even if there are no obvious changes in the spectrum pattern, equipment faults may already have occurred.

[0004] Currently, there is a lack of effective and reliable automated methods for identifying and detecting abnormal spectrum components. This process is mostly carried out manually by engineers. Limited by individual experience differences among engineers, there is also a lack of a unified standard for the relationship between the degree of amplitude change of frequency components and the severity of faults.

[0005] Therefore, how to automatically analyze the characteristics of the vibration signal spectrum of nuclear power equipment to accurately identify abnormal trends and avoid potential risks has become an urgent problem to be solved. Summary of the Invention

[0006] In view of this, embodiments of this application provide a method, device, equipment and medium for identifying abnormal operation trends of nuclear power equipment to solve the problem of how to automatically analyze the characteristics of the vibration signal spectrum of nuclear power equipment to accurately identify abnormal trends and avoid potential risks.

[0007] Use a vibration sensor provided at a preset position of the nuclear power equipment to collect a first vibration spectrum in the current time period, obtain a second vibration spectrum collected by the vibration sensor in the first historical time period, and determine a first energy difference between the first vibration spectrum and the second vibration spectrum, where the durations of the current time period and the first historical time period are the same;

[0008] Extract the second vibration spectrum using a sliding window with variable width to obtain N historical sub - spectra with time sequence, determine the initial energy of each historical sub - spectrum, perform normalization processing on the initial energy value to obtain the standard energy of each historical sub - spectrum, and calculate the corresponding standard energy mean according to the standard energy of each historical sub - spectrum, where N is an integer greater than zero;

[0009] Perform time - series decomposition on each historical sub - spectrum to obtain the trend sub - spectrum and residual sub - spectrum of the corresponding historical sub - spectrum, determine the trend energy mean of all trend sub - spectra and the residual energy standard deviation, determine the reference threshold according to the standard energy mean and the trend energy mean, and determine the reference adjustment amount according to the residual energy standard deviation;

[0010] Determine the dynamic difference threshold according to the reference threshold and the reference adjustment amount. When the first energy difference is greater than the dynamic difference threshold, perform frequency - division extraction on the first vibration spectrum to obtain the first - band spectra of M frequency bands, where M is an integer greater than zero;

[0011] For any frequency band, extract the corresponding second - band spectrum from the second vibration spectrum, determine the second energy difference between the first - band spectrum corresponding to the frequency band and the corresponding second - band spectrum. When the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum.

[0012] In a second aspect, an embodiment of the present application provides an abnormal operation trend recognition device for nuclear power equipment, and the abnormal operation trend recognition device includes:

[0013] A current data analysis module, configured to collect the first vibration spectrum of the current time period using a vibration sensor set at a preset position of the nuclear power equipment, obtain the second vibration spectrum collected by the vibration sensor in the first historical time period, and determine the first energy difference between the first vibration spectrum and the second vibration spectrum, where the duration of the current time period is the same as that of the first historical time period;

[0014] A historical data analysis module, configured to extract the second vibration spectrum using a sliding window with variable width to obtain N historical sub - spectra with time sequence, determine the initial energy of each historical sub - spectrum, perform normalization processing on the initial energy value to obtain the standard energy of each historical sub - spectrum, and calculate the corresponding standard energy mean according to the standard energy of each historical sub - spectrum, where N is an integer greater than zero;

[0015] A dynamic difference threshold analysis module, which is used to perform time series decomposition on each historical sub-spectrum to obtain a trend sub-spectrum and a residual sub-spectrum corresponding to the historical sub-spectrum, determine the trend energy mean value of all trend sub-spectrums and the residual energy standard deviation, determine a reference threshold according to the standard energy mean value and the trend energy mean value, and determine a reference adjustment amount according to the residual energy standard deviation;

[0016] A frequency division extraction module, which is used to determine a dynamic difference threshold according to the reference threshold and the reference adjustment amount, and perform frequency division extraction on the first vibration spectrum when the first energy difference is greater than the dynamic difference threshold to obtain first frequency band spectra in M frequency bands, where M is an integer greater than zero;

[0017] An abnormal spectrum determination module, which is used to extract a corresponding second frequency band spectrum from the second vibration spectrum for any frequency band, determine a second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum when the second energy difference is greater than the dynamic difference threshold.

[0018] In a third aspect, an embodiment of the present application provides a computer device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the running abnormal trend recognition method described in the first aspect is implemented.

[0019] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the running abnormal trend recognition method described in the first aspect is implemented.

[0020] The beneficial effects of the embodiments of the present application compared with the prior art are as follows: The present application uses vibration sensors set at preset positions in nuclear power equipment to collect the first vibration spectrum in the current time period, obtains the second vibration spectrum collected by the vibration sensors in the first historical time period, determines the first energy difference between the first vibration spectrum and the second vibration spectrum, uses a sliding window with variable width to extract the second vibration spectrum, obtains N historical sub-spectrums with time sequence, determines the initial energy of each historical sub-spectrum, performs normalization processing on the initial energy values to obtain the standard energy of each historical sub-spectrum, calculates the corresponding standard energy mean according to the standard energy of each historical sub-spectrum, performs time series decomposition on each historical sub-spectrum to obtain the trend sub-spectrum and residual sub-spectrum of the corresponding historical sub-spectrum, determines the trend energy mean and residual energy standard deviation of all trend sub-spectrums, determines the benchmark threshold according to the standard energy mean and the trend energy mean, determines the benchmark adjustment amount according to the residual energy standard deviation, determines the dynamic difference threshold according to the benchmark threshold and the benchmark adjustment amount, when the first energy difference is greater than the dynamic difference threshold, performs frequency division extraction on the first vibration spectrum to obtain the first frequency band spectrum of M frequency bands, for any frequency band, extracts the corresponding second frequency band spectrum from the second vibration spectrum, determines the second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determines that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum. Through the analysis of the energy difference of the vibration spectrum and the determination of the dynamic difference threshold through the analysis of the historical vibration spectrum, the comparison between the energy difference and the dynamic difference threshold more accurately describes the trend change, so as to accurately determine the abnormality, realizing automatic analysis and accurate abnormal trend recognition. Brief Description of the Drawings

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings.

[0022] Figure 1 It is a schematic diagram of an application environment of a method for identifying abnormal operation trends of nuclear power equipment provided in Embodiment 1 of the present application;

[0023] Figure 2 It is a schematic flowchart of a method for identifying abnormal operation trends of nuclear power equipment provided in Embodiment 2 of the present application;

[0024] Figure 3 It is a schematic flowchart of a method for identifying abnormal operation trends of nuclear power equipment provided in Embodiment 3 of the present application;

[0025] Figure 4 It is a schematic flowchart of a method for identifying abnormal operation trends of nuclear power equipment provided in Embodiment 4 of the present application;

[0026] Figure 5 It is a schematic structural diagram of a device for identifying abnormal operation trends of nuclear power equipment provided in Embodiment 5 of the present application;

[0027] Figure 6 It is a schematic structural diagram of a computer device provided in Embodiment 6 of the present application. Detailed implementation manners

[0028] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0029] It should be understood that when used in the specification of the present application and the appended claims, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0030] It should also be understood that the term "and / or" used in the specification of the present application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0031] As used in the specification of the present application and the appended claims, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if detecting [the described condition or event]" can be interpreted as meaning "once determined", "in response to determining", "once detecting [the described condition or event]", or "in response to detecting [the described condition or event]" according to the context.

[0032] In addition, in the description of the specification of the present application and the appended claims, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0033] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized.

[0034] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence is the theory, method, technology and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results.

[0035] Artificial intelligence basic technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, robotics, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0036] It should be understood that the magnitudes of the sequence numbers of the steps in the following embodiments do not mean the order of execution is prior or subsequent. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0037] In order to illustrate the technical solution of this application, the following will be described through specific embodiments.

[0038] An abnormal operation trend recognition method for nuclear power equipment provided in the first embodiment of this application can be applied in an application environment such as Figure 1 where the server is connected to a corresponding database, etc., and the server executes a corresponding method to analyze the data stored in the database. Of course, the server can also be directly connected to the vibration sensors set on the nuclear power equipment to obtain the corresponding collected data. Among them, the server can include but is not limited to computer devices such as a palm computer, a desktop computer, a notebook computer, an ultra-mobile personal computer (UMPC), a netbook, a cloud computer device, a personal digital assistant (PDA), etc., and can also be implemented by an independent server.

[0039] The nuclear power equipment in this application includes, but is not limited to, rotating equipment, etc. For example, a steam turbine. By setting the corresponding vibration sensor at a preset position of the equipment, the vibration condition at this position can be collected, thereby forming a vibration spectrum.

[0040] See Figure 2 , which is a schematic flowchart of a method for identifying an abnormal operation trend of nuclear power equipment provided in the second embodiment of this application. The above method for identifying an abnormal operation trend is applied to Figure 1 the server in, and the server is associated with a corresponding database to obtain data. As Figure 2 shown, the method for identifying an abnormal operation trend of the nuclear power equipment may include the following steps:

[0041] Step S201: Use the vibration sensor set at the preset position of the nuclear power equipment to collect the first vibration spectrum in the current time period, obtain the second vibration spectrum collected by the vibration sensor in the first historical time period, and determine the first energy difference between the first vibration spectrum and the second vibration spectrum.

[0042] Among them, the current time period is a time period after the first historical time period. The two can be two consecutive time periods or there can be a certain time interval between them. The duration of the current time period is the same as that of the first historical time period.

[0043] To ensure the accuracy and consistency of energy calculation, the vibration spectrum can first be preprocessed. Among them, filtering and denoising is to perform band-pass filtering on the signal to remove high-frequency and low-frequency noise interference signals in the signal, so that the signal energy is concentrated in the target frequency range; normalization processing is to assume that the historical spectrum (i.e., the second vibration spectrum) and the current spectrum (i.e., the first vibration spectrum) are S history and S current , respectively perform normalization processing on S history and S current . The normalization formula is as follows:

[0044] S normallized =(S - min(S)) / (max(S) - min(S));

[0045] In the formula, S normallized is the normalization result, S represents any vibration spectrum, min(S) represents the minimum value in the corresponding vibration spectrum, and max(S) represents the maximum value in the corresponding vibration spectrum.

[0046] Energy difference calculation requires calculating the energy of each vibration spectrum. Energy calculation is a key step in spectrum analysis and is calculated by accumulating the power spectral density (PSD) of the signal in each frequency band.

[0047] First, perform spectrum segmentation and integration, dividing the spectrum into n frequency bands:

[0048]

[0049] Among them, f i represents the starting frequency of the i-th frequency band, and f i+1 represents the ending frequency of the i-th frequency band. E i represents the energy of the i-th frequency band. S(f) represents the power spectral density at frequency f, where frequency f is a frequency value in any frequency band.

[0050] Secondly, calculate the total energy of the historical spectrum and the current spectrum. Let the total energies of the historical spectrum and the current spectrum be E history_total and E current_total .

[0051] Here, both the historical spectrum and the current spectrum are expressed using the total energy E total . The calculation formula for this E total is as follows:

[0052]

[0053] Among them, n is the number of frequency segments, and each E i is the energy calculated within the corresponding frequency band.

[0054] Finally, after obtaining the total energies of the historical spectrum and the current spectrum, calculate the energy difference by taking the difference. This difference is used for subsequent anomaly detection.

[0055] The calculation formula for the energy difference is as follows:

[0056] ΔE = |E history_total - E current_total |

[0057] Among them, ΔE represents the energy difference. When this value can be compared with a designed dynamic difference threshold, the comparison result can be used to determine whether there is an anomaly in the current spectrum.

[0058] Step S202: Extract the second vibration spectrum using a sliding window with variable width to obtain N historical sub-spectra with time order. Determine the initial energy of each historical sub-spectrum, perform normalization processing on the initial energy value to obtain the standard energy of each historical sub-spectrum, and calculate the corresponding standard energy mean based on the standard energy of each historical sub-spectrum.

[0059] Wherein, N is an integer greater than zero. The variable width may mean that the window width of the sliding window can be adjusted according to requirements. Specifically, the variable width can be widened or narrowed according to the corresponding feedback information. If the window is wider, the adjustment of the subsequent dynamic difference threshold will be smoother. If the window is narrower, the dynamic difference threshold will be more sensitive to data fluctuations.

[0060] By analyzing historical spectrum data, adaptively calculate the dynamic difference threshold for marking anomalies when the spectrum energy difference exceeds the threshold. The setting of the dynamic difference threshold is mainly based on the statistical characteristics of historical spectrum data, such as mean, standard deviation, etc. Dynamically update the threshold using the fluctuation range of historical data to adapt to short-term changes and sudden anomalies in the spectrum. Through this adaptive mechanism, the system can avoid frequent false alarms and missed detections, and improve the robustness and sensitivity of detection.

[0061] Using a sliding window to extract historical spectra can obtain historical sub-spectra within the corresponding window. Each historical sub-spectrum corresponds to its own energy data. Constructing time series data in the order of the sliding of the sliding window can be expressed as: E history ={E1, E2,..., E i}(i ≤ N), where E i is the total energy value (i.e., the initial energy) of the historical sub-spectrum intercepted by the corresponding sliding window for the t i time period.

[0062] To ensure the comparability of energy data in each period, standardize the initial energy within the sliding window. After standardization, the standard energies of each historical sub-spectrum are comparable. Calculate the corresponding standard energy mean based on the standard energies of all historical sub-spectra. The standard energy mean, as part of the reference threshold in the dynamic difference threshold, can effectively ensure the accuracy of the threshold reference.

[0063] Optionally, perform standardization processing on the initial energy value to obtain the standard energy of each historical sub-spectrum, including:

[0064] Determine the historical energy mean and historical energy standard deviation of the corresponding second vibration spectrum according to the initial energies of all historical sub-spectra;

[0065] For any historical sub-spectrum, subtract the historical energy mean from the initial energy value corresponding to the historical sub-spectrum to obtain a difference result. Divide the difference result by the historical energy standard deviation to obtain the standard energy of the corresponding historical sub-spectrum. Traverse all historical sub-spectra to obtain the standard energy of each historical sub-spectrum.

[0066] Wherein, according to the previous time series data, calculate the historical energy mean and historical energy standard deviation of the second vibration spectrum. The standardization formula is as follows:

[0067] E i '=(E i -μ history ) / σ history

[0068] wherein, μ history is the mean of historical data and σ history is the standard deviation of historical data.

[0069] In addition, the calculation formulas for the standard deviation and the mean are as follows:

[0070]

[0071] wherein, μ is the mean, σ is the standard deviation, and E i is the standard energy corresponding to the i-th historical sub-spectrum. Of course, for the historical energy mean and the historical energy standard deviation, the data object is the energy in the historical spectrum.

[0072] Step S203: Perform time series decomposition on each historical sub-spectrum to obtain the trend sub-spectrum and the residual sub-spectrum corresponding to the historical sub-spectrum, determine the trend energy mean of all trend sub-spectra and the residual energy standard deviation, determine the reference threshold according to the standard energy mean and the trend energy mean, and determine the reference adjustment amount according to the residual energy standard deviation.

[0073] Among them, for the vibration spectrum, it includes a trend part, a seasonal part, and a residual part. These three parts constitute all the characteristics of the vibration spectrum. In the embodiments of the present application, in order to implement anomaly analysis, the seasonal part is not processed. Therefore, time series decomposition is used to obtain the trend part and the residual part, which respectively correspond to the trend sub-spectrum and the residual sub-spectrum. Time series decomposition is performed on each historical sub-spectrum to obtain all the trend sub-spectra and residual sub-spectra.

[0074] Decompose the historical sub-spectrum into a trend T, a seasonality S, and a residual R through time series decomposition. The corresponding energy representation is E i =T i +S i +R i , where T i is the trend energy, S i is the seasonal energy, and R i is the residual energy. Here, retain the mean of the trend part T i as the reference for the dynamic difference threshold, and calculate the adjustment amount of the dynamic difference threshold based on the fluctuation of the residual part R i .

[0075] Step S204: Determine the dynamic difference threshold according to the reference threshold and the reference adjustment amount. When the first energy difference is greater than the dynamic difference threshold, perform frequency division extraction on the first vibration spectrum to obtain the first band spectra of M bands.

[0076] Where M is an integer greater than zero.

[0077] The dynamic difference threshold can be adjusted according to the transformation of the second vibration spectrum. At each abnormal trend recognition time point, as the second vibration spectrum changes, the corresponding dynamic difference threshold changes, making the comparison between the new dynamic difference threshold and the newly collected first energy difference more accurate and more in line with the actual situation, avoiding inaccurate judgment caused by using artificially defined thresholds for abnormal judgment.

[0078] The dynamic difference threshold can be obtained by adding the reference threshold and the reference adjustment amount. Compare the first energy difference with this dynamic difference threshold. If the first energy difference is not greater than this dynamic difference threshold, it indicates that the difference between the first vibration spectrum and the second vibration spectrum is small and there is no significant change, and it can be considered that the first vibration spectrum has not shown abnormal conditions. If the first energy difference is greater than this dynamic difference threshold, it indicates that an abnormal situation may have occurred, and further analysis of this first vibration spectrum is required to identify the abnormal position, etc.

[0079] By dividing the frequency bands, the first vibration spectrum can be extracted in each band to obtain the band spectra of the corresponding bands, which is used to locate the abnormal bands and abnormal spectra.

[0080] If an abnormality is detected, record the current timestamp, energy difference value, and spectrum data for subsequent analysis. These abnormal information can be used to generate alarm information or trigger further band analysis.

[0081] Optionally, determine the reference threshold according to the standard energy mean and the trend energy mean, and determine the reference adjustment amount according to the residual energy standard deviation, including:

[0082] Obtain the updated weight coefficient, and use the updated weight coefficient to perform weighted summation on the standard energy mean and the trend energy mean to determine that the weighted summation result is the reference threshold;

[0083] Obtain the updated sensitivity coefficient, and multiply the updated sensitivity coefficient by the residual energy standard deviation to determine that the formed result is the reference adjustment amount;

[0084] Determine the dynamic difference threshold according to the reference threshold and the reference adjustment amount, including:

[0085] Add the reference threshold and the reference adjustment amount to determine that the added result is the dynamic difference threshold.

[0086] Among them, the standard energy mean value and the trend energy mean value calculated by the sliding window are combined to obtain a comprehensive reference threshold Base Threshold, as follows:

[0087] Base Threshold = α·μ W +(1 - α)·T

[0088] Among them, α is a weight factor used to adjust the balance between the sliding window mean value and the long-term trend, μ W is the standard energy mean value, and T is the trend energy mean value.

[0089] The final dynamic difference threshold is based on the adjustment amount DynamicThreshold composed of the reference threshold and the standard deviation of the residual energy, as follows:

[0090] Dynamic Threshold = Base Threshold + k·σ R

[0091] In the formula, σ R is the standard deviation of the residual energy, and k is the sensitivity coefficient used to control the sensitivity of the threshold.

[0092] Optionally, before determining the reference threshold according to the standard energy mean value and the trend energy mean value and determining the reference adjustment amount according to the standard deviation of the residual energy, it further includes:

[0093] After the last recognition is completed, obtain the false alarm rate and the missed alarm rate, and adjust the current sensitivity coefficient according to the false alarm rate and the missed alarm rate to obtain an updated sensitivity coefficient;

[0094] Obtain the historical vibration spectrum collected by the vibration sensor in the second historical time period, analyze the historical vibration spectrum, determine the trend value in the second historical time period, and adjust the current weight coefficient according to the trend value to obtain an updated weight coefficient, where the second historical time period is greater than the first historical time period.

[0095] Among them, the sensitivity coefficient and the weight coefficient are variable, and the sensitivity coefficient k is dynamically adjusted according to the false alarm rate and the missed alarm rate of the detection result. When the false alarm rate is too high, the system will increase the value of k, thereby expanding the threshold range; when the missed alarm rate is too high, the system will decrease the value of k to make the threshold more stringent. If it is detected that the spectral environment has changed significantly (such as a long-term trend rising or falling), the value of the weight factor α can be adjusted to make the threshold more sensitive to the long-term trend.

[0096] Through the above update mechanism, the dynamic difference threshold will be adjusted in each detection cycle, and the new adjustment amount Dynamic Threshold new , as follows:

[0097] Dynamic Threshold new =(α new ·μ W +(1-α new )·T)+k new ·σ R

[0098] Among them, α new is the updated weight coefficient, k new is the updated sensitivity coefficient.

[0099] Optionally, after determining the dynamic difference threshold according to the benchmark threshold and the benchmark adjustment amount, the method further includes:

[0100] If the first energy difference is greater than the dynamic difference threshold, the variable width is increased to obtain an increased variable width;

[0101] When entering the next time period to identify abnormal operation trends, the current time period is used as the first historical time period, the next time period is used as the current time period, the increased variable width is used as the variable width, and the execution is returned to use the vibration sensor set at the preset position of the nuclear power equipment to collect the first vibration spectrum of the current time period until the identification is completed.

[0102] Among them, the mean and standard deviation are updated regularly according to the vibration spectrum of the historical time period. If the fluctuation of the energy difference is detected to exceed the threshold, the sliding window width W can be increased to smooth the short-term fluctuation. The width and update frequency of the sliding window can be dynamically adjusted according to the detection requirements.

[0103] Step S205, for any frequency band, extract the corresponding second frequency band spectrum from the second vibration spectrum, determine the second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum.

[0104] Among them, for any frequency band, the spectrum of the second vibration spectrum is extracted to obtain the second frequency band spectrum of the corresponding frequency band, and the second energy difference is obtained by comparing the first frequency band spectrum and the second frequency band spectrum of each frequency band. If the energy difference is greater than the dynamic difference threshold, the frequency band can be located as an abnormal frequency band. Of course, the spectrum extracted from the corresponding frequency band is the abnormal spectrum.

[0105] In an embodiment of the present application, a vibration sensor provided by a nuclear power device at a preset position is used to collect a first vibration spectrum in the current time period, a second vibration spectrum collected by the vibration sensor in a first historical time period is obtained, a first energy difference between the first vibration spectrum and the second vibration spectrum is determined, the second vibration spectrum is extracted using a sliding window with a variable width to obtain N historical sub-spectrums with a time sequence, an initial energy of each historical sub-spectrum is determined, the initial energy value is normalized to obtain a standard energy of each historical sub-spectrum, a corresponding standard energy mean value is calculated according to the standard energy of each historical sub-spectrum, each historical sub-spectrum is decomposed by time series to obtain a trend sub-spectrum and a residual sub-spectrum corresponding to the historical sub-spectrum, a trend energy mean value and a residual energy standard deviation of all trend sub-spectrums are determined, a reference threshold is determined according to the standard energy mean value and the trend energy mean value, a reference adjustment amount is determined according to the residual energy standard deviation, a dynamic difference threshold is determined according to the reference threshold and the reference adjustment amount, when the first energy difference is greater than the dynamic difference threshold, the first vibration spectrum is frequency-divided and extracted to obtain first frequency-band spectrums of M frequency bands, for any frequency band, a corresponding second frequency-band spectrum is extracted from the second vibration spectrum, a second energy difference between the first frequency-band spectrum corresponding to the frequency band and the corresponding second frequency-band spectrum is determined, and when the second energy difference is greater than the dynamic difference threshold, it is determined that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum. Through the analysis of the energy difference of the vibration spectrum and the determination of the dynamic difference threshold through the analysis of the historical vibration spectrum, the comparison between the energy difference and the dynamic difference threshold can more accurately describe the trend change, so as to accurately determine the abnormality, realizing automatic analysis and accurate abnormal trend recognition.

[0106] See Figure 3 , which is a schematic flowchart of a method for identifying an abnormal operation trend of a nuclear power device provided in Embodiment 3 of the present application. As Figure 3 shown, after frequency-dividing and extracting the first vibration spectrum in the above step S204 to obtain first frequency-band spectrums of M frequency bands, the following steps may be included:

[0107] Step S301: Use the encoder in the trained anomaly detection model to convert each first frequency-band spectrum to obtain a corresponding feature vector.

[0108] Among them, the anomaly detection model may be a neural network model composed of an encoder and a decoder. For example, a convolutional network model, etc. The encoder in the anomaly detection model may use an adversarial autoencoder, which can encode the spectrum data well to obtain more accurate features. Inputting the first frequency-band spectrum into the encoder can learn the feature vector representing the key information in the frequency band, which can be used as the input for subsequent anomaly detection.

[0109] The training of the anomaly detection model can be carried out using existing labeled spectrum data, so as to accurately perform feature vector transformation on the spectrum.

[0110] Step S302: Use the decoder in the trained anomaly detection model to reconstruct the feature vector corresponding to each first-band spectrum to obtain the corresponding reconstructed spectrum.

[0111] Among them, the decoder is trained simultaneously with the encoder, and the overall training objective is to use the encoder to learn the normal spectrum distribution and detect abnormal samples that deviate from this distribution. The training steps are as follows:

[0112] Train the encoder and decoder: Given normal spectrum data, train the encoder-decoder structure so that for the input frequency band F i , the encoder encodes it as: z i = Encoder(F i ), where z i is the encoding result, and the decoder reconstructs it as F i ' = Decoder(z i ), where F i ' is the reconstruction result.

[0113] Optimize the model by minimizing the reconstruction loss L recon = ||F i - F i '|| 2 to make the reconstruction error for the normal frequency band the lowest. For new spectrum data, if the reconstruction error of a certain frequency band exceeds the set threshold, it is considered that there is an anomaly in this frequency band.

[0114] Step S303: For any first-band spectrum, calculate the error between the first-band spectrum and the corresponding reconstructed spectrum of the first-band spectrum to obtain the reconstruction error. If the reconstruction error is greater than the reconstruction error threshold, determine that the frequency band corresponding to the first-band spectrum is an abnormal frequency band.

[0115] Among them, based on the reconstruction error of the frequency band corresponding to the frequency spectrum of the frequency band, the abnormal frequency band is determined, so as to analyze the abnormal frequency band in subsequent analysis, which can effectively reduce the amount of data for analysis and improve the analysis efficiency.

[0116] For any frequency band in step S204 above, extract the corresponding second-band spectrum from the second vibration spectrum, determine the second energy difference between the first-band spectrum corresponding to the frequency band and the corresponding second-band spectrum. When the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum, specifically including:

[0117] Step S304: For any abnormal frequency band, extract the corresponding second - band spectrum from the second vibration spectrum, determine the second energy difference between the first - band spectrum corresponding to the abnormal frequency band and the corresponding second - band spectrum. When the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum.

[0118] In the embodiments of the present application, an abnormal detection model can be trained by means of artificial intelligence, and the data in the divided frequency bands can be identified based on the trained abnormal detection model, so as to accurately obtain the abnormal frequency bands, facilitating subsequent identification preferentially from the abnormal frequency bands during abnormal positioning and improving the overall identification efficiency of the method.

[0119] See Figure 4 , which is a schematic flowchart of a method for identifying the abnormal operation trend of a nuclear power equipment provided in the fourth embodiment of the present application. As Figure 4 shown, after determining that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum in the above - mentioned step S304, the following steps can be included:

[0120] Step S401: Obtain the abnormal center frequency and abnormal bandwidth of the abnormal spectrum, as well as the historical center frequency and historical bandwidth of the abnormal spectrum in the second historical time period.

[0121] Step S402: Compare the abnormal center frequency and abnormal bandwidth with the historical center frequency and historical bandwidth respectively to obtain the comparison result. According to the comparison result, conduct a reliability analysis on the abnormal spectrum, and determine the abnormal spectrum that meets the reliability analysis conditions as the final abnormal spectrum.

[0122] Among them, after determining the abnormal spectrum, information such as the center frequency, bandwidth, and corresponding energy difference of the abnormal spectrum can be output and recorded for subsequent analysis and processing.

[0123] The above - recorded data can be sent to the user for processing. Before that, the abnormal spectrum is verified through this part of the data, thereby avoiding the occurrence of unreliable abnormalities and being able to reduce the false alarm rate.

[0124] If it is determined as the final abnormal spectrum, information such as the detection time, frequency range, and energy change of all abnormal frequency bands is recorded for subsequent backtracking and analysis. An alarm can also be generated, an automatic adjustment strategy can be triggered, or further in - depth analysis can be carried out.

[0125] In addition, the comparison in the above - mentioned step S402 can specifically adopt similarity comparison, and the reliability analysis can adopt a method of comparing the degree of deviation from the associated spectrum. If there is a large deviation, the reliability analysis conditions are not met, and only when there is no large deviation is it determined that the reliability analysis conditions are met.

[0126] The method for identifying the abnormal operation trend of nuclear power equipment corresponding to the above embodiments Figure 5 shows the structural block diagram of the device for identifying the abnormal operation trend of nuclear power equipment provided in the fifth embodiment of the present application. The above device for identifying the abnormal operation trend is applied to Figure 1 the server in. For the sake of convenience of description, only the parts related to the embodiments of the present application are shown.

[0127] See Figure 5 , the device for identifying the abnormal operation trend includes:

[0128] The current data analysis module 51 is used to collect the first vibration spectrum in the current time period by using the vibration sensor set at the preset position of the nuclear power equipment, obtain the second vibration spectrum collected by the vibration sensor in the first historical time period, and determine the first energy difference between the first vibration spectrum and the second vibration spectrum, where the duration of the current time period is the same as that of the first historical time period;

[0129] The historical data analysis module 52 is used to extract the second vibration spectrum by using a sliding window with variable width to obtain N historical sub-spectrums with time sequence, determine the initial energy of each historical sub-spectrum, perform normalization processing on the initial energy value to obtain the standard energy of each historical sub-spectrum, and calculate the corresponding standard energy mean according to the standard energy of each historical sub-spectrum, where N is an integer greater than zero;

[0130] The dynamic difference threshold analysis module 53 is used to perform time series decomposition on each historical sub-spectrum to obtain the trend sub-spectrum and the residual sub-spectrum corresponding to the historical sub-spectrum, determine the trend energy mean and the residual energy standard deviation of all trend sub-spectrums, determine the reference threshold according to the standard energy mean and the trend energy mean, and determine the reference adjustment amount according to the residual energy standard deviation;

[0131] The frequency division extraction module 54 is used to determine the dynamic difference threshold according to the reference threshold and the reference adjustment amount, and perform frequency division extraction on the first vibration spectrum when the first energy difference is greater than the dynamic difference threshold to obtain the first frequency band spectrum of M frequency bands, where M is an integer greater than zero;

[0132] The abnormal spectrum determination module 55 is used to extract the corresponding second frequency band spectrum from the second vibration spectrum for any frequency band, determine the second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum when the second energy difference is greater than the dynamic difference threshold.

[0133] Optionally, the device for identifying the abnormal operation trend further includes:

[0134] A sliding window adjustment module, which is used to determine a dynamic difference threshold according to a reference threshold and a reference adjustment amount. If the first energy difference is greater than the dynamic difference threshold, it increases the variable width to obtain an increased variable width.

[0135] A loop execution module, which is used to, when entering the next time period for running anomaly trend recognition, take the current time period as the first historical time period, the next time period as the current time period, and the increased variable width as the variable width, and return to execute collecting the first vibration spectrum of the current time period by a vibration sensor set at a preset position of a nuclear power equipment until the recognition ends.

[0136] Optionally, the dynamic difference threshold analysis module 53 includes:

[0137] A reference threshold calculation unit, which is used to obtain an updated weight coefficient, use the updated weight coefficient to perform weighted summation on the standard energy mean and the trend energy mean, and determine that the weighted summation result is the reference threshold.

[0138] A reference adjustment amount calculation unit, which is used to obtain an updated sensitivity coefficient, multiply the updated sensitivity coefficient by the residual energy standard deviation, and determine that the formed result is the reference adjustment amount.

[0139] A frequency division extraction module 54 includes:

[0140] A dynamic difference threshold calculation unit, which is used to add the reference threshold and the reference adjustment amount, and determine that the added result is the dynamic difference threshold.

[0141] Optionally, the running anomaly trend recognition device further includes:

[0142] A sensitivity update module, which is used to, before determining the reference threshold according to the standard energy mean and the trend energy mean and determining the reference adjustment amount according to the residual energy standard deviation, obtain the false alarm rate and the miss rate after the last recognition is completed, and adjust the current sensitivity coefficient according to the false alarm rate and the miss rate to obtain an updated sensitivity coefficient.

[0143] A weight update module, which is used to obtain the historical vibration spectrum collected by the vibration sensor in the second historical time period, analyze the historical vibration spectrum, determine the trend value in the second historical time period, and adjust the current weight coefficient according to the trend value to obtain an updated weight coefficient, where the second historical time period is greater than the first historical time period.

[0144] Optionally, the historical data analysis module 52 includes:

[0145] An initial energy analysis unit, which is used to determine the historical energy mean and the historical energy standard deviation corresponding to the second vibration spectrum according to the initial energy of all historical sub-spectrums.

[0146] A normalization unit is used to subtract the initial energy value corresponding to a historical sub-spectrum from the historical energy mean for any historical sub-spectrum, obtain a difference result, divide the difference result by the historical energy standard deviation to obtain the standard energy of the corresponding historical sub-spectrum, and traverse all historical sub-spectra to obtain the standard energy of each historical sub-spectrum.

[0147] Optionally, the running anomaly trend recognition device further includes:

[0148] A feature encoding module is used to, after performing frequency division extraction on the first vibration spectrum to obtain the first frequency band spectra of M frequency bands, use the encoder in the trained anomaly detection model to convert each first frequency band spectrum to obtain the corresponding feature vector;

[0149] A feature reconstruction module is used to use the decoder in the trained anomaly detection model to reconstruct the feature vector corresponding to each first frequency band spectrum to obtain the corresponding reconstructed spectrum;

[0150] An abnormal frequency band determination module is used to calculate the error between the first frequency band spectrum and the reconstructed spectrum corresponding to the first frequency band spectrum for any first frequency band spectrum to obtain the reconstruction error. If the reconstruction error is greater than the reconstruction error threshold, it is determined that the frequency band corresponding to the first frequency band spectrum is an abnormal frequency band;

[0151] The abnormal spectrum determination module 55 includes:

[0152] An abnormal spectrum determination unit is used to, for any abnormal frequency band, extract the corresponding second frequency band spectrum from the second vibration spectrum, determine the second energy difference between the first frequency band spectrum corresponding to the abnormal frequency band and the corresponding second frequency band spectrum. When the second energy difference is greater than the dynamic difference threshold, it is determined that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum.

[0153] Optionally, the running anomaly trend recognition device further includes:

[0154] An abnormal information acquisition module is used to, after determining that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum, acquire the abnormal center frequency and abnormal bandwidth of the abnormal spectrum, as well as the historical center frequency and historical bandwidth of the abnormal spectrum in the second historical time period;

[0155] A reliability analysis module is used to respectively compare the abnormal center frequency and abnormal bandwidth with the historical center frequency and historical bandwidth to obtain a comparison result. According to the comparison result, perform a reliability analysis on the abnormal spectrum to determine the abnormal spectrum that meets the reliability analysis conditions as the final abnormal spectrum.

[0156] It should be noted that the content such as information interaction and execution process between the above modules, due to being based on the same concept as the method embodiment of this application, for its specific functions and the technical effects brought, reference can be specifically made to the method embodiment part, and details will not be elaborated here.

[0157] Figure 6 FIG. 6 is a schematic structural diagram of a computer device provided in Embodiment 6 of this application. As Figure 6 shown, the computer device in this embodiment includes: at least one processor ( Figure 6 only one is shown in FIG. 6), a memory, and a computer program stored in the memory and executable on at least one processor. When the processor executes the computer program, it implements the steps in any of the above method embodiments for identifying the abnormal trend of the operation of nuclear power equipment.

[0158] The computer device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that Figure 6 this is only an example of a computer device and does not constitute a limitation on the computer device. The computer device may include more or fewer components than those shown in the figure, or combine some components, or different components. For example, it may also include a network interface, a display screen, and an input device, etc.

[0159] The so-called processor may be a CPU, and this processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0160] The memory includes a readable storage medium, an internal memory, etc. Among them, the internal memory can be the memory of a computer device, and the internal memory provides an environment for the operation of the operating system and computer-readable instructions in the readable storage medium. The readable storage medium can be the hard disk of a computer device, and in some other embodiments, it can also be an external storage device of the computer device. For example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device. Further, the memory can also include both the internal storage unit of the computer device and the external storage device. The memory is used to store the operating system, application programs, a BootLoader, data, and other programs, etc. The other programs such as the program code of a computer program, etc. The memory can also be used to temporarily store the data that has been output or will be output.

[0161] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is used as an example. In practical applications, the above-mentioned functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above-mentioned device can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-mentioned method embodiments of this application, a computer program can be used to instruct the relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0162] All or part of the processes in the above-mentioned method embodiments of this application can also be completed by a computer program product. When the computer program product runs on a computer device, it enables the computer device to execute and implement the steps in the above-mentioned method embodiments.

[0163] In the above-mentioned embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0164] Those of ordinary skill in the art will realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0165] In the embodiments provided in this application, it should be understood that the disclosed device / computer device and method can be implemented in other ways. For example, the device / computer device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in an electrical, mechanical or other form.

[0166] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] The above embodiments are only used to illustrate the technical solutions of this application, rather than to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application.

Claims

1. A method for identifying abnormal operation trends of nuclear power equipment, characterized in that, The described abnormal operation trend recognition method includes: Collecting a first vibration spectrum in the current time period using a vibration sensor set at a preset position of the nuclear power equipment, obtaining a second vibration spectrum collected by the vibration sensor in a first historical time period, and determining a first energy difference between the first vibration spectrum and the second vibration spectrum, where the duration of the current time period is the same as that of the first historical time period; Extracting the second vibration spectrum using a sliding window with a variable width to obtain N historical sub-spectra with a time sequence, determining the initial energy of each historical sub-spectrum, performing a normalization process on the initial energy values to obtain the standard energy of each historical sub-spectrum, and calculating the corresponding standard energy mean based on the standard energy of each historical sub-spectrum, where N is an integer greater than zero; Performing a time series decomposition on each historical sub-spectrum to obtain a trend sub-spectrum and a residual sub-spectrum corresponding to the historical sub-spectrum, determining the trend energy mean and the residual energy standard deviation of all trend sub-spectra, determining a reference threshold based on the standard energy mean and the trend energy mean, and determining a reference adjustment amount based on the residual energy standard deviation; Determining a dynamic difference threshold based on the reference threshold and the reference adjustment amount, and when the first energy difference is greater than the dynamic difference threshold, performing a frequency division extraction on the first vibration spectrum to obtain first band spectra in M bands, where M is an integer greater than zero; For any band, extracting a corresponding second band spectrum from the second vibration spectrum, determining a second energy difference between the first band spectrum corresponding to the band and the corresponding second band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determining that the spectrum of the band in the first vibration spectrum is an abnormal spectrum.

2. The method for identifying the abnormal operation trend according to claim 1, wherein After determining the dynamic difference threshold according to the reference threshold and the reference adjustment amount, it further includes: If the first energy difference is greater than the dynamic difference threshold, increasing the variable width to obtain an increased variable width; When entering the next time period for abnormal operation trend recognition, taking the current time period as the first historical time period, taking the next time period as the current time period, and taking the increased variable width as the variable width, and returning to execute the step of collecting the first vibration spectrum in the current time period using the vibration sensor set at the preset position of the nuclear power equipment until the recognition ends.

3. The operating anomaly trend recognition method according to claim 1, wherein The step of determining the reference threshold according to the standard energy mean and the trend energy mean, and determining the reference adjustment amount according to the residual energy standard deviation includes: Obtaining an updated weight coefficient, using the updated weight coefficient to perform a weighted sum of the standard energy mean and the trend energy mean, and determining the weighted sum result as the reference threshold; Obtaining an updated sensitivity coefficient, multiplying the updated sensitivity coefficient by the residual energy standard deviation, and determining the resulting product as the reference adjustment amount; The step of determining the dynamic difference threshold according to the reference threshold and the reference adjustment amount includes: Adding the reference threshold and the reference adjustment amount, and determining the added result as the dynamic difference threshold.

4. The operating anomaly trend recognition method according to claim 3, wherein, Before determining the reference threshold according to the standard energy mean value and the trend energy mean value and determining the reference adjustment amount according to the residual energy standard deviation, the following steps are further included: After the last recognition is completed, obtain the false alarm rate and the miss rate, and adjust the current sensitivity coefficient according to the false alarm rate and the miss rate to obtain an updated sensitivity coefficient; Obtain the historical vibration spectrum collected by the vibration sensor in the second historical time period, analyze the historical vibration spectrum, determine the trend value in the second historical time period, and adjust the current weight coefficient according to the trend value to obtain an updated weight coefficient, where the second historical time period is longer than the first historical time period.

5. The method for identifying the running anomaly trend according to claim 1, wherein The step of normalizing the initial energy value to obtain the standard energy of each historical sub-spectrum includes: Determine the historical energy mean value and the historical energy standard deviation corresponding to the second vibration spectrum according to the initial energy of all historical sub-spectrums; For any historical sub-spectrum, subtract the initial energy value corresponding to the historical sub-spectrum from the historical energy mean value to obtain a difference result, divide the difference result by the historical energy standard deviation to obtain the standard energy corresponding to the historical sub-spectrum, and traverse all historical sub-spectrums to obtain the standard energy of each historical sub-spectrum.

6. The method for identifying the running anomaly trend according to any one of claims 1 to 5, characterized in that After performing frequency division extraction on the first vibration spectrum to obtain the first frequency band spectra of M frequency bands, the following steps are further included: Use the encoder in the trained anomaly detection model to convert each first frequency band spectrum to obtain a corresponding feature vector; Use the decoder in the trained anomaly detection model to reconstruct each feature vector corresponding to the first frequency band spectrum to obtain a corresponding reconstructed spectrum; For any first frequency band spectrum, calculate the error between the first frequency band spectrum and the reconstructed spectrum corresponding to the first frequency band spectrum to obtain a reconstruction error. If the reconstruction error is greater than the reconstruction error threshold, determine that the frequency band corresponding to the first frequency band spectrum is an abnormal frequency band; For any frequency band, extract the corresponding second frequency band spectrum from the second vibration spectrum, determine the second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the frequency band in the first vibration spectrum is an abnormal spectrum, including: For any abnormal frequency band, extract the corresponding second frequency band spectrum from the second vibration spectrum, determine the second energy difference between the first frequency band spectrum corresponding to the abnormal frequency band and the corresponding second frequency band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determine that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum.

7. The operating anomaly trend recognition method according to claim 6, wherein After determining that the spectrum of the abnormal frequency band in the first vibration spectrum is an abnormal spectrum, the following steps are further included: Obtain the abnormal center frequency and abnormal bandwidth of the abnormal spectrum, as well as the historical center frequency and historical bandwidth of the abnormal spectrum in the second historical time period; Compare the abnormal center frequency and the abnormal bandwidth with the historical center frequency and the historical bandwidth respectively to obtain a comparison result. According to the comparison result, perform a reliability analysis on the abnormal spectrum, and determine the abnormal spectrum that meets the reliability analysis conditions as the final abnormal spectrum.

8. An abnormal operation trend recognition device for nuclear power equipment, characterized in that, The operation abnormal trend recognition device includes: A current data analysis module, configured to collect a first vibration spectrum in a current time period by using a vibration sensor arranged at a preset position of the nuclear power equipment, obtain a second vibration spectrum collected by the vibration sensor in a first historical time period, and determine a first energy difference between the first vibration spectrum and the second vibration spectrum, wherein the duration of the current time period is the same as that of the first historical time period; A historical data analysis module, configured to extract the second vibration spectrum by using a sliding window with a variable width to obtain N historical sub-spectrums with a time sequence, determine an initial energy of each historical sub-spectrum, perform a normalization process on the initial energy value to obtain a standard energy of each historical sub-spectrum, and calculate a corresponding standard energy mean according to the standard energy of each historical sub-spectrum, where N is an integer greater than zero; A dynamic difference threshold analysis module, configured to perform a time series decomposition on each historical sub-spectrum to obtain a trend sub-spectrum and a residual sub-spectrum corresponding to the historical sub-spectrum, determine a trend energy mean and a residual energy standard deviation of all trend sub-spectrums, determine a reference threshold according to the standard energy mean and the trend energy mean, and determine a reference adjustment amount according to the residual energy standard deviation; A frequency division extraction module, configured to determine a dynamic difference threshold according to the reference threshold and the reference adjustment amount, and when the first energy difference is greater than the dynamic difference threshold, perform a frequency division extraction on the first vibration spectrum to obtain a first frequency band spectrum of M frequency bands, where M is an integer greater than zero; An abnormal spectrum determination module, configured to extract a corresponding second frequency band spectrum from the second vibration spectrum for any frequency band, determine a second energy difference between the first frequency band spectrum corresponding to the frequency band and the corresponding second frequency band spectrum, and when the second energy difference is greater than the dynamic difference threshold, determine the spectrum of the frequency band in the first vibration spectrum as an abnormal spectrum.

9. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation abnormal trend recognition method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the operation abnormal trend recognition method according to any one of claims 1 to 7.

Images