Wind turbine generator fault diagnosis method and device

Abstract

The invention provides a wind turbine generator fault diagnosis method and device. The method comprises the following steps: acquiring real-time operation condition data and real-time vibration signal data of a wind turbine generator; performing noise reduction processing on the real-time vibration signal data to obtain noise reduction data; performing frequency domain conversion on the noise reduction data to obtain frequency domain data; obtaining vibration prediction data according to the real-time operation condition data and a trained vibration prediction model; the vibration prediction model is obtained by training a preset network model according to historical operation condition data of the wind turbine generator and corresponding historical vibration signal data; and performing fault diagnosis on the wind turbine generator according to the frequency domain data and the vibration prediction data to obtain a fault diagnosis result. The fault diagnosis accuracy of the wind turbine generator can be effectively improved, and the operation efficiency of the wind turbine generator is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind turbine fault diagnosis, and further relates to a wind turbine fault diagnosis method and device. BACKGROUND

[0002] As a typical clean energy unit, wind turbine is an important part of power energy. In the development process of rotating machinery equipment, which is rapidly moving towards large-scale, complex, precision, high-speed, heavy-load and intelligent, the application scenarios of wind turbine are also more modern and diversified. For power generation enterprises, this is not only an opportunity to improve power generation efficiency and strengthen reform and innovation, but also a challenge to ensure safety in production and improve digitalization capability. With the progress of the times and the rapid improvement of technology, enterprises are increasingly concerned about the safety and sustainable operation of production equipment. Wind turbines operate in complex environments and often need to run under heavy load. Once a fault occurs, it will not only affect production safety and the power grid, but even threaten personal safety. Therefore, ensuring the safe operation of wind turbines during their life cycle, diagnosing and warning early mechanical faults, and preventing controllable faults from developing into major out-of-control faults can effectively avoid economic losses and is of great significance to the stability of the power grid.

[0003] However, the operating environment of wind turbines is complex, and the collected signals often have unknown intensity of noise, which causes certain difficulties in extracting fault features of wind turbines. If the energy of the noise signal is strong, the fault feature is likely to be submerged in the noise signal, which will cause inaccurate diagnosis of the fault. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a wind turbine fault diagnosis method and device to improve the accuracy of wind turbine fault diagnosis.

[0005] To solve the above technical problems, the technical solutions of the present application are as follows: In a first aspect, the present application provides a wind turbine fault diagnosis method, comprising: obtaining real-time operating condition data and real-time vibration signal data of a wind turbine; performing noise reduction processing on the real-time vibration signal data to obtain noise-reduced data; performing frequency domain conversion on the noise-reduced data to obtain frequency domain data; obtaining vibration prediction data according to the real-time operating condition data and a trained vibration prediction model; the vibration prediction model is obtained by training a preset network model according to historical operating condition data and corresponding historical vibration signal data of the wind turbine; performing fault diagnosis on the wind turbine according to the frequency domain data and the vibration prediction data to obtain a fault diagnosis result.

[0006] Optionally, the real-time running condition data and the real-time vibration signal data of the wind turbine generator are acquired, including: acquiring real-time running condition data of the wind turbine generator; the running condition data includes temperature parameter data, electrical parameter data, environmental parameter data and control parameter data; acquiring real-time vibration signal data of the wind turbine generator through sensors installed at preset positions in the wind turbine generator.

[0007] Optionally, the real-time vibration signal data is subjected to noise reduction processing to obtain noise reduction data, including: decomposing the real-time vibration signal data to obtain low-frequency components and high-frequency components; performing threshold processing on the high-frequency components to obtain detail coefficients; obtaining noise reduction data according to the low-frequency components and the detail coefficients.

[0008] Optionally, the noise reduction data is subjected to frequency domain conversion to obtain frequency domain data, including: by subjecting the noise reduction data to frequency domain conversion to obtain frequency domain data; wherein, is the frequency domain data, is the noise reduction data, n is a time domain index, N is the total number of sampling points, k is a frequency domain index, , is a rotation factor.

[0009] Optionally, vibration prediction data is obtained according to the real-time running condition data and the trained vibration prediction model, including: inputting the real-time running condition data into an input layer of the vibration prediction model to obtain a first output result; inputting the first output result into at least one processing layer of the vibration prediction model to obtain a second output result; inputting the second output result into an output layer of the vibration prediction model to obtain vibration prediction data.

[0010] Optionally, the training process of the vibration prediction model includes: acquiring historical running condition data and corresponding historical vibration signal data of the wind turbine generator under preset conditions; preprocessing the historical running condition data and the historical vibration signal data to obtain preprocessed historical data; extracting historical feature data from the preprocessed historical data; training a preset network model according to the historical feature data to obtain an initial training result; According to the initial training result, a preset network model is optimized to obtain a vibration prediction model.

[0011] Optionally, according to the frequency domain data and the vibration prediction data, a fault diagnosis is performed on the wind turbine generator to obtain a fault diagnosis result, including: According to the frequency domain data and the vibration prediction data, a residual error is obtained. According to the residual error and a preset warning threshold, a warning result is determined. According to the warning result and a preset fault scheme library, a fault scheme is obtained. According to the frequency domain data, the vibration prediction data, the warning result and the fault scheme, a fault diagnosis is performed on the wind turbine generator to obtain a fault diagnosis result.

[0012] In a second aspect of the present application, a wind turbine generator fault diagnosis device is provided, including: An acquisition module is configured to acquire real-time operating condition data and real-time vibration signal data of a wind turbine generator. A processing module is configured to perform noise reduction processing on the real-time vibration signal data to obtain noise reduction data, perform frequency domain conversion on the noise reduction data to obtain frequency domain data, obtain vibration prediction data according to the real-time operating condition data and a trained vibration prediction model, wherein the vibration prediction model is obtained by training a preset network model according to historical operating condition data and corresponding historical vibration signal data of the wind turbine generator, and perform fault diagnosis on the wind turbine generator according to the frequency domain data and the vibration prediction data to obtain a fault diagnosis result.

[0013] In a third aspect of the present application, a computing device is provided, including a processor and a memory storing a computer program, wherein the computer program is executed by the processor to perform the method of the first aspect.

[0014] In a fourth aspect of the present application, a computer readable storage medium is provided, storing instructions, wherein the instructions are executed on a computer to cause the computer to perform the method of the first aspect.

[0015] The above-mentioned scheme of the present application has at least the following beneficial effects: The above-mentioned scheme of the present application can effectively improve the fault diagnosis accuracy of the wind turbine generator and improve the operating efficiency of the wind turbine generator by acquiring real-time operating condition data and real-time vibration signal data of the wind turbine generator, performing noise reduction processing on the real-time vibration signal data to obtain noise reduction data, performing frequency domain conversion on the noise reduction data to obtain frequency domain data, obtaining vibration prediction data according to the real-time operating condition data and a trained vibration prediction model, and comparing the frequency domain data and the vibration prediction data to perform fault diagnosis on the wind turbine generator to obtain a fault diagnosis result. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is a flow diagram of a wind turbine fault diagnosis method in an embodiment of the present application; Figure 2 is a structural diagram of a wind turbine fault diagnosis device in an embodiment of the present application. DETAILED DESCRIPTION

[0017] Exemplary embodiments of the present application will be described herein below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood, and so that the scope of the present application can be conveyed to those skilled in the art.

[0018] As shown in Figure 1 , an embodiment of the present application proposes a wind turbine fault diagnosis method, comprising the following steps: Step 101, acquiring real-time operating condition data and real-time vibration signal data of a wind turbine; Step 102, performing noise reduction processing on the real-time vibration signal data to obtain noise reduction data; Step 103, performing frequency domain conversion on the noise reduction data to obtain frequency domain data; Step 104, obtaining vibration prediction data according to the real-time operating condition data and a trained vibration prediction model; the vibration prediction model is obtained by training a preset network model according to historical operating condition data and corresponding historical vibration signal data of the wind turbine; Step 105, performing fault diagnosis on the wind turbine according to the frequency domain data and the vibration prediction data to obtain a fault diagnosis result.

[0019] The wind turbine fault diagnosis method proposed in the embodiment of the present application can effectively improve the fault diagnosis accuracy of the wind turbine and improve the operating efficiency of the wind turbine by acquiring real-time operating condition data and real-time vibration signal data of the wind turbine, then performing noise reduction processing on the real-time vibration signal data to obtain noise reduction data, performing frequency domain conversion on the noise reduction data to obtain frequency domain data, obtaining vibration prediction data according to the real-time operating condition data and a trained vibration prediction model, and comparing the frequency domain data and the vibration prediction data to perform fault diagnosis on the wind turbine and obtain a fault diagnosis result.

[0020] In an optional embodiment of the present application, step 101 comprises: Step 1011, acquiring real-time operating condition data of a wind turbine; the operating condition data comprises temperature parameter data, electrical parameter data, environmental parameter data and control parameter data; Specifically, the real-time operation condition data of the wind turbine is the basis for subsequent prediction of mechanical vibration signal data, and also provides a diagnosis basis for subsequent fault diagnosis. Here, temperature parameter data such as gear box oil temperature, generator winding temperature and bearing temperature can be collected by a temperature sensor, electrical parameter data such as power, voltage, current and frequency can be collected by a power analyzer or a current transformer, wind speed, wind direction and air density data can be collected by a wind speed meter, and temperature, humidity and air pressure data in the environment can be collected by a temperature and humidity sensor or a barometer. Here, the environmental parameter data can include wind speed, wind direction, air density and the like, as well as temperature, humidity, air pressure and the like, and control parameter data such as blade pitch angle, yaw angle and rotational speed can be collected by an encoder or an angle sensor. According to the actual application scenario, the real-time operation condition data of other types of wind turbines can be collected, and the above is only an example.

[0021] In step 1012, real-time vibration signal data of the wind turbine is obtained by a sensor installed at a preset position in the wind turbine.

[0022] Specifically, the preset position can include at least one key part such as a main bearing, a gear box input shaft, a gear box output shaft, a generator driving end / non-driving end, and a sensor is installed at the key part to collect at least one mechanical vibration signal data of the main bearing, the gear box input shaft, the gear box output shaft and the generator driving end / non-driving end, thereby providing a data basis for subsequent fault diagnosis.

[0023] In an optional embodiment of the present application, step 102 includes: In step 1021, the real-time vibration signal data is decomposed to obtain a low-frequency component and a high-frequency component. Specifically, the real-time vibration signal data is decomposed by to obtain the low-frequency component, and the real-time vibration signal data is decomposed by to obtain the high-frequency component, wherein, the low-frequency component is obtained by the coefficient of the low-pass filter, is a part of the downsampling and filtering operation on the real-time vibration signal data, the high-frequency component is obtained by the coefficient of the high-pass filter, n is a data index, and m is an index of the filter coefficient.

[0024] In step 1022, the high-frequency component is threshold processed to obtain a detail coefficient. Specifically, the high-frequency component is threshold processed by or to obtain the detail coefficient, wherein, the detail coefficient is obtained by the high-frequency component is obtained by the preset threshold, , is a standard deviation of noise, M is a length of signal to determine, is a symbol function.

[0025] In step 1023, the de-noising data is obtained according to the low-frequency component and the detail coefficient.

[0026] Specifically, the de-noising data is obtained by wherein, is a low-frequency component at the b-1th level, is a coefficient of a reconstruction low-pass filter, is a low-frequency component at the bth level, is a coefficient of a reconstruction high-pass filter, is a down rounding function, used for implementing an up-sampling operation, is a detail coefficient at the bth level, m is an index of a filter coefficient, and n is a data index.

[0027] Through the de-noising processing of the above steps, the mechanical vibration signal data with high accuracy can be obtained, which is beneficial to providing the accuracy of subsequent fault diagnosis.

[0028] In an optional embodiment of the present application, step 103 comprises: Through the frequency domain conversion is performed on the de-noising data to obtain frequency domain data; wherein, is frequency domain data, is de-noising data, n is a time domain index, , N is a total number of sampling points, and k is a frequency domain index, , is a rotation factor.

[0029] Specifically, the frequency domain data comprises frequency and amplitude, and is basic data for subsequent fault diagnosis.

[0030] In an optional embodiment of the present application, the vibration prediction data is obtained according to the real-time running condition data and the trained vibration prediction model in step 104, comprising: In step 10411, the real-time running condition data is input into an input layer of the vibration prediction model to obtain a first output result; Specifically, the input layer of the vibration prediction model performs normalization processing on the real-time running condition data to eliminate dimension influence and improve processing efficiency, and here, the input layer adopts The real-time running condition data is normalized to obtain the first output result, wherein, is a first output result, x is real-time running condition data, is a minimum value of the real-time running condition data, to obtain a maximum value of the real-time operating condition data.

[0031] Step 10412, input the first output result into at least one processing layer of the vibration prediction model to obtain a second output result; Specifically, the processing layer of the vibration prediction model performs nonlinear transformation on the first output result to extract high-order features. Here, the processing layer performs nonlinear transformation on the first output result to extract high-order features by nonlinear transformation on the first output result to obtain a second output result, wherein, is the output result of the jth neuron in the kth processing layer, is an activation function, is a weighted sum of the output results of all neurons in the k-1th layer, is the connection weight from the ith neuron in the k-1th layer to the jth neuron in the kth layer, is the output of the ith neuron in the k-1th layer, and p is the number of neurons in the k-1th processing layer, is the bias term of the jth neuron in the kth layer. It should be noted that the output of the last processing layer is the second output result. If the vibration prediction model has k processing layers, the output results of all neurons in the kth processing layer are the second output results.

[0032] Step 10413, input the second output result into an output layer of the vibration prediction model to obtain vibration prediction data.

[0033] Specifically, the output layer of the vibration prediction model obtains vibration prediction data by obtaining vibration prediction data, wherein, is the vibration prediction data, is the second output result, is the connection weight from the neuron in the kth processing layer to the output layer, M is the number of neurons in the kth processing layer, and b is the bias term of the output layer.

[0034] In an optional embodiment of the present application, the training process of the vibration prediction model in step 104 includes: Step 10421, obtaining historical operating condition data and corresponding historical vibration signal data of the wind turbine under a preset condition; Specifically, the historical operating condition data and corresponding historical vibration signal data of the wind turbine under normal operation (i.e., under the preset condition without failure) can be obtained from the management database or platform of the wind turbine as training samples to train the preset network model.

[0035] Step 10422, preprocessing the historical operating condition data and the historical vibration signal data to obtain preprocessed historical data; Specifically, the preprocessing mode can be at least one of data cleaning, deleting abnormal values, and normalization, so as to improve the accuracy of the historical data and the accuracy of subsequent model training.

[0036] In step 10423, historical feature extraction is performed on the preprocessed historical data to obtain historical feature data. Specifically, the historical frequency domain data can be extracted from the historical vibration signal data in the preprocessed historical data in a manner as in steps 102 and 103, and the historical operating condition data in the preprocessed historical data are combined to form the historical feature data.

[0037] In step 10424, the preset network model is trained according to the historical feature data to obtain an initial training result. Specifically, according to the number of features of the historical operating condition data in the historical feature data, the number of neurons of each processing layer of the preset network model is set, for example, if the number of features of the historical operating condition data is 3, each processing layer of the preset network model contains 3 neurons, and the number of processing layers can be 1 to 3. The historical feature data is divided into a training set, a validation set, and a test set according to a ratio of 70%, 15%, and 15%. The training set is input into the preset network model for training. During the training process, the parameters of the preset network model, such as the weights and bias terms of the processing layers and the weights and bias terms of the output layer, are adjusted and optimized. After the training is completed, a preliminarily trained preset network model is obtained. Then, the performance of the preliminarily trained preset network model is verified using the validation set. During the verification process, the loss function is monitored. If the loss function starts to rise, it means that the model may be overfitting, and the training needs to be stopped to prevent overfitting. Wherein, is the loss function, is the historical vibration signal data, is the predicted vibration signal data. After the performance of the model is verified using the validation set, the trained preset network model, i.e., the initial training result, is obtained.

[0038] In step 10425, the preset network model is optimized according to the initial training result to obtain a vibration prediction model.

[0039] Specifically, the test set is input into the preset network model in the initial training result for testing. If the loss function is lower than a preset value (such as 0.01), it means that the performance of the model is good, and the model can be used as a vibration prediction model. Otherwise, data needs to be re-collected for training.

[0040] In an optional embodiment of the present application, step 105 includes: In step 1051, residual error is obtained according to the frequency domain data and the vibration prediction data. Specifically, according to The residual is calculated, wherein e is y is The vibration prediction data is, and the purpose of calculating the residual is to determine the difference between the frequency domain data and the vibration prediction data, thereby providing a basis for subsequent diagnosis of whether the wind turbine has a fault.

[0041] Step 1052, determining an early warning result according to the residual and a preset early warning threshold; Specifically, if the residual is greater than the preset early warning threshold, it indicates that the wind turbine has a fault, and early warning is needed. The early warning result can include a comparison result of the residual and the preset early warning threshold and a corresponding early warning mode.

[0042] Step 1053, obtaining a fault scheme according to the early warning result and a preset fault scheme library; Specifically, the preset fault scheme library is set according to historical experience, and includes a fault type, a fault mechanism, a characteristic phenomenon and the like corresponding to the early warning result, and can also include a corresponding historical fault scheme, so as to facilitate rapid fault diagnosis and determination of a solution for the wind turbine.

[0043] Step 1054, performing fault diagnosis on the wind turbine according to the frequency domain data, the vibration prediction data, the early warning result and the fault scheme, to obtain a fault diagnosis result.

[0044] Specifically, the fault diagnosis result includes the frequency domain data, the vibration prediction data, the early warning result and the fault scheme, so that a management personnel can intuitively and quickly understand the fault of the wind turbine according to the fault diagnosis result and improve the fault solution efficiency.

[0045] One specific embodiment of the wind turbine fault diagnosis method of the embodiment of the present application includes: Step 111, acquiring real-time running condition data and real-time vibration signal data of the wind turbine; The temperature parameter data, the electrical parameter data, the environmental parameter data and the control parameter data and the like of the real-time running condition data of the wind turbine are acquired, thereby providing a data basis for subsequent prediction of the mechanical vibration signal data; the real-time vibration signal data of the wind turbine is acquired through a sensor installed at a preset position in the wind turbine, thereby providing a data basis for subsequent fault diagnosis.

[0046] Step 112, performing noise reduction processing on the real-time vibration signal data to obtain noise reduction data; The real-time vibration signal data is denoised to reduce noise interference and improve the accuracy of the data. The real-time vibration signal data can be denoised by wavelet denoising and ensemble empirical mode decomposition threshold denoising. If the abnormal interference during the data acquisition of the real-time vibration signal data causes the time domain waveform to have a jump point, the peak-to-peak value will be larger. In this case, the waveform signal needs to be filtered to remove the jump point data, that is, to find the periodic impact signal, and then to perform denoising.

[0047] In step 113, the denoised data is converted into frequency domain data. The denoised vibration data is subjected to feature extraction (frequency domain conversion), and the empirical mode decomposition and its optimization algorithm (such as periodic impact signal extraction) are used for feature extraction to obtain early signals that can reflect the mechanical fault characteristics of the wind turbine generator. The filtered high-frequency waveform signal is subjected to feature extraction, and the characteristic frequencies in the frequency spectrum and envelope spectrum are extracted in combination with the fault type, such as 1X amplitude, 2X amplitude, BPFI (rolling body outer ring fault frequency), BPFO (rolling body inner ring fault frequency), meshing frequency, frequency band noise, vibration phase, etc. The denoised data is a one-dimensional array, , and the frequency domain data is a frequency and amplitude array, such as [[0, 0.01], [0.75, 0.11], [1.5, 0.13], …]. For the frequency domain data, in combination with the wind turbine equipment parameter information and the real-time speed, the required characteristic components (i.e. amplitudes) can be extracted by transforming the results, such as the real-time speed of 600 rpm. The frequency corresponding to the 1X amplitude is 60 Hz, and the amplitude corresponding to the 60 Hz frequency index is the 1X characteristic component.

[0048] In step 114, vibration prediction data is obtained based on the real-time operating condition data and the trained vibration prediction model. First, the preset network model is trained based on the historical operating data of the wind turbine generator to obtain a vibration prediction model. The vibration prediction model obtains vibration prediction data based on real-time operating condition data, In step 115, the wind turbine generator is subjected to fault diagnosis based on the frequency domain data and the vibration prediction data to obtain a fault diagnosis result.

[0049] Residual analysis is performed on the frequency domain data and the vibration prediction data, and according to the 3σ rule, if the deviation exceeds 3σ at multiple times, it is considered that the wind turbine generator is abnormal and a warning signal is issued. According to the warning signal, in combination with the dynamics mechanism of the wind power equipment, the precise positioning and early warning of common mechanical faults of the wind turbine generator are realized. The expert knowledge in the wind power industry is sorted and packaged as a mechanism model (i.e. a preset fault scheme library), as shown in Table 1. After identifying the abnormal state of the equipment in the warning signal, the precise positioning of the wind turbine generator equipment fault can be realized by searching the preset fault scheme library.

[0050] Table 1 Preset Fault Solution Library

[0051] The wind turbine fault diagnosis method of this invention preprocesses the collected real-time vibration waveform data to extract the frequency domain data from the vibration signal; then, it obtains vibration prediction data based on the vibration prediction model, performs residual feature identification on the frequency domain data and the vibration prediction data, and considers the equipment to be abnormal if the residual exceeds a threshold. It calculates the current characteristic amplitude under the operating condition based on statistical methods, and finds the specific fault mode and degree of the equipment according to the preset fault scheme library, so as to provide support for the stable operation of the wind turbine equipment.

[0052] like Figure 2 As shown, an embodiment of the present invention provides a wind turbine fault diagnosis device 200, comprising: The acquisition module 201 is used to acquire real-time operating condition data and real-time vibration signal data of the wind turbine. The processing module 202 is used to perform noise reduction processing on the real-time vibration signal data to obtain noise-reduced data; to perform frequency domain conversion on the noise-reduced data to obtain frequency domain data; to obtain vibration prediction data based on the real-time operating condition data and the trained vibration prediction model; the vibration prediction model is obtained by training a preset network model based on the historical operating condition data of the wind turbine and the corresponding historical vibration signal data; and to perform fault diagnosis on the wind turbine based on the frequency domain data and the vibration prediction data to obtain fault diagnosis results.

[0053] Optionally, real-time operating condition data and real-time vibration signal data of the wind turbine can be acquired, including: Acquire real-time operating condition data of the wind turbine generator set; the operating condition data includes temperature parameter data, electrical parameter data, environmental parameter data, and control parameter data; Real-time vibration signal data of the wind turbine is acquired by sensors installed at preset locations in the wind turbine.

[0054] Optionally, the real-time vibration signal data is subjected to noise reduction processing to obtain noise-reduced data, including: The real-time vibration signal data is decomposed to obtain low-frequency and high-frequency components; Thresholding is performed on the high-frequency components to obtain detail coefficients; The noise reduction data is obtained based on the low-frequency components and the detail coefficients.

[0055] Optionally, the noise-reduced data is frequency-domain transformed to obtain frequency-domain data, including: pass performing frequency domain conversion on the noise reduction data to obtain frequency domain data; wherein, is the frequency domain data, is the noise reduction data, n is a time domain index, N is a total number of sampling points, and k is a frequency domain index, , is a rotation factor.

[0056] Optionally, according to the real-time running condition data and the trained vibration prediction model, vibration prediction data is obtained, including: inputting the real-time running condition data into an input layer of the vibration prediction model to obtain a first output result; inputting the first output result into at least one processing layer of the vibration prediction model to obtain a second output result; inputting the second output result into an output layer of the vibration prediction model to obtain the vibration prediction data.

[0057] Optionally, the training process of the vibration prediction model includes: acquiring historical running condition data and corresponding historical vibration signal data of a wind turbine under a preset condition; preprocessing the historical running condition data and the historical vibration signal data to obtain preprocessed historical data; performing historical feature extraction on the preprocessed historical data to obtain historical feature data; training a preset network model according to the historical feature data to obtain an initial training result; optimizing the preset network model according to the initial training result to obtain the vibration prediction model.

[0058] Optionally, according to the frequency domain data and the vibration prediction data, fault diagnosis is performed on the wind turbine to obtain a fault diagnosis result, including: obtaining a residual error according to the frequency domain data and the vibration prediction data; determining a warning result according to the residual error and a preset warning threshold; obtaining a fault scheme according to the warning result and a preset fault scheme library; performing fault diagnosis on the wind turbine according to the frequency domain data, the vibration prediction data, the warning result, and the fault scheme to obtain the fault diagnosis result.

[0059] The wind turbine fault diagnosis device provided in the embodiment of the present application can effectively improve the fault diagnosis accuracy of the wind turbine and improve the operation efficiency of the wind turbine.

[0060] It should be noted that the device corresponds to the above method, and all implementation manners in the above method embodiments are applicable to the device embodiments and can achieve the same technical effects. Details are not described herein again.

[0061] The embodiment of the present application also provides a computing device, comprising a processor and a memory storing a computer program, wherein the computer program is executed by the processor to perform the method in any one of the above embodiments. All implementation manners in the above method embodiments are applicable to the device embodiments and can achieve the same technical effects. Details are not described herein again.

[0062] The embodiment of the present application also provides a computer readable storage medium, which stores instructions, and when the instructions are executed on a computer, the computer executes the method in any one of the above embodiments. All implementation manners in the above method embodiments are applicable to the device embodiments and can achieve the same technical effects. Details are not described herein again.

[0063] It should be noted that in the device and method of the present application, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombination should be regarded as equivalent solutions of the present application. Moreover, the steps of performing the above series of processes can be executed in time sequence according to the order of description, but it is not necessary to be executed in time sequence. Some steps can be executed in parallel, cross or independently of each other.

[0064] It is to be understood that the terminology "including", "comprising", or any other variation thereof, is intended to cover a non-exclusive inclusion such that process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Exclusion of such elements is only present if it is expressly stated that these elements are excluded. Further, it is to be understood that the description of the embodiments of the present application is not limited to the order of the steps of the methods described herein and that unless otherwise specified, the steps of the methods described herein can be performed in any order. Additionally, features described with respect to certain examples can be combined in other examples.

[0065] The above description is considered that the preferred embodiments of the present application, it should be noted that for those skilled in the art, without departing from the principles of the present application described, can also make a number of improvements and refinements, these improvements and refinements should also be considered the scope of protection of the present application.

Claims

1. A method for diagnosing faults in wind turbine generators, characterized in that, include: Acquire real-time operating condition data and real-time vibration signal data of wind turbine units; The real-time vibration signal data is subjected to noise reduction processing to obtain noise-reduced data; The noise-reduced data is transformed into frequency domain data. Vibration prediction data is obtained based on the real-time operating condition data and the trained vibration prediction model; the vibration prediction model is obtained by training a preset network model based on the historical operating condition data of the wind turbine and the corresponding historical vibration signal data. The wind turbine is diagnosed based on the frequency domain data and the vibration prediction data to obtain the fault diagnosis results.

2. The wind turbine fault diagnosis method according to claim 1, characterized in that, Acquire real-time operating condition data and real-time vibration signal data of wind turbine units, including: Acquire real-time operating condition data of the wind turbine generator set; the operating condition data includes temperature parameter data, electrical parameter data, environmental parameter data, and control parameter data; Real-time vibration signal data of the wind turbine is acquired by sensors installed at preset locations in the wind turbine.

3. The wind turbine fault diagnosis method according to claim 1, characterized in that, The real-time vibration signal data is denoised to obtain denoised data, including: The real-time vibration signal data is decomposed to obtain low-frequency and high-frequency components; Thresholding is performed on the high-frequency components to obtain detail coefficients; The noise reduction data is obtained based on the low-frequency components and the detail coefficients.

4. The wind turbine fault diagnosis method according to claim 1, characterized in that, The noise-reduced data is transformed in the frequency domain to obtain frequency domain data, including: pass The noise-reduced data is transformed into frequency domain data. in, For frequency domain data, The data is denoised, and n is the time-domain index. N is the total number of sampling points, and k is the frequency domain index. , is the rotation factor.

5. The wind turbine fault diagnosis method according to claim 1, characterized in that, Based on the real-time operating condition data and the trained vibration prediction model, vibration prediction data is obtained, including: The real-time operating condition data is input into the input layer of the vibration prediction model to obtain the first output result; The first output result is input into at least one processing layer of the vibration prediction model to obtain the second output result; The second output result is input into the output layer of the vibration prediction model to obtain vibration prediction data.

6. The wind turbine fault diagnosis method according to claim 5, characterized in that, The training process of the vibration prediction model includes: Acquire historical operating condition data and corresponding historical vibration signal data of wind turbines under preset conditions; The historical operating condition data and the historical vibration signal data are preprocessed to obtain preprocessed historical data. Historical features are extracted from the preprocessed historical data to obtain historical feature data; The preset network model is trained based on the historical feature data to obtain the initial training results; The preset network model is optimized based on the initial training results to obtain the vibration prediction model.

7. The wind turbine fault diagnosis method according to claim 1, characterized in that, Based on the frequency domain data and the vibration prediction data, fault diagnosis is performed on the wind turbine to obtain fault diagnosis results, including: The residual is obtained based on the frequency domain data and the vibration prediction data; The warning result is determined based on the residual and the preset warning threshold; Based on the warning results and the preset fault solution library, a fault solution is obtained; Based on the frequency domain data, the vibration prediction data, the early warning results, and the fault diagnosis scheme, the wind turbine is diagnosed to obtain the fault diagnosis results.

8. A wind turbine fault diagnosis device, characterized in that, include: The acquisition module is used to acquire real-time operating condition data and real-time vibration signal data of the wind turbine. The processing module is used to perform noise reduction processing on the real-time vibration signal data to obtain noise-reduced data; to perform frequency domain conversion on the noise-reduced data to obtain frequency domain data; to obtain vibration prediction data based on the real-time operating condition data and the trained vibration prediction model; the vibration prediction model is obtained by training a preset network model based on the historical operating condition data of the wind turbine and the corresponding historical vibration signal data; and to perform fault diagnosis on the wind turbine based on the frequency domain data and the vibration prediction data to obtain fault diagnosis results.

9. A computing device, characterized in that, include: A processor, a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The system stores instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images