Method and device for determining working state of hydroelectric generating set, medium and product

The method employs non-uniform sampling and integrated bagging classification to decompose and analyze voiceprint signals, addressing the challenge of overlapping sound sources in hydropower units, enabling accurate assessment of operation status and flow properties.

CN119091919APending Publication Date: 2024-12-06CHINA INST OF WATER RESOURCES & HYDROPOWER RES
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411178075.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The challenge in hydropower stations is accurately determining the internal conditions and health level of hydropower units due to overlapping soundprint signals from various excitation sources, making it difficult to assess operating conditions and internal flow properties.

Method used

A method using non-uniformly sampled multivariate empirical mode decomposition to decompose voiceprint signals, followed by logarithmic energy calculation and integrated bagging classification to construct a voiceprint signal analysis model, enabling accurate determination of the unit's working state.

Benefits of technology

This method effectively extracts regular voiceprint signals, allowing for high recognition and extraction of sound features, thereby accurately judging the operation status and internal flow properties of hydropower units.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091919A_ABST
    Figure CN119091919A_ABST
Patent Text Reader

Abstract

The invention discloses a hydroelectric generating set working state determination method and device, a medium and a product, and relates to the field of voiceprint signal processing, and the method comprises the steps: generating a multivariate signal based on an obtained voiceprint signal; decomposing the multivariate signal by adopting a non-uniform sampling multivariate empirical mode decomposition method to obtain a plurality of stationary components; performing logarithmic energy calculation on the stationary component to obtain logarithmic energy; taking the logarithmic energy as a characteristic parameter to generate a characteristic value vector; constructing an integrated Bagging classifier, training and testing the integrated Bagging classifier by adopting the training set and the test set, and taking the trained integrated Bagging classifier as a voiceprint signal analysis model until a set condition is met; inputting the characteristic value vector into a voiceprint signal analysis model to obtain a voiceprint signal analysis result; and determining the working state of the hydroelectric generating set based on the voiceprint signal analysis result. The method can accurately judge the operation condition and the internal flow attribute of the hydroelectric generating set.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of voiceprint signal processing, and in particular to a method, device, medium and product for determining the working status of a hydropower unit. Background Art

[0002] In a hydropower station, there are many potential sources of vibration around the hydropower units, including vibration of rotating machinery, impact of fluid dynamics, and electromagnetic effects. The soundprint signals generated by these excitation sources overlap with each other, making it extremely difficult to directly identify the internal conditions of the unit through acoustic monitoring. In particular, it is challenging to obtain key parameters of internal conditions such as speed and water pressure, which brings obvious difficulties to accurately assessing the health level of the unit. Although the on-site technicians pay special attention to fluid dynamic phenomena that may affect the safety of the unit and appear in the form of specific sounds, such as the Karman vortex street effect, vortices and cavitation in the tailwater pipe, they are interfered by the technicians' abilities and external conditions, and cannot accurately listen to abnormal sounds, which leads to the inability to accurately judge the working state (i.e., operating conditions) and internal flow properties of the hydropower unit. Summary of the invention

[0003] The purpose of this application is to provide a method, device, medium and product for determining the working status of a hydropower unit, which can accurately judge the operating conditions and internal flow properties of the hydropower unit.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for determining the working state of a hydropower unit, comprising:

[0006] Acquire a voiceprint signal of a hydropower unit, and generate a multivariate signal based on the voiceprint signal;

[0007] Decomposing the multivariate signal by using a non-uniformly sampled multivariate empirical mode decomposition method to obtain a plurality of stationary components;

[0008] Determine the logarithmic energy of the stationary component, and use the logarithmic energy as a characteristic parameter to generate an eigenvalue vector;

[0009] Construct an integrated bagging classifier, use a training set and a test set to train and test the integrated bagging classifier, and use the trained integrated bagging classifier as a voiceprint signal analysis model when the set conditions are met; the training set and the test set are both constructed using voiceprint signals under different operating conditions of the hydropower unit;

[0010] Inputting the eigenvalue vector into the voiceprint signal analysis model to obtain a voiceprint signal analysis result;

[0011] The working state of the hydropower unit is determined based on the analysis result of the voiceprint signal; the working state includes a healthy operation state and a fault state.

[0012] Optionally, the multivariate signal is decomposed by using a non-uniformly sampled multivariate empirical mode decomposition method to obtain a plurality of stationary components, including:

[0013] Performing covariance decomposition on the multivariate signal to obtain an eigenvector matrix and an eigenvalue matrix;

[0014] A uniform sampling sphere is constructed based on the eigenvalue matrix, and a Hammerseley sequence is used to uniformly sample the uniform sampling sphere to obtain a uniform sample;

[0015] Determining the azimuth and inclination of the projection of the Hammerseley sequence on the uniform sampling sphere;

[0016] Constructing a rotation ellipsoid based on the azimuth, the inclination and the eigenvalue matrix;

[0017] Determine a non-uniform projection of the eigenvector matrix on the rotating ellipsoid to obtain a non-uniform sample;

[0018] Based on the uniform sample and the non-uniform sample, a multivariate empirical mode decomposition method is used to perform local mean estimation to obtain n stationary components c i (t), i = 1, 2, 3, ..., n; t represents the tth point of the signal.

[0019] Optionally, the rotating ellipsoid is represented as:

[0020]

[0021] In the formula, represents a rotating ellipsoid, φ H represents the azimuth, θ H represents the inclination angle, λ 1 ,λ 2 ,λ 3 are all eigenvalues ​​of the eigenvalue matrix, and λ 1 >λ 2 >λ 3 .

[0022] Optionally, the non-uniform projection of the eigenvector matrix on the rotating ellipsoid is expressed as:

[0023]

[0024] In the formula, represents non-uniform projection, represents a rotating ellipsoid, and V represents the eigenvector matrix.

[0025] Optionally, determining the logarithmic energy of the stationary component comprises:

[0026] Performing logarithmic energy calculation on the stationary component to obtain intermediate total energy;

[0027] The logarithmic energy of the stationary component is obtained based on the intermediate total energy.

[0028] Optionally, logarithmic energy calculation is performed on each of the n stationary components to obtain the energy of the component, including:

[0029] Assume that the i-th stationary component c i (t) There are N data points, that is, t = 1, 2, 3, ..., N; for t = 2, ..., N-1, the formula Y is used t =Y t-1 +((c i (t)^2)-c i (t-1)*c i (t+1)) Iterate and solve the intermediate energy of each stationary component to obtain the energy of each component; where Y t represents the logarithmic energy of the t-th point, Y t-1 represents the logarithmic energy of the t-1th point, c i (t-1) represents the t-1th point of the i-th stationary component, c i (t+1) represents the t+1th point of the i-th stationary component, Y 1 =0.

[0030] Optionally, an integrated bagging classifier is constructed, and the integrated bagging classifier is trained and tested using a training set and a test set until a set condition is met, and the trained integrated bagging classifier is used as a voiceprint signal analysis model, including:

[0031] The Bagging algorithm is used to obtain the base classifiers of multiple single classification regression trees through resampling technology;

[0032] Iteratively training the base classifier of each single classification regression tree by using the training set until a set number of iterations is met, thereby obtaining a trained base classifier of the single classification regression tree;

[0033] All the trained single classification regression tree base classifiers are integrated by majority voting to obtain a trained integrated Bagging classifier.

[0034] Use the test set to test whether the trained integrated Bagging classifier meets the set conditions;

[0035] When the trained integrated Bagging classifier does not meet the set conditions, the sampling method is adjusted, and the base classifier of multiple single classification regression trees obtained by resampling technology using the Bagging algorithm is returned to be executed;

[0036] When the trained integrated Bagging classifier meets the set conditions, the trained integrated Bagging classifier is used as the voiceprint signal analysis model.

[0037] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method for determining the working status of a hydropower unit provided above.

[0038] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for determining the working state of a hydropower unit provided above.

[0039] In a fourth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the method for determining the working status of a hydropower unit provided above.

[0040] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0041] The present application provides a method, device, medium and product for determining the working status of a hydropower unit. By using a non-uniformly sampled multivariate empirical mode decomposition method to decompose multivariate signals, it can effectively display the energy distribution pattern of voiceprint signals in multiple frequency ranges, making it easier to extract voiceprint signals with better regularity, thereby laying a solid data foundation for accurately obtaining abnormal voiceprint signals. The present application uses a trained integrated Bagging classifier as a voiceprint signal analysis model to obtain a voiceprint signal analysis result, achieves high recognition and extraction of sound features, can accurately determine the working status of a hydropower unit, and thereby achieves accurate evaluation of the operation status and internal flow properties of a hydropower unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1This is an application environment diagram of a method for determining the working state of a hydropower unit in one embodiment of the present application;

[0044] Figure 2 A schematic flow chart of a method for determining the working state of a hydropower unit provided in one embodiment of the present application;

[0045] Figure 3 A schematic diagram of a voiceprint signal under 150MW power generation conditions provided in an embodiment of the present application;

[0046] Figure 4 A schematic diagram of a voiceprint signal under a 300MW power generation condition provided in an embodiment of the present application;

[0047] Figure 5 A schematic diagram of a soundprint signal under a pumping condition provided by an embodiment of the present application;

[0048] Figure 6 A schematic diagram of a soundprint signal under a pumping phase adjustment condition provided by an embodiment of the present application;

[0049] Figure 7 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0052] The method for determining the working state of a hydropower unit provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the voiceprint signal of the hydropower unit to the server 104. After the server 104 receives the voiceprint signal of the hydropower unit, the server 104 generates a multivariate signal based on the voiceprint signal of the hydropower unit; the multivariate empirical mode decomposition method of non-uniform sampling is used to decompose the multivariate signal to obtain multiple stationary components; the logarithmic energy of the stationary component is determined, and the logarithmic energy is used as a feature parameter to generate an eigenvalue vector; the eigenvalue vector is input into the pre-constructed voiceprint signal analysis model to obtain the voiceprint signal analysis result; the working state of the hydropower unit is determined based on the voiceprint signal analysis result. Among them, the voiceprint signal analysis model is a classification model built based on the integrated Bagging classifier and implanted in the server 104. The server 104 can feedback the working state of the hydropower unit to the terminal 102. In addition, in some embodiments, the method for determining the working status of the hydropower unit can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly determine the working status of the hydropower unit based on the voiceprint signal of the hydropower unit, or the server 104 can obtain the voiceprint signal of the hydropower unit from the data storage system and determine the working status of the hydropower unit based on the voiceprint signal of the hydropower unit.

[0053] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0054] In an exemplary embodiment, Figure 2 As shown, a method for determining the working state of a hydropower unit is provided. The method is executed by a computer device, and can be executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 200 to 205. Among them:

[0055] Step 200: Acquire the voiceprint signal of the hydropower unit, and generate a multivariate signal based on the voiceprint signal.

[0056] Step 201: Decompose the multivariate signal using a non-uniformly sampled multivariate empirical mode decomposition method to obtain multiple stationary components.

[0057] Step 202: Determine the logarithmic energy of the stationary component, and use the logarithmic energy as a characteristic parameter to generate an eigenvalue vector.

[0058] Step 203: Construct an integrated bagging classifier, use the training set and the test set to train and test the integrated bagging classifier, until the set conditions are met, and use the trained integrated bagging classifier as the voiceprint signal analysis model. Both the training set and the test set are constructed using voiceprint signals under different operating conditions of the hydropower unit.

[0059] Step 204: Input the eigenvalue vector into the voiceprint signal analysis model to obtain the voiceprint signal analysis result.

[0060] Step 205: Determine the working state of the hydropower unit based on the analysis result of the voiceprint signal. The working state includes a healthy operation state and a fault state.

[0061] By implementing the above steps 200 to 205, high recognition and extraction of sound features are achieved, and the working state of the hydropower unit can be accurately determined, thereby achieving accurate judgment of the operation status and internal flow properties of the hydropower unit. In addition, this application also lays a solid data foundation for accurately obtaining abnormal soundprint signals.

[0062] In another exemplary embodiment of the present application, in order to obtain a component with better regularity than the original voiceprint signal, to well reflect the characteristics of the tail voiceprint signal, to effectively display the energy distribution pattern of the voiceprint signal in multiple frequency ranges, and to facilitate the extraction of a voiceprint signal with better regularity, the implementation process of the above step 201 may include:

[0063] Step 211: Perform covariance decomposition on the multivariate signal to obtain the eigenvector matrix and eigenvalue matrix. For example, given a ternary signal x(t), perform covariance decomposition on it, expressed as E{xx T}=VΛV T . Where V is the eigenvector matrix, Λ=diag{λ 1 ,λ 2 ,λ 3} is the eigenvalue matrix, λ 1 ,λ 2 ,λ 3 are all eigenvalues ​​of the eigenvalue matrix, and λ 1 >λ 2 >λ 3 .

[0064] Step 2012: construct a uniform sampling sphere based on the eigenvalue matrix, and use the Hammerseley sequence to uniformly sample the uniform sampling sphere to obtain a uniform sample.

[0065] Step 2013: Determine the azimuth and inclination of the projection of the Hammerseley sequence on the uniform sampling sphere, thereby determining the Cartesian coordinates of the uniform sampling sphere.

[0066] Step 2014: construct a rotation ellipsoid based on the azimuth angle, inclination angle and eigenvalue matrix. The rotation ellipsoid is represented as:

[0067]

[0068] In the formula, represents a rotating ellipsoid, φ H represents the azimuth, θ H Indicates the inclination angle.

[0069] Step 2015: Determine the non-uniform projection of the eigenvector matrix on the rotating ellipsoid to obtain a non-uniform sample. The non-uniform projection is expressed as:

[0070]

[0071] In the formula, represents non-uniform projection, represents a rotating ellipsoid, and V represents the eigenvector matrix.

[0072] Step 2016: Based on the uniform samples and the non-uniform samples, the multielement empirical mode decomposition (MEMD) method is used to estimate the local mean and obtain n stationary components c i (t), i = 1, 2, 3, ..., n; t represents the tth point of the signal.

[0073] In another exemplary embodiment of the present application, in order to improve the sensitivity, stability and reliability of voiceprint signal analysis, multi-scale decomposition combined with logarithmic energy value can be used for analysis. Based on this, the process of determining the logarithmic energy of the stationary component in the above step 202 may include:

[0074] Step 2021: Calculate the logarithmic energy of the stationary component to obtain the intermediate total energy. For example, assuming that the ith stationary component c i (t) There are N data points, that is, t = 1, 2, 3, ..., N; for t = 2, ..., N-1, the formula Y is used t =Y t-1 +((c i (t)^2)-c i (t-1)*c i(t+1)) Iterate and solve the intermediate energy of each stationary component to obtain the energy of each component; where Y t represents the logarithmic energy of the t-th point, Y t-1 represents the logarithmic energy of the t-1th point, c i (t-1) represents the t-1th point of the i-th stationary component, c i (t+1) represents the t+1th point of the i-th stationary component, Y 1 =0.

[0075] Step 2022: Obtain the logarithmic energy of the stationary component based on the intermediate total energy. The logarithmic energy is expressed as:

[0076] h i =log(Y).

[0077] In the formula, h i represents logarithmic energy, and Y represents the intermediate total energy.

[0078] In another exemplary embodiment of the present application, the Bagging algorithm obtains the base classifiers of several single classification regression trees through resampling technology, and the final classification result is generated by majority voting on the classification results of several single classification regression tree base classifiers. The difference of the base classifier comes from the difference of the training subset. Based on this, the implementation process of the above step 203 may include:

[0079] Step 2031: Using the Bagging algorithm and resampling technology, base classifiers of multiple single classification regression trees are obtained.

[0080] Step 2032: Use the training set to iteratively train the base classifier of each single classification regression tree until the set number of iterations is met, thereby obtaining the trained base classifier of the single classification regression tree.

[0081] For example, select a training set D of size m = {(x 1 ,y 1 ),(x 2 ,y 2 ),···,(x i ,y i ),···,(x m ,y m )}, where x i is any sample in the training set space X, which is a vector consisting of multiple attribute values. i ∈Y={1,2,···,k}, is the same as x i The corresponding category.

[0082] Set the number of iterations (i.e. the number of base classifiers generated) to t TFor a given single classification regression tree-based classifier, in the tth iteration, the training set D is first resampled to obtain the training subset D for this iteration. t ={(x 1 ,y 1 ),(x 2 ,y 2 ),···,(x j ,y j ),···,(x n ,y n )}, using the obtained training subset D t The single classification regression tree-based classifier is trained to obtain the prediction function h(t).

[0083] Step 2033: Integrate all the trained base classifiers of the single classification regression tree by majority voting to obtain a trained integrated Bagging classifier. T After rounds of iterations, all single classification regression tree base classifiers are integrated by majority voting to obtain a trained integrated Bagging classifier.

[0084] Step 2034: Use the test set to test whether the trained integrated bagging classifier meets the set conditions.

[0085] Step 2035: When the trained integrated bagging classifier does not meet the set conditions, adjust the sampling method and return to step 2031.

[0086] Step 2036: When the trained integrated bagging classifier meets the set conditions, the trained integrated bagging classifier is used as the voiceprint signal analysis model.

[0087] In another exemplary embodiment of the present application, taking the collection of the soundprint signal of the tailwater pipe entry gate under four operating conditions, namely, 150MW power generation condition, 300MW power generation condition, pumping condition, and pumping phase adjustment as an example, the implementation process and the effect achieved of the method for determining the working state of the hydropower unit provided above in the present application are explained.

[0088] In this embodiment, the implementation process of the method for determining the working state of a hydropower unit includes:

[0089] 1) Through field tests on hydropower units, the soundprint signal data of the tailwater pipe entry gate were collected under four operating conditions: power generation of 150MW, power generation of 300MW, pumping, and pumping phase adjustment.

[0090] 2) The non-uniform sampling ternary EMD method is used to decompose the tailwater pipe entry gate soundprint signals under four working conditions, and several stationary components are obtained corresponding to the soundprint signals under the four conditions.

[0091] 3) Calculate the logarithmic energy of each stationary component to obtain the logarithmic energy of the stationary component, and convert the logarithmic energy h i As the characteristic parameter, generate the characteristic value vector X. i = 1, 2, ..., n, h i represents the logarithmic energy of the ith stationary component.

[0092] 4) Input the feature value vector X into the integrated Bagging classifier to train the classifier.

[0093] 5) Collect the test signal, construct the test sample feature value vector X′ according to step 2), step 3) and step 4), and use it as the input of the integrated Bagging classifier to determine the working status of the unit with the output of the integrated Bagging classifier.

[0094] In another exemplary embodiment of the present application, taking determining the working status of a power station unit as an example, the implementation process and the achieved effect of the method for determining the working status of the hydropower unit provided above in the present application are explained.

[0095] In this embodiment, the main parameters of a power station unit include: the rotor diameter is 4.158m, the rated head is 430m, the rated output is 306.1MW, the rated speed is 428.6r / min, the number of rotor blades is 9, and the rated capacity of the generator / motor is 333.3 / 325MVA / MW. The field test data of the No. 4 unit of the power station is used as a sample for research to verify the validity of this application, which specifically includes:

[0096] (1) Collect the soundprint signal of the tailwater pipe entering the door.

[0097] When the unit is operating normally, a sound detection sensor is arranged 1m away from the tailwater pipe entrance door. After the system reaches a stable state, the soundprint signal of the tailwater pipe entrance door is collected under four operating conditions: power generation of 150MW, power generation of 300MW, pumping condition, and pumping phase adjustment, to record the sound wave signal near the tailwater pipe entrance door. The sampling frequency is 1000Hz and the number of sampling points is 4096. The soundprint signals of the tailwater pipe entrance door under the four operating conditions are as follows: Figures 3 to 6 As shown. Figures 3 to 6 It can be seen that due to the influence of hydraulic, electromagnetic, mechanical and other factors, the voiceprint signal is very complex, which makes it very difficult to directly identify its operating status through graphics. Figures 3 to 6 A in the figure represents the voiceprint signal value.

[0098] (2) The non-uniform sampling ternary EMD method is used to decompose the tailwater pipe entry gate soundprint signal under four different operating conditions.

[0099] Through analysis, it can be seen that the components obtained after the signal is decomposed by the non-uniformly sampled ternary EMD method have better regularity than the original voiceprint signal, and can well reflect the characteristics of the voiceprint signal of the tailwater pipe entry gate.

[0100] (3) Calculate the logarithmic energy of each stationary component, obtain the logarithmic energy of the stationary component as the characteristic parameter, and generate the eigenvalue vector X.

[0101] Among them, the logarithmic energy values ​​of the soundprint signal of the tailwater pipe entering the gate at multiple scales and the obtained eigenvalue vectors are shown in Table 1 under four operating conditions: power generation of 150MW, power generation of 300MW, pumping condition, and pumping phase adjustment.

[0102] Table 1 Characteristic value vector table of tailwater entry gate voiceprint signal based on logarithmic energy value

[0103]

[0104]

[0105] As can be seen from Table 1, when the unit is in four operating conditions, namely, power generation of 150MW, power generation of 300MW, pumping condition, and pumping phase adjustment, the multi-scale component logarithmic energy values ​​of the soundprint signal of the tailwater pipe entering the door show significant differences, which shows that the analysis using multi-scale decomposition combined with logarithmic energy value has high sensitivity, stability and reliability.

[0106] (4) Input the eigenvalue vector X into the integrated Bagging classifier and train the classifier.

[0107] (5) Collect the test signal. According to steps (2), (3) and (4), construct the test sample feature value vector X′ and use it as the input of the integrated Bagging classifier to determine the working status of the unit by integrating the output of the Bagging classifier.

[0108] For example, from 240 sets of voiceprint signal data containing four different working conditions, 120 sets (30 sets for each working condition) were randomly selected as training samples and input into the nearest neighbor classifier for learning. Then, the remaining 120 sets of data were used as test samples for verification. The results showed that all samples involved in the test could be accurately identified, thus verifying the feasibility of using voiceprint signals to judge the operating status (i.e., working status) of hydropower units.

[0109] Based on the above description, this application uses the non-uniform sampling ternary EMD method, logarithmic energy operator and integrated bagging analysis technology to analyze and process the soundprint signals of the hydropower unit. First, the non-stationary soundprint signals under 4 different working conditions were decomposed using the non-uniform sampling ternary EMD method, and then the logarithmic energy value was used to further analyze the stable components obtained by these decompositions, thereby effectively showing the energy distribution pattern of the soundprint signal in multiple frequency ranges. Finally, an integrated bagging classifier is used to identify the operating status of the unit. The results of the example analysis show that the method provided in this application has shown excellent results in field applications and can effectively distinguish the different operating states of the unit.

[0110] Through the logarithmic energy operator analysis of multi-scale components, the operating status of the hydropower unit can be monitored, and the abnormal sounds emitted by the hydropower unit can be identified in time. This provides an effective tool for field staff to discover potential risks, and can provide intelligent suggestions to help hydropower stations make operation and maintenance decisions. This abnormality detection method can provide key operating information, help prevent major accidents, reduce the failure rate and unreliability of the unit, and improve resource utilization efficiency. Taking these advantages into account, the method provided in this application has broad application potential in the field of operation and maintenance of hydropower units.

[0111] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data for determining the working status of a hydropower unit. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for determining the working status of a hydropower unit is implemented.

[0112] Those skilled in the art will understand that Figure 7The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0113] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0114] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0115] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0116] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0117] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for determining the working state of a hydropower unit, characterized in that: The method for determining the working state of a hydroelectric unit comprises: Acquire a voiceprint signal of a hydropower unit, and generate a multivariate signal based on the voiceprint signal; Decomposing the multivariate signal by using a non-uniformly sampled multivariate empirical mode decomposition method to obtain a plurality of stationary components; Determine the logarithmic energy of the stationary component, and use the logarithmic energy as a characteristic parameter to generate an eigenvalue vector; Construct an integrated bagging classifier, use a training set and a test set to train and test the integrated bagging classifier, and use the trained integrated bagging classifier as a voiceprint signal analysis model when the set conditions are met; the training set and the test set are both constructed using voiceprint signals under different operating conditions of the hydropower unit; Inputting the eigenvalue vector into the voiceprint signal analysis model to obtain a voiceprint signal analysis result; The working state of the hydropower unit is determined based on the analysis result of the voiceprint signal; the working state includes a healthy operation state and a fault state.

2. The method for determining the working state of a hydropower unit according to claim 1, characterized in that: The multivariate signal is decomposed by using a non-uniformly sampled multivariate empirical mode decomposition method to obtain multiple stationary components, including: Performing covariance decomposition on the multivariate signal to obtain an eigenvector matrix and an eigenvalue matrix; A uniform sampling sphere is constructed based on the eigenvalue matrix, and a Hammerseley sequence is used to uniformly sample the uniform sampling sphere to obtain a uniform sample; Determining the azimuth and inclination of the projection of the Hammerseley sequence on the uniform sampling sphere; Constructing a rotation ellipsoid based on the azimuth, the inclination and the eigenvalue matrix; Determine a non-uniform projection of the eigenvector matrix on the rotating ellipsoid to obtain a non-uniform sample; Based on the uniform sample and the non-uniform sample, a multivariate empirical mode decomposition method is used to perform local mean estimation to obtain n stationary components c i (t), i = 1, 2, 3, ..., n; t represents the tth point of the signal.

3. The method for determining the working state of a hydroelectric unit according to claim 2, characterized in that: The rotating ellipsoid is represented as: In the formula, represents a rotating ellipsoid, φ H represents the azimuth, θ H represents the inclination angle, λ1, λ2, λ3 are all eigenvalues ​​of the eigenvalue matrix, and λ1>λ2>λ3.

4. The method for determining the working state of a hydroelectric unit according to claim 2, characterized in that: The non-uniform projection of the eigenvector matrix on the rotating ellipsoid is expressed as: In the formula, represents non-uniform projection, represents a rotating ellipsoid, and V represents the eigenvector matrix.

5. The method for determining the working state of a hydroelectric unit according to claim 1, characterized in that: Determining the logarithmic energy of the stationary component includes: Performing logarithmic energy calculation on the stationary component to obtain intermediate total energy; The logarithmic energy of the stationary component is obtained based on the intermediate total energy.

6. The method for determining the working state of a hydroelectric unit according to claim 5, characterized in that: The logarithmic energy of n stationary components is calculated respectively to obtain the energy of the component, including: Assume that the i-th stationary component c i (t) There are N data points, that is, t = 1, 2, 3, ..., N; for t = 2, ..., N-1, the formula Y is used t =Y t-1 +((c i (t)^2)-c i (t-1)*c i (t+1)) Iterate and solve the intermediate energy of each stationary component to obtain the energy of each component respectively; where Y t represents the logarithmic energy of the t-th point, Y t-1 represents the logarithmic energy of the t-1th point, c i (t-1) represents the t-1th point of the i-th stationary component, c i (t+1) represents the t+1th point of the i-th stationary component, Y1=0.

7. The method for determining the working state of a hydroelectric unit according to claim 1, characterized in that: Construct an integrated Bagging classifier, use a training set and a test set to train and test the integrated Bagging classifier until the set conditions are met, and use the trained integrated Bagging classifier as a voiceprint signal analysis model, including: The Bagging algorithm is used to obtain the base classifiers of multiple single classification regression trees through resampling technology; Iteratively training the base classifier of each single classification regression tree by using the training set until a set number of iterations is met, thereby obtaining a trained base classifier of the single classification regression tree; All the trained single classification regression tree base classifiers are integrated by majority voting to obtain a trained integrated Bagging classifier. Use the test set to test whether the trained integrated Bagging classifier meets the set conditions; When the trained integrated Bagging classifier does not meet the set conditions, the sampling method is adjusted, and the base classifier of multiple single classification regression trees obtained by resampling technology using the Bagging algorithm is returned to be executed; When the trained integrated Bagging classifier meets the set conditions, the trained integrated Bagging classifier is used as the voiceprint signal analysis model.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for determining the working state of a hydropower unit according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for determining the working state of a hydropower unit according to any one of claims 1 to 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for determining the working state of a hydropower unit according to any one of claims 1 to 7 is implemented.