Multidimensional analysis methods, devices, equipment and media for wind turbine vibration data

By constructing a three-dimensional time-frequency spectrum and integrating operating parameters, the problem of traditional methods being unable to identify wind turbine faults under complex operating conditions was solved, and accurate analysis and fault identification of wind turbine vibration signals were achieved.

CN123064635APending Publication Date: 2026-11-10THREE GORGES NEW ENERGY QIXIAN POWER GENERATION CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610972655.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-01
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Traditional vibration analysis methods struggle to accurately identify fault characteristics in wind turbines under complex operating conditions, leading to misdiagnosis or missed diagnosis, and are unable to effectively distinguish between normal vibrations caused by operating conditions and potential faults.

Method used

A wavelet coefficient matrix is ​​generated by continuous wavelet transform, a three-dimensional time-frequency spectrum is constructed and external operating parameters are fused, and the operating status and potential fault characteristics of wind turbine units are identified by analyzing the three-dimensional time-frequency spectrum.

Benefits of technology

This paper presents an intuitive and comprehensive multidimensional analysis method for vibration data, which can effectively identify fault characteristics in non-stationary signals, eliminate operating condition interference, and improve diagnostic accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123064635A_ABST
    Figure CN123064635A_ABST
Patent Text Reader

Abstract

This disclosure relates to a method, apparatus, equipment, and medium for multidimensional analysis of vibration data from wind turbine generators. The method includes: acquiring the original vibration signal of the wind turbine generator and synchronously collected external operating parameters, and preprocessing the original vibration signal; analyzing the preprocessed vibration signal using continuous wavelet transform to generate a wavelet coefficient matrix; constructing a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, and visually integrating the external operating parameters into the three-dimensional time-frequency spectrum; and analyzing the integrated three-dimensional time-frequency spectrum to identify the operating status and potential fault characteristics of the wind turbine generator. Therefore, by integrating time-frequency analysis with the visualization of operating parameters, an intuitive and comprehensive multidimensional vibration data analysis method is provided, which can effectively identify fault characteristics in non-stationary signals, eliminate operating condition interference, and improve diagnostic accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to the field of vibration monitoring and analysis technology, and in particular to a method, device, equipment and medium for multidimensional analysis of vibration data of wind turbine generators. Background Technology

[0002] Wind turbines typically operate in complex and variable natural environments. Their vibration signals are influenced by various operating conditions such as wind speed, wind direction, blade angle of attack, rotor speed, generator torque, grid frequency, and ambient temperature, exhibiting typical non-stationary and nonlinear characteristics. Traditional vibration analysis methods mainly include time-domain analysis, frequency-domain analysis, and trend analysis. Time-domain analysis directly observes the change in amplitude over time, but it is difficult to intuitively interpret signal characteristics when complex frequency components are superimposed. Frequency-domain analysis converts the signal to the frequency domain through Fourier transform, which can clearly show the frequency components, but it assumes that the signal is stationary and cannot reflect the change in frequency over time, thus losing time information. Trend analysis judges the trend of amplitude changes through long-term data accumulation, but it is difficult to locate specific fault types.

[0003] During wind turbine operation, frequent adjustments to operating conditions (such as pitch, yaw, and torque control) cause changes in the frequency components and energy distribution of vibration signals. Ignoring these changes and directly analyzing vibration data using traditional methods can easily lead to misinterpreting normal vibration fluctuations caused by operating conditions as faults, or obscuring early, subtle fault characteristics under the influence of these changes, resulting in misdiagnosis or missed diagnosis. Therefore, a multidimensional analysis method that simultaneously considers time-frequency domain characteristics and operating parameters is urgently needed to improve the accuracy and reliability of wind turbine vibration monitoring. Summary of the Invention

[0004] To address the aforementioned technical issues, this disclosure provides a method, apparatus, equipment, and medium for multidimensional analysis of vibration data from wind turbine generators.

[0005] Firstly, this disclosure provides a multidimensional analysis method for wind turbine vibration data, including: The original vibration signal of the wind turbine and the external operating parameters acquired synchronously are obtained, and the original vibration signal is preprocessed. The preprocessed vibration signal is analyzed using continuous wavelet transform to generate a wavelet coefficient matrix; A three-dimensional time-frequency spectrum is constructed based on the wavelet coefficient matrix, and the external operating condition parameters are integrated into the three-dimensional time-frequency spectrum in a visual form; By analyzing the fused three-dimensional time-frequency spectrum, the operating status and potential fault characteristics of wind turbine units can be identified.

[0006] Secondly, this disclosure provides a multi-dimensional analysis device for wind turbine vibration data, including: The data acquisition module is used to acquire the original vibration signal of the wind turbine and the synchronously acquired external operating parameters, and to preprocess the original vibration signal. The first processing module is used to analyze the preprocessed vibration signal using continuous wavelet transform and generate a wavelet coefficient matrix. The second processing module is used to construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, and to integrate the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form; The third processing module is used to analyze the fused three-dimensional time-frequency spectrum to identify the operating status and potential fault characteristics of the wind turbine.

[0007] Thirdly, this disclosure provides a multi-dimensional analysis device for wind turbine vibration data, including: processor; Memory, used to store executable instructions; The processor is used to read executable instructions from memory and execute the executable instructions to implement the multidimensional analysis method of wind turbine vibration data in the first aspect.

[0008] Fourthly, this disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the multidimensional analysis method for wind turbine vibration data of the first aspect.

[0009] The technical solution provided in this disclosure has the following advantages compared with the prior art: The multidimensional vibration data analysis method for wind turbines disclosed in this embodiment can acquire the original vibration signal of the wind turbine and synchronously collected external operating parameters. The original vibration signal is preprocessed, and then continuous wavelet transform is used to analyze the preprocessed vibration signal to generate a wavelet coefficient matrix. A three-dimensional time-frequency spectrum is then constructed based on the wavelet coefficient matrix, and the external operating parameters are visualized and integrated into the three-dimensional time-frequency spectrum. Finally, by analyzing the fused three-dimensional time-frequency spectrum, the operating status and potential fault characteristics of the wind turbine are identified. Therefore, by integrating time-frequency analysis with the visualization of operating parameters, an intuitive and comprehensive multidimensional vibration data analysis method is provided, which can effectively identify fault characteristics in non-stationary signals, eliminate operating condition interference, and improve diagnostic accuracy. Attached Figure Description

[0010] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0011] Figure 1 A flowchart illustrating a multidimensional analysis method for wind turbine vibration data provided in this embodiment of the present disclosure; Figure 2 A flowchart illustrating another method for multidimensional analysis of wind turbine vibration data provided in this embodiment of the present disclosure; Figure 3 A schematic diagram of a three-dimensional time-frequency spectrum provided in an embodiment of this disclosure; Figure 4 This is a schematic diagram of the structure of a multidimensional analysis device for vibration data of a wind turbine provided in an embodiment of this disclosure; Figure 5 This is a schematic diagram of the structure of a multidimensional analysis device for vibration data of a wind turbine provided in an embodiment of this disclosure. Detailed Implementation

[0012] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0013] It should be understood that the various steps described in the method implementation of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method implementation may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0014] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0015] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0016] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0017] The names of messages or information exchanged between multiple devices in this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0018] To address the aforementioned problems, this disclosure provides a method, apparatus, equipment, and medium for multidimensional analysis of wind turbine vibration data. The following is a detailed description... Figure 1-3 The multidimensional analysis method for wind turbine vibration data provided in the embodiments of this disclosure will be described in detail.

[0019] Figure 1 A flowchart illustrating a multidimensional analysis method for wind turbine vibration data provided in an embodiment of this disclosure is shown.

[0020] In this embodiment of the disclosure, the multidimensional analysis method for wind turbine vibration data can be executed by an electronic device. This electronic device may include, but is not limited to, devices such as computer equipment, cloud servers, or cloud server clusters.

[0021] like Figure 1 As shown, the multidimensional analysis method for wind turbine vibration data can include the following steps.

[0022] S110. Acquire the original vibration signal of the wind turbine and the synchronously acquired external operating parameters, and preprocess the original vibration signal.

[0023] In this embodiment of the disclosure, the electronic device can acquire the original vibration signal of the wind turbine and the synchronously acquired external operating parameters, and preprocess the original vibration signal.

[0024] Optionally, external operating parameters may include wind speed, impeller speed, generator power, etc., which are not limited here.

[0025] Specifically, electronic devices can collect raw vibration signals through acceleration sensors installed in key parts of the wind turbine (such as main shaft bearings, gearboxes, generators, etc.), and simultaneously acquire external operating parameters, including wind speed, rotor speed, generator power, etc., through the unit monitoring system, and preprocess the raw vibration signals.

[0026] S120. Continuous wavelet transform is used to analyze the preprocessed vibration signal and generate a wavelet coefficient matrix.

[0027] In this embodiment of the disclosure, the electronic device can use continuous wavelet transform to analyze the preprocessed vibration signal and generate a wavelet coefficient matrix.

[0028] Optionally, in the Continuous Wavelet Transform (CWT), the Symlet wavelet function is selected, which allows the wavelet function to vary continuously in time and scale. It analyzes the characteristics of the signal at different times and frequencies by performing inner product operations on the signal with wavelet functions at different scales and locations.

[0029] Specifically, after obtaining the preprocessed vibration signal, the electronic device can use continuous wavelet transform to analyze the preprocessed vibration signal. For example, the signal can be multiplied by wavelet functions at different scales and time positions using the Symlet wavelet function to obtain a wavelet coefficient matrix. Each row of this matrix corresponds to a scale (frequency), and each column corresponds to a time point. The element values ​​reflect the energy intensity of the frequency component at that moment.

[0030] S130. Construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, and integrate the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form.

[0031] In this embodiment of the disclosure, the electronic device can construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix and integrate the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form.

[0032] Specifically, electronic devices can construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, such as using time as the horizontal axis and frequency as the vertical axis, with the absolute value of the wavelet coefficients representing height (or color intensity) to indicate energy. Simultaneously, synchronously acquired external operating parameters (such as wind speed) are aligned with the vibration signal on the time axis and superimposed as curves on top of the spectrum or at the maximum frequency boundary, forming a fused spectrum containing vibration time-frequency characteristics and operating condition information.

[0033] S140. By analyzing the fused three-dimensional time-frequency spectrum, the operating status and potential fault characteristics of the wind turbine are identified.

[0034] In this embodiment of the disclosure, the electronic device can identify the operating status and potential fault characteristics of the wind turbine by analyzing the fused three-dimensional time-frequency spectrum.

[0035] Specifically, electronic devices can analyze the fused three-dimensional time-frequency spectrum to identify the operating status and potential fault characteristics of wind turbines. For example, based on the energy accumulation area, the trend of change, and its correlation with the operating condition curve, it can determine whether there are any abnormalities in the wind turbine. For instance, if there is a continuous increase in energy at a certain fixed frequency while the operating condition curve remains stable, it indicates a mechanical fault; if the energy change is consistent with the fluctuation of the operating condition curve, it is a normal response caused by the operating condition.

[0036] This allows for the acquisition of raw vibration signals from wind turbines and synchronously collected external operating parameters. The raw vibration signals are preprocessed, and then continuous wavelet transform is used to analyze them, generating a wavelet coefficient matrix. A three-dimensional time-frequency spectrum is then constructed based on this matrix, and the external operating parameters are visually integrated into the spectrum. Finally, analysis of the integrated three-dimensional time-frequency spectrum identifies the operating status and potential fault characteristics of the wind turbines. Therefore, by integrating time-frequency analysis with the visualization of operating parameters, an intuitive and comprehensive multi-dimensional vibration data analysis method is provided. This method effectively identifies fault characteristics in non-stationary signals, eliminates operating condition interference, and improves diagnostic accuracy.

[0037] Optionally, S110 may specifically include: converting the original vibration signal into a one-dimensional numerical array; and performing moving average filtering on the one-dimensional numerical array based on a preset filtering window length.

[0038] In this embodiment of the disclosure, the electronic device can convert the original vibration signal into a one-dimensional numerical array.

[0039] Specifically, after acquiring the original vibration signal, the electronic device can convert the original vibration signal into a one-dimensional numerical array. The original vibration signal is generally data on the change of vibration amplitude over time. The np.array function is used to convert the original vibration signal into a one-dimensional numerical array.

[0040] Furthermore, the electronic device can perform moving average filtering on the one-dimensional numerical array based on a preset filtering window length. For example, the preset filtering window length can be 10 sampling points. The average value of the amplitude of each sampling point and the 10 points before and after it is taken as the new value of that point. The window slides through the entire array to obtain the smoothed signal.

[0041] Therefore, in this embodiment, the moving average filtering can effectively suppress high-frequency spikes caused by electromagnetic interference and sensor noise during signal transmission, while retaining the main impact and modulation characteristics in the vibration signal. The selection of a window length of 10 has been experimentally verified, achieving a good balance between denoising effect and signal fidelity, avoiding excessive smoothing that could lead to the loss of useful information. Noise removal through moving average filtering improves the signal-to-noise ratio of subsequent wavelet transform, making the wavelet coefficient matrix more reflective of the true vibration characteristics, thereby enhancing the reliability of fault identification.

[0042] Optionally, S120 may specifically include: performing convolution operations on the preprocessed vibration signal with continuous wavelet transforms at different scales and time positions to obtain the corresponding convolution calculation results; and normalizing the convolution calculation results to generate a wavelet coefficient matrix.

[0043] In this embodiment of the disclosure, the electronic device can perform convolution operations on the preprocessed vibration signal and continuous wavelet transforms at different scales and time positions to obtain the corresponding convolution calculation results.

[0044] Specifically, electronic devices can perform convolution operations on preprocessed vibration signals with continuous wavelet transforms at different scales and time positions. For example, the continuous wavelet transform can use the Symlet wavelet family (such as sym4, sym8, etc.) as wavelet basis functions. Symlet wavelets possess approximate symmetry, enabling them to better match the periodic waveforms in vibration signals; their compact support characteristics ensure computational efficiency; orthogonality avoids redundancy between wavelet coefficients; simultaneously, Symlet wavelets exhibit good localization capabilities in both the time and frequency domains, capable of simultaneously capturing transient impacts (high frequencies) and slow changes (low frequencies) in the signal. Electronic devices then use Symlet wavelet functions to perform inner product operations on the signal with wavelet functions at different scales and time positions to obtain the corresponding convolution calculation results.

[0045] Furthermore, the electronic device normalizes the convolution calculation results to generate a wavelet coefficient matrix.

[0046] Therefore, it is possible to simultaneously extract the high-frequency impact characteristics of early faults and the low-frequency modulation characteristics under operating conditions, thereby improving the accuracy and applicability of time-frequency analysis.

[0047] Optionally, the multidimensional analysis method for wind turbine vibration data may further include: determining the scale parameters of the continuous wavelet transform in response to a user's scale range selection operation, wherein the scale range selection operation is used to select either a small-scale range or a large-scale range.

[0048] In this embodiment of the disclosure, the electronic device can determine the scale parameters of the continuous wavelet transform in response to a user's scale range selection operation.

[0049] Because early wear (such as bearing micro-scratching) in wind turbine vibration signals typically manifests as high-frequency impacts, while structural resonance and imbalance faults often exhibit low-frequency components, a single scale range is insufficient to account for both. Therefore, this embodiment sets the scale parameter as a variable that can be selected by the user based on the analysis objective. Specifically, the algorithm interface provides a scale range selection interface, allowing users to choose between "high-frequency analysis mode" (small-scale range) or "low-frequency analysis mode" (large-scale range). The algorithm discretizes the scale according to the selected range and calculates the corresponding wavelet coefficients to generate a targeted time-frequency spectrum. The relationship between the scale parameter and frequency is determined by the sampling interval.

[0050] Therefore, by adjusting the scale parameter, the analysis method can flexibly adapt to the diagnostic needs of different fault types. It can be used for the fine detection of early and weak faults, as well as for the monitoring of low-frequency faults under operating conditions, thus improving the versatility and pertinence of the method.

[0051] Optionally, S130 may specifically include: determining the time parameter as the horizontal axis based on the time series of the original vibration signal according to the wavelet coefficient matrix, calculating the frequency parameter as the vertical axis based on the scale parameter, and constructing the three-dimensional time-frequency spectrum.

[0052] In this embodiment of the disclosure, the electronic device can construct the three-dimensional time-frequency spectrum by determining the time parameter as the horizontal axis through the time sequence of the original vibration signal based on the wavelet coefficient matrix, and calculating the frequency parameter as the vertical axis through the scale parameter.

[0053] Specifically, electronic devices can generate a time axis based on the time series of the original vibration signal using a wavelet coefficient matrix, with a time step Δt as the sampling interval. The frequency axis is obtained by scaling parameters: for each scale, the corresponding frequency f = 1 / (scale × Δt). The absolute value of each time-scale point in the wavelet coefficient matrix is ​​used as the energy value of that point, and a surface or color map is plotted in a three-dimensional coordinate system. The higher the energy, the darker the color or the greater the height. The constructed three-dimensional time-frequency map can intuitively display the distribution of signal energy in both time and frequency dimensions, clearly presenting the time-varying characteristics of non-stationary signals.

[0054] Therefore, the three-dimensional time-frequency spectrum contains richer information than the traditional two-dimensional spectrum, and can reveal the changing patterns of frequency components over time, which helps to identify phenomena such as intermittent faults, impulse characteristics, and frequency drift.

[0055] Optionally, S130 may specifically include: aligning the external operating condition parameters with the original vibration signal on the time axis; and overlaying the aligned external operating condition parameters onto the three-dimensional time-frequency spectrum in the form of an independent curve visualization.

[0056] In this embodiment of the disclosure, the electronic device can align the external operating condition parameters with the original vibration signal on the time axis.

[0057] Specifically, the synchronously collected operating parameters (such as wind speed and rotational speed) are time-aligned with the vibration signals to ensure that the data at the same moment correspond.

[0058] Furthermore, the electronic device can overlay the aligned external operating condition parameters onto the three-dimensional time-frequency spectrum as independent curves. For example, on the plotted three-dimensional time-frequency spectrum, operating condition parameter curves can be drawn along the time axis at the maximum frequency boundary or in the blank area above the spectrum. The vertical axis of the curve represents the value of the operating condition parameter, and the horizontal axis is consistent with the time axis of the spectrum. For clarity, a dual vertical axis can be set, with the frequency on the left and the operating condition parameter on the right.

[0059] In this way, analysts can intuitively compare the relationship between changes in vibration energy and fluctuations in operating conditions. For example, when wind speed increases, vibration energy rises at certain frequencies. If the energy change trend is highly consistent with the wind speed curve, it can be judged as a normal response caused by airflow excitation; conversely, if the energy anomaly is unrelated to the operating conditions, potential faults should be suspected.

[0060] Therefore, the visualization and fusion of operating condition parameters integrates previously separate data into a single view, greatly facilitating correlation analysis, effectively avoiding misdiagnosis caused by changes in operating conditions, and improving the accuracy of fault identification.

[0061] Optionally, S130 may specifically include: analyzing the fused three-dimensional time-frequency spectrum to determine the accumulation characteristics of vibration energy in the target frequency range, its trend over time, and its correlation with operating parameters, thereby identifying the operating status and potential fault characteristics of the wind turbine.

[0062] In this embodiment of the disclosure, the electronic device can analyze the fused three-dimensional time-frequency spectrum to determine the accumulation characteristics of vibration energy in the target frequency range, the trend of change over time, and the correlation with operating parameters, thereby identifying the operating status and potential fault characteristics of the wind turbine.

[0063] Specifically, electronic devices can analyze the fused three-dimensional time-frequency spectrum to determine factors such as whether energy is continuously concentrated in one or several frequency ranges, whether the energy increases, decreases, or appears periodically over time, and whether energy changes are synchronized with operating parameter curves. This identifies the concentration characteristics of vibration energy in the target frequency range, its trend over time, and its correlation with operating parameters. If energy is found to continuously increase at a fixed frequency (such as gear meshing frequency or bearing characteristic frequency) while operating parameters remain stable, it is identified as a mechanical fault, such as gear wear or bearing failure. If energy changes are highly correlated with operating parameters (such as rotational speed), such as increasing with increasing rotational speed, it is identified as normal vibration caused by the operating conditions. Through comprehensive judgment, an assessment of the wind turbine's operating status and early warning suggestions are provided. Furthermore, a fault case library can be established using historical data, and machine learning methods can be used to automatically identify fault modes in the spectrum, achieving intelligent diagnosis.

[0064] Therefore, the fault identification method based on multi-dimensional fusion map makes full use of time and frequency features and operating condition information, improves the reliability of fault judgment criteria, provides a scientific basis for condition-based maintenance of wind turbine units, and helps to reduce unplanned downtime and extend equipment life.

[0065] Figure 2 A flowchart illustrating another method for multidimensional analysis of wind turbine vibration data provided in an embodiment of this disclosure is shown.

[0066] like Figure 2 As shown, the electronic device can collect the raw vibration signal of the wind turbine as input data for analysis, convert the raw signal into a one-dimensional numerical array suitable for algorithm processing, and perform moving average filtering for noise reduction. The Symlet wavelet is selected as the wavelet basis function for analysis, providing a basis for convolution calculation. The denoised vibration signal is convolved with the selected wavelet basis function. The user inputs or selects the scale parameter to determine the frequency range of analysis (high frequency / low frequency mode selection). The convolution result is normalized to eliminate amplitude differences at different scales, making each frequency component comparable. A wavelet coefficient matrix containing time, scale (frequency), and energy information is generated. External operating parameters such as wind speed and rotational speed are collected synchronously. A three-dimensional time-frequency spectrum is constructed based on the wavelet coefficient matrix, and external variable data is integrated for visualization output.

[0067] Figure 3 A schematic diagram of a three-dimensional time-frequency spectrum provided in an embodiment of this disclosure is shown.

[0068] like Figure 3 As shown, in this three-dimensional time-frequency spectrum, the time parameter (Time / sec) determined based on the time series of the original vibration signal is used as the horizontal axis, and the frequency parameter (Frequency / Hz) calculated based on the scale parameter is used as the vertical axis. Color / grayscale represents energy intensity (0-1.6), and wavelet coefficient amplitude represents vibration energy. The darker / brighter the color, the stronger the energy. The color scale on the right represents the energy scale (0-1.6), quantifying the energy value corresponding to the color. In the 0-2 second time period, there is strong energy in the 0-4Hz frequency band (darker color); in the 2-3 second time period, energy diffuses to higher frequencies (4-10Hz); in the 3-4 second time period, a new energy accumulation appears in the 8-12Hz frequency band. The signal energy changes significantly over time, reflecting the non-stationary characteristics of the wind turbine vibration signal. The frequency components drift over time, reflecting the changes in the unit's operating conditions. If there is a continuous energy accumulation at a certain fixed frequency (as shown in the figure, the stable energy of 8-10Hz at 4-6 seconds), it indicates the characteristics of a periodic fault. The area where the energy suddenly increases (as shown in the figure, around 2.5 seconds) corresponds to a transient impact event.

[0069] Figure 4A schematic diagram of the structure of a multidimensional analysis device for vibration data of a wind turbine provided in an embodiment of this disclosure is shown.

[0070] like Figure 4 As shown, the multidimensional analysis device 400 for wind turbine vibration data may include a data acquisition module 410, a first processing module 420, a second processing module 430, and a third processing module 440.

[0071] The data acquisition module 410 can be used to acquire the original vibration signal of the wind turbine and the synchronously acquired external operating parameters, and to preprocess the original vibration signal.

[0072] The first processing module 420 can be used to analyze the preprocessed vibration signal using continuous wavelet transform to generate a wavelet coefficient matrix.

[0073] The second processing module 430 can be used to construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, and to integrate the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form.

[0074] The third processing module 440 can be used to identify the operating status and potential fault characteristics of wind turbine units by analyzing the fused three-dimensional time-frequency spectrum.

[0075] Therefore, in this embodiment, the original vibration signal of the wind turbine and synchronously acquired external operating parameters can be obtained. The original vibration signal is preprocessed, and then continuous wavelet transform is used to analyze the preprocessed vibration signal to generate a wavelet coefficient matrix. A three-dimensional time-frequency spectrum is then constructed based on the wavelet coefficient matrix, and the external operating parameters are visualized and integrated into the three-dimensional time-frequency spectrum. Finally, by analyzing the fused three-dimensional time-frequency spectrum, the operating status and potential fault characteristics of the wind turbine can be identified. Thus, by integrating time-frequency analysis with the visualization of operating parameters, an intuitive and comprehensive multi-dimensional vibration data analysis method is provided, which can effectively identify fault characteristics in non-stationary signals, eliminate operating condition interference, and improve diagnostic accuracy.

[0076] In some embodiments of this disclosure, the data acquisition module 410 may specifically include: The first processing unit is used to convert the original vibration signal into a one-dimensional numerical array.

[0077] The second processing unit is used to perform moving average filtering on the one-dimensional numerical array based on a preset filtering window length.

[0078] In some embodiments of this disclosure, the first processing module 420 may specifically include: The third processing unit is used to perform convolution operations on the preprocessed vibration signal and continuous wavelet transforms at different scales and time positions to obtain the corresponding convolution calculation results.

[0079] The fourth processing unit is used to normalize the convolution calculation results and generate a wavelet coefficient matrix.

[0080] In some embodiments of this disclosure, the multidimensional analysis device 400 for wind turbine vibration data may further include: The parameter determination module is used to determine the scale parameters of the continuous wavelet transform in response to the user's scale range selection operation, wherein the scale range selection operation is used to select either a small scale range or a large scale range.

[0081] In some embodiments of this disclosure, the second processing module 430 may be specifically used to construct the three-dimensional time-frequency spectrum by determining the time parameter as the horizontal axis based on the time series of the original vibration signal according to the wavelet coefficient matrix and calculating the frequency parameter as the vertical axis based on the scale parameter.

[0082] In some embodiments of this disclosure, the second processing module 430 may specifically include: The fifth processing unit is used to align the external operating parameters with the original vibration signal on the time axis.

[0083] The sixth processing unit is used to overlay the aligned external operating condition parameters onto the three-dimensional time-frequency spectrum in the form of independent curves.

[0084] In some embodiments of this disclosure, the third processing module 440 may be specifically used to analyze the fused three-dimensional time-frequency spectrum to determine the accumulation characteristics of vibration energy in the target frequency range, the trend of change over time, and the correlation with operating parameters, thereby identifying the operating status and potential fault characteristics of the wind turbine.

[0085] It should be noted that, Figure 4 The wind turbine vibration data multidimensional analysis device 400 shown can perform... Figure 1-3 The various steps in the method embodiment shown are implemented. Figure 1-3 The processes and effects in the method embodiments shown are not described in detail here.

[0086] Figure 5 A schematic diagram of the structure of a multidimensional analysis device for vibration data of a wind turbine provided in an embodiment of this disclosure is shown.

[0087] In some embodiments of this disclosure, Figure 5The multidimensional analysis device for wind turbine vibration data shown can be an electronic device. Specifically, the electronic device may include, but is not limited to, devices such as computer equipment, cloud servers, or cloud server clusters.

[0088] like Figure 5 As shown, the multidimensional analysis device for wind turbine vibration data may include a processor 501 and a memory 502 storing computer program instructions.

[0089] Specifically, the processor 501 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0090] Memory 502 may include a large-capacity storage device for information or instructions. For example, and not limitingly, memory 502 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 502 may include removable or non-removable (or fixed) media. Where appropriate, memory 502 may be internal or external to the integrated gateway device. In a particular embodiment, memory 502 is a non-volatile solid-state memory. In a particular embodiment, memory 502 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (Electrically Programmable ROM, EPROM), an electrically erasable programmable PROM (EEPROM), an electrically alterable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0091] The processor 501 reads and executes the computer program instructions stored in the memory 502 to perform the steps of the multidimensional analysis method for wind turbine vibration data provided in this embodiment of the present disclosure.

[0092] In one example, the multidimensional analysis device for wind turbine vibration data may also include a transceiver 503 and a bus 504. For example, Figure 5As shown, the processor 501, memory 502 and transceiver 503 are connected via bus 504 and communicate with each other.

[0093] Bus 504 may include hardware, software, or both. For example, and not limitingly, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 504 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0094] This disclosure also provides a computer-readable storage medium that can store a computer program. When the computer program is executed by a processor, the processor enables the processor to implement the multi-dimensional analysis method for wind turbine vibration data provided in this disclosure.

[0095] The aforementioned storage medium may, for example, include a memory 502 containing computer program instructions, which can be executed by the processor 501 of the wind turbine vibration data multidimensional analysis device to complete the wind turbine vibration data multidimensional analysis method provided in this embodiment. Optionally, the storage medium may be a non-transitory computer-readable storage medium, such as a ROM, random access memory (RAM), compact disc ROM (CD-ROM), magnetic tape, floppy disk, and optical data storage device.

[0096] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the term "comprising" is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0097] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for multidimensional analysis of vibration data of wind turbine generators, characterized in that, include: The original vibration signal of the wind turbine and the external operating parameters acquired synchronously are obtained, and the original vibration signal is preprocessed. The preprocessed vibration signal is analyzed using continuous wavelet transform to generate a wavelet coefficient matrix; A three-dimensional time-frequency spectrum is constructed based on the wavelet coefficient matrix, and the external operating condition parameters are integrated into the three-dimensional time-frequency spectrum in a visual form; By analyzing the fused three-dimensional time-frequency spectrum, the operating status and potential fault characteristics of wind turbine units can be identified.

2. The method according to claim 1, characterized in that, The preprocessing of the original vibration signal includes: The original vibration signal is converted into a one-dimensional numerical array; The one-dimensional numerical array is subjected to moving average filtering based on a preset filtering window length.

3. The method according to claim 1, characterized in that, The analysis of the preprocessed vibration signal using continuous wavelet transform to generate a wavelet coefficient matrix includes: The preprocessed vibration signal is convolved with continuous wavelet transforms at different scales and time positions to obtain the corresponding convolution calculation results. The convolution calculation results are normalized to generate a wavelet coefficient matrix.

4. The method according to claim 3, characterized in that, The method further includes: In response to the user's scale range selection operation, the scale parameters of the continuous wavelet transform are determined, wherein the scale range selection operation is used to select either a small scale range or a large scale range.

5. The method according to claim 4, characterized in that, The construction of the three-dimensional time-frequency spectrum based on the wavelet coefficient matrix includes: Based on the wavelet coefficient matrix, the time parameter is determined as the horizontal axis based on the time series of the original vibration signal, and the frequency parameter is calculated as the vertical axis based on the scale parameter to construct the three-dimensional time-frequency spectrum.

6. The method according to claim 1, characterized in that, The step of integrating the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form includes: Align the external operating parameters with the original vibration signal on the time axis; The aligned external operating parameters are plotted as independent curves overlaid on the three-dimensional time-frequency spectrum.

7. The method according to claim 1, characterized in that, The analysis of the fused three-dimensional time-frequency spectrum identifies the operating status and potential fault characteristics of wind turbine units, including: By analyzing the fused three-dimensional time-frequency spectrum, the accumulation characteristics of vibration energy in the target frequency range, the trend of change over time, and the correlation with operating parameters are determined, thereby identifying the operating status and potential fault characteristics of the wind turbine.

8. A multidimensional analysis device for vibration data of wind turbine generators, characterized in that, include: The data acquisition module is used to acquire the original vibration signal of the wind turbine and the synchronously acquired external operating parameters, and to preprocess the original vibration signal. The first processing module is used to analyze the preprocessed vibration signal using continuous wavelet transform and generate a wavelet coefficient matrix. The second processing module is used to construct a three-dimensional time-frequency spectrum based on the wavelet coefficient matrix, and to integrate the external operating condition parameters into the three-dimensional time-frequency spectrum in a visual form; The third processing module is used to analyze the fused three-dimensional time-frequency spectrum to identify the operating status and potential fault characteristics of the wind turbine.

9. A multidimensional analysis device for vibration data of wind turbine generators, characterized in that, include: processor; Memory, used to store executable instructions; The processor is configured to read the executable instructions from the memory and execute the executable instructions to implement the multidimensional analysis method for wind turbine vibration data as described in any one of claims 1-7.

10. A non-volatile computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, causes the processor to implement the multidimensional analysis method for wind turbine vibration data as described in any one of claims 1-7.