A Fault Diagnosis Method, Device and Electronic Equipment for Wind Turbine Generators

Abstract

The present invention discloses a fault diagnosis method, device and electronic device for a wind turbine. The method includes: obtaining various types of fault vibration data corresponding to the wind turbine; for any type of fault vibration data, obtaining the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data; obtaining target impact signals with peak values greater than a preset peak value threshold and extracting impact segments with a preset length from the fault vibration data according to the positions where the peak values of each target impact signal are located; performing frequency domain transformation on the extracted impact segments corresponding to each target impact signal to obtain the spectrogram of each impact segment; superimposing the spectrograms of each impact segment and determining the target frequency band energy range in the superimposed spectrogram; constructing fault diagnosis comparison data for the wind turbine according to each fault type in the wind turbine and the corresponding target frequency band energy range, so as to realize accurate diagnosis of non-periodic impact faults.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine fault diagnosis, and particularly relates to a wind turbine fault diagnosis method, device and electronic equipment. Background Art

[0002] With the development of the wind power industry, the fault diagnosis technology based on vibration signals is more and more widely applied. At present, vibration sensors and data acquisition devices are installed on most components of wind turbines, and various types of faults of wind turbines can be diagnosed and warned by analyzing vibration data. At present, the wind power vibration fault diagnosis can only perform time-domain analysis and frequency-spectrum analysis on periodic signals or quasi-periodic signals. For faults such as component cracks, cracks, wear and collisions, the vibration signals usually have irregular impacts, and conventional time-domain and frequency-spectrum analysis methods cannot analyze such data. Therefore, it is urgent to propose a new wind turbine fault diagnosis method to ensure accurate diagnosis of faults with non-periodic impacts. Summary of the Invention

[0003] Therefore, the technical problem to be solved by the present invention is to overcome the defect that for faults such as component cracks, the vibration signals have irregular impacts and conventional time-domain and frequency-spectrum analysis cannot analyze such data, so as to provide a wind turbine fault diagnosis method, device and electronic equipment.

[0004] According to a first aspect, an embodiment of the present invention discloses a wind turbine fault diagnosis method, the method comprising: acquiring various types of fault vibration data corresponding to a wind turbine; for any type of fault vibration data, acquiring the peak value of an impact signal in the fault vibration data and the position of the peak value in the fault vibration data; acquiring target impact signals with peak values greater than a preset peak value threshold and extracting impact segments with a preset length from the fault vibration data according to the position of the peak value of each target impact signal; performing frequency-domain transformation on each extracted impact segment corresponding to a target impact signal to obtain a spectrogram of each impact segment; superimposing the spectrograms of each impact segment and determining an energy range of a target frequency band in the superimposed spectrogram; and constructing wind turbine fault diagnosis comparison data according to each type of fault in the wind turbine and the corresponding energy range of the target frequency band.

[0005] Optionally, the extracting impact segments with a preset length from the fault vibration data according to the position of the peak value of each target impact signal comprises: taking the position of the peak value of each target impact signal as a center, respectively extracting vibration sub-data with a target length from the left and right sides; and taking the extracted vibration sub-data as the impact segments.

[0006] Optionally, obtaining the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data includes: dividing the fault vibration data into multiple equal parts and sequentially numbering each equal part; determining the position of the peak value in the fault vibration data according to the length of each equal part, the number of the equal part where the peak value is located, and the coordinate of the peak value in the equal part.

[0007] Optionally, before obtaining the peak value of the impact signal in the fault vibration data, it includes: obtaining the crest factor in the fault vibration data; comparing the crest factor with a preset crest factor threshold; when the crest factor is greater than or equal to the preset crest factor threshold, there is the impact signal in the fault vibration data.

[0008] Optionally, the method further includes: training a preset machine learning model according to the constructed wind turbine fault diagnosis comparison data to obtain a fault diagnosis model; when obtaining the vibration data of the wind turbine to be detected, processing the vibration data to obtain the corresponding target frequency band energy range; inputting the target frequency band energy range into the fault diagnosis model to obtain the fault type existing in the wind turbine.

[0009] According to a second aspect, an embodiment of the present invention also discloses a wind turbine fault diagnosis device, and the device includes: a data acquisition module, configured to acquire various types of fault vibration data corresponding to a wind turbine; a peak position acquisition module, configured to, for any type of fault vibration data, acquire the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data; an impact segment extraction module, configured to acquire a target impact signal with a peak value greater than a preset peak value threshold and extract an impact segment with a preset length from the fault vibration data according to the position where the peak value of each target impact signal is located; a frequency domain transformation module, configured to perform a frequency domain transformation on each extracted impact segment corresponding to a target impact signal to obtain a spectrogram of each impact segment; an energy range determination module, configured to superimpose the spectrograms of each impact segment and determine the target frequency band energy range in the superimposed spectrogram; a diagnosis data construction module, configured to construct wind turbine fault diagnosis comparison data according to each fault type in the wind turbine and the corresponding target frequency band energy range.

[0010] Optionally, the peak position acquisition module includes: a numbering sub-module, configured to divide the fault vibration data into multiple equal parts and sequentially number each equal part; a position calculation sub-module, configured to determine the position of the peak value in the fault vibration data according to the length of each equal part, the number of the equal part where the peak value is located, and the coordinate of the peak value in the equal part.

[0011] Optionally, the device further includes: a crest factor acquisition module, configured to acquire the crest factor in the fault vibration data; a comparison module, configured to compare the crest factor with a preset crest factor threshold; and an impact signal determination module, configured to determine that the impact signal exists in the fault vibration data when the crest factor is greater than or equal to the preset crest factor threshold.

[0012] According to a third aspect, an embodiment of the present invention further discloses an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the steps of the wind turbine fault diagnosis method according to the first aspect or any optional implementation manner of the first aspect.

[0013] According to a fourth aspect, an embodiment of the present invention further discloses a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the wind turbine fault diagnosis method according to the first aspect or any optional implementation manner of the first aspect.

[0014] The technical solution of the present invention has the following advantages:

[0015] The wind turbine fault diagnosis method provided by the present invention acquires the peak value of the impact signal in the fault diagnosis data and the position of the peak value in the fault vibration data, extracts the impact segment from the fault vibration data according to the position where the peak value of the impact signal is located for frequency domain transformation to obtain a spectrogram, determines the target frequency band energy range in the superimposed spectrogram according to the spectrogram, and constructs wind turbine fault diagnosis comparison data according to any fault type and the corresponding target frequency band energy range. Through the constructed wind turbine fault diagnosis comparison data, the present invention can not only analyze periodic impact signals, but also analyze non-periodic signals. For data with irregular impact signals, the energy range of the target frequency band of the impact segment in the vibration data can be extracted and analyzed and compared with the wind turbine fault diagnosis comparison data to achieve accurate diagnosis of non-periodic impact faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of a specific example of the wind turbine fault diagnosis method in the embodiment of the present invention;

[0018] Figure 2 It is a principle block diagram of a specific example of the wind turbine fault diagnosis device in the embodiment of the present invention;

[0019] Figure 3 It is a specific example diagram of the electronic device in the embodiment of the present invention. Detailed implementation manners

[0020] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0022] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and can also be the communication inside two elements. It can be a wireless connection or a wired connection. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0023] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0024] The embodiment of the present invention discloses a wind turbine fault diagnosis method. As Figure 1 shown, the method includes the following steps:

[0025] Step S1, obtain various types of fault vibration data corresponding to a wind turbine; exemplarily, install vibration sensors and data acquisition devices on different components of the wind turbine to collect corresponding vibration data and obtain fault vibration data therefrom, or obtain it from the data storing various types of historical fault vibration data of the wind turbine. The specific obtaining method can be determined according to the actual situation by itself.

[0026] Step S2, for any type of fault vibration data, obtain the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data;

[0027] Exemplarily, the embodiment of the present application does not limit the method for obtaining the position of the peak value in the fault vibration data. The embodiment of the present application calculates the position of the impact signal in the fault vibration data, finds the coordinates of the peak value in the impact signal, so as to calculate the position of the peak value in the fault vibration data. In a specific embodiment, the coordinates of the peak value of the impact signal can be directly read in the device for obtaining vibration data.

[0028] Step S3, obtain target impact signals with peak values greater than a preset peak value threshold, and extract impact segments with a preset length from the fault vibration data according to the position where the peak value of each target impact signal is located;

[0029] Exemplarily, in the embodiment of the present application, a preset peak value threshold is set (this preset peak value threshold is an adjustable parameter, and those skilled in the art can determine it according to actual needs). In the embodiment of the present application, this preset peak value threshold is 3*R; where R is the effective value of the vibration data, and can be calculated by the following formula:

[0030]

[0031] where N is the number of data points in the obtained fault vibration data, and X i is the i-th fault vibration data, where i = 1, 2,..., n.

[0032] When the maximum peak value of each impact signal is greater than the set preset peak value threshold, it means that the peak value needs to be extracted. In the embodiment of the present application, centered on the position of the peak value in the fault vibration data obtained in step S2, an impact segment with a preset length (this preset length is also an adjustable parameter) is extracted from the fault vibration data. This impact segment can include a segment with S vibration data points, and all the extracted impact segments are stored. By adjusting the preset peak value threshold of the impact signal, the number of target impact signals extracted can be adjusted, making the method adjustable.

[0033] Step S4: Perform frequency-domain transformation on each extracted impact segment corresponding to the target impact signal to obtain the spectrogram of each impact segment. Exemplarily, in the embodiments of the present application, the frequency-domain transformation of the extracted impact segment may, but is not limited to, use the windowed Fourier transform to obtain the spectrogram. The selection of different window functions is determined according to the actual situation. In the embodiments of the present application, the Hanning window is used for Fourier transform.

[0034] Step S5: Superimpose the spectrograms of each impact segment and determine the target frequency band energy range in the superimposed spectrogram.

[0035] Exemplarily, in the embodiments of the present application, the spectrograms obtained by performing frequency-domain transformation on each extracted impact segment in sequence are superimposed. According to the superimposed spectrogram, auto-power spectrum analysis is performed (the method of analyzing the spectrogram is only for example and is not limited thereto), the frequency band energy range of multiple impact signal segments is determined, and the target frequency band energy range is obtained therefrom. In the embodiments of the present application, the target frequency band is the frequency band in which the energy value within the target frequency band range is greater than the preset energy value threshold, where the preset energy threshold is determined according to actual requirements.

[0036] Step S6: Construct the fault diagnosis comparison data of the wind turbine according to each fault type in the wind turbine and the corresponding target frequency band energy range.

[0037] Exemplarily, in the embodiments of the present application, the above steps are repeated for each known type of fault to determine the target frequency band energy range of each fault type, and then a diagnostic comparison data list for each fault type is constructed, so that the faults of the wind turbine can be diagnosed according to the normal data and the diagnostic comparison data list of each fault type.

[0038] The wind turbine fault diagnosis method provided by the present invention calculates the position of the peak value of the impact signal in the fault vibration data in the fault vibration data, extracts the impact segment of the preset length from the fault vibration data according to the position of the peak value greater than the preset peak threshold, performs frequency-domain transformation on each extracted impact segment to obtain the spectrogram of the impact segment, superimposes the spectrograms of each impact segment and determines the target frequency band energy range in the superimposed spectrogram, constructs the fault diagnosis comparison data of the wind turbine according to each fault type of the wind turbine and the corresponding target frequency band energy range. Through the constructed fault diagnosis comparison data of the wind turbine, not only can periodic impact signals be analyzed, but also non-periodic signals can be analyzed. For data with irregular impact signals, the energy range of the target frequency band of the impact segment in the vibration data can be extracted and compared with the fault diagnosis comparison data of the wind turbine to achieve accurate diagnosis of non-periodic impact faults.

[0039] As an alternative embodiment of the present invention, step S3 includes: taking the position where the peak of each target impact signal is located as the center, and extracting the vibrator data of the target length from the left and right sides respectively; using the extracted vibrator data as the impact segment.

[0040] Exemplarily, the embodiment of the present application does not limit the method of intercepting the vibrator data of the target length. The extracted vibrator data may contain not only impact signals but also periodic signals or quasi-periodic signals. The vibrator data corresponding to the left and right sides extracted are merged to obtain the impact segment. The embodiment of the present application does not limit this target length, and those skilled in the art can determine it according to actual needs. In the embodiment of the present application, the vibrator data including 0.55 faulty vibration data points on each of the left and right sides is selected.

[0041] As an alternative embodiment of the present invention, to obtain the peak of the impact signal in the faulty vibration data and the position of the peak in the faulty vibration data, step S2 includes: dividing the faulty vibration data into multiple equal parts and numbering each equal part; determining the position of the peak in the faulty vibration data according to the length of each equal part, the number of the equal part where the peak is located, and the coordinate of the peak in the equal part. In the embodiment of the present application, the faulty vibration data is divided into multiple equal parts to calculate the position of the peak in the faulty vibration data, ensuring that the preset peak threshold does not need to be too high, avoiding missing impact segments with smaller peaks, and the number of extracted impact segments can be controlled by adjusting the number of divided equal parts.

[0042] Exemplarily, in the embodiment of the present application, the faulty vibration data is divided into multiple equal parts. The coordinate of the peak in this equal part is obtained as m, the coordinate length of each equal part is S, and the sequence of this equal part in the faulty vibration data is calculated as k. Then the coordinate of the peak in the faulty vibration data is m+(k - 1)*S.

[0043] As an alternative embodiment of the present invention, before obtaining the peak of the impact signal in the faulty vibration data, it includes: obtaining the crest factor in the faulty vibration data; comparing the crest factor with the preset crest factor threshold; when the crest factor is greater than or equal to the preset crest factor threshold, there is an impact signal in the faulty vibration data.

[0044] Exemplarily, the crest factor in the faulty vibration data can be calculated by the following formula:

[0045] C = Pk / R

[0046] where C is the crest factor, Pk = MAX(D), representing the maximum peak in the faulty vibration data, D represents the peak in the faulty vibration data, and R is the effective value of the vibration data.

[0047] Determine the crest factor threshold according to the actual situation. In the embodiments of the present application, the crest factor threshold can be set to 3. Compare the crest factor of the calculated fault vibration data with the set crest factor threshold. If the calculated crest factor is greater than or equal to the set crest factor threshold, there is an impact signal in the fault data. If the calculated crest factor is less than the set crest factor threshold, it means that there is no impact signal in the vibration data.

[0048] As an alternative embodiment of the present invention, the method further includes: training a preset machine learning model according to the constructed wind turbine fault diagnosis comparison data to obtain a fault diagnosis model; when obtaining the vibration data of the wind turbine to be detected, processing the vibration data to obtain the corresponding target frequency band energy range; inputting the target frequency band energy range into the fault diagnosis model to obtain the fault types existing in the wind turbine. Training the preset machine learning model according to the fault comparison data improves the intelligence and accuracy of the wind turbine fault diagnosis, and reduces the labor cost.

[0049] The wind turbine fault diagnosis method provided by the embodiments of the present invention processes the extracted impact signal and performs corresponding spectrum and envelope analysis, and establishes a priori knowledge base according to the target frequency band energy range corresponding to different faults to warn different types of unit faults.

[0050] The embodiments of the present invention also disclose a wind turbine fault diagnosis device, as Figure 2 shown. The device includes: a data acquisition module 101 for acquiring various types of fault vibration data corresponding to the wind turbine; a peak position acquisition module 102 for, for any type of fault vibration data, acquiring the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data; an impact segment extraction module 103 for acquiring target impact signals with peak values greater than a preset peak threshold and extracting impact segments with a preset length from the fault vibration data according to the positions of the peak values of each target impact signal; a frequency domain transformation module 104 for performing frequency domain transformation on each extracted impact segment corresponding to the target impact signal to obtain the spectrogram of each impact segment; an energy range determination module 105 for superimposing the spectrograms of each impact segment and determining the target frequency band energy range in the superimposed spectrogram; a diagnosis data construction module 106 for constructing wind turbine fault diagnosis comparison data according to each fault type in the wind turbine and the corresponding target frequency band energy range.

[0051] The wind turbine fault diagnosis device provided by the present invention acquires various types of fault diagnosis data corresponding to the wind turbine. For any type of fault vibration data, it obtains the peak value of the impact signal and the position where the peak value is located, extracts an impact segment with a preset length from the fault vibration data centered on the peak position greater than the preset peak threshold, performs a frequency-domain transformation on the extracted impact segment to obtain the spectrogram of each impact segment, superimposes the spectrograms of each impact segment to determine the target frequency band energy range, and obtains the wind turbine fault diagnosis comparison data according to each fault type and the corresponding target frequency band energy range. Through the constructed wind turbine fault diagnosis comparison data, not only can periodic impact signals be analyzed, but also non-periodic signals can be analyzed. For data with irregular impact signals, the energy range of the target frequency band of the impact segment in the vibration data can be extracted and analyzed and compared with the wind turbine fault diagnosis comparison data to achieve accurate diagnosis of non-periodic impact faults.

[0052] As an optional implementation manner of the present invention, the impact segment extraction module 103 is further configured to extract vibration sub-data with a target length from the left and right sides respectively centered on the peak position of each target impact signal, and use the extracted vibration sub-data as the impact segment.

[0053] As an optional implementation manner of the present invention, the peak position acquisition module 102 further includes: a numbering sub-module for dividing the fault diagnosis data into multiple equal parts and sequentially numbering each equal part; a position calculation sub-module for calculating and determining the position of the peak value in the fault vibration data according to the length of each equal part, the number of the equal part where the peak value is located, and the coordinates of the peak value in the equal part.

[0054] As an optional implementation manner of the present invention, the device further includes: a crest factor acquisition module for acquiring the crest factor in the fault vibration data; a comparison module for comparing the crest factor with a preset crest factor threshold; an impact signal determination module for determining that there is an impact signal in the fault vibration data when the crest factor is greater than or equal to the preset crest factor threshold.

[0055] As an optional implementation manner of the present invention, the device further includes: a model training module for training a preset machine learning model according to the constructed wind turbine fault diagnosis comparison data to obtain a fault diagnosis model; a data processing module for processing the vibration data to obtain the corresponding target frequency band energy range when the vibration data of the wind turbine to be detected is acquired; a fault diagnosis module for inputting the target frequency band energy range into the fault diagnosis model to obtain the fault types existing in the wind turbine.

[0056] The embodiment of the present invention also provides an electronic device, such as Figure 3As shown, the electronic device may include a processor 401 and a memory 402, where the processor 401 and the memory 402 may be connected through a bus or other means. Figure 3 Take the connection through the bus as an example.

[0057] The processor 401 may be a central processing unit (CPU). The processor 401 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or a combination of the above types of chips.

[0058] As a non-transitory computer-readable storage medium, the memory 402 can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the wind turbine fault diagnosis method in the embodiments of the present invention. The processor 401 executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions, and modules stored in the memory 402, that is, implements the wind turbine fault diagnosis method in the above method embodiments.

[0059] The memory 402 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created by the processor 401, etc. In addition, the memory 402 may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 402 may optionally include a memory remotely set relative to the processor 401, and these remote memories may be connected to the processor 401 through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0060] The one or more modules are stored in the memory 402 and, when executed by the processor 401, execute the Figure 1 wind turbine fault diagnosis method in the embodiments shown.

[0061] For specific details of the above electronic device, reference may be made to Figure 1 the corresponding relevant descriptions and effects in the embodiments shown, and details are not described herein again.

[0062] Those skilled in the art can understand that to implement all or part of the processes in the above-described method embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-described method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memories.

[0063] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the defined scope.

Claims

1. A fault diagnosis method for a wind turbine unit, characterized in that, The method includes: Obtaining various types of fault vibration data corresponding to a wind turbine; For any type of fault vibration data, obtaining the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data; The obtaining the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data includes: dividing the fault vibration data into multiple equal parts and numbering each equal part; determining the position of the peak value in the fault vibration data according to the length of each equal part, the number of the equal part where the peak value is located, and the coordinates of the peak value in the equal part; Obtaining target impact signals with peak values greater than a preset peak value threshold and extracting impact segments with a preset length from the fault vibration data according to the positions of the peak values of each target impact signal; The extracting impact segments with a preset length from the fault vibration data according to the positions of the peak values of each target impact signal includes: taking the position of the peak value of each target impact signal as the center, extracting vibration sub-data with a target length from the left and right sides respectively; taking the extracted vibration sub-data as the impact segment; Performing frequency domain transformation on each extracted impact segment corresponding to a target impact signal to obtain a spectrogram of each impact segment; Superimposing the spectrograms of each impact segment and determining the target frequency band energy range in the superimposed spectrogram; Constructing wind turbine fault diagnosis comparison data according to each fault type in the wind turbine and the corresponding target frequency band energy range.

2. The wind turbine fault diagnosis method according to claim 1, wherein Before obtaining the peak value of the impact signal in the fault vibration data, it includes: Obtaining the crest factor in the fault vibration data; Comparing the crest factor with a preset crest factor threshold; When the crest factor is greater than or equal to the preset crest factor threshold, there is the impact signal in the fault vibration data.

3. The wind turbine fault diagnosis method according to claim 1, characterized in that, The method further includes: Training a preset machine learning model according to the constructed wind turbine fault diagnosis comparison data to obtain a fault diagnosis model; When obtaining the vibration data of a wind turbine to be detected, processing the vibration data to obtain the corresponding target frequency band energy range; Inputting the target frequency band energy range into the fault diagnosis model to obtain the fault types existing in the wind turbine.

4. A wind turbine fault diagnosis device, characterized in that, The device includes: A data acquisition module, configured to obtain various types of fault vibration data corresponding to a wind turbine; A peak position acquisition module, configured to, for any type of fault vibration data, obtain the peak value of the impact signal in the fault vibration data and the position of the peak value in the fault vibration data; The peak position acquisition module includes: a numbering sub-module, configured to divide the fault diagnosis data into multiple equal parts and number each equal part; a position calculation sub-module, configured to calculate and determine the position of the peak value in the fault vibration data according to the length of each equal part, the number of the equal part where the peak value is located, and the coordinates of the peak value in the equal part; An impact segment extraction module, configured to obtain target impact signals with peak values greater than a preset peak value threshold and extract impact segments with a preset length from the fault vibration data according to the positions of the peak values of each target impact signal; The impact segment extraction module is further configured to extract the vibrator data with a target length from the left and right sides respectively centered on the position where the peak value of each target impact signal is located, and use the extracted vibrator data as the impact segment; A frequency domain transformation module, configured to perform frequency domain transformation on the impact segment corresponding to each target impact signal to obtain the spectrogram of each impact segment; An energy range determination module, configured to superimpose the spectrograms of each impact segment and determine the target frequency band energy range in the superimposed spectrogram; A diagnostic data construction module, configured to construct the fault diagnosis comparison data of the wind turbine according to each fault type in the wind turbine and the corresponding target frequency band energy range.

5. The wind turbine fault diagnosis device according to claim 4, characterized in that The device further includes: A crest factor acquisition module, configured to acquire the crest factor in the fault vibration data; A comparison module, configured to compare the crest factor with a preset crest factor threshold; An impact signal determination module, configured to determine that there is the impact signal in the fault vibration data when the crest factor is greater than or equal to the preset crest factor threshold.

6. An electronic device, characterized in that, Including: At least one processor; And a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the steps of the wind turbine fault diagnosis method according to any one of claims 1-3.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the wind turbine fault diagnosis method according to any one of claims 1-3 are implemented.

Images