Wind driven generator transmission chain vibration test analysis method for downsampling processing
By using downsampling and signal filtering techniques, the problem of high data storage costs in wind turbine vibration monitoring systems has been solved, resulting in reduced data volume and improved accuracy of condition assessment, thus promoting intelligent operation and maintenance in the wind power industry.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA THREE GORGES CORP FUJIAN ENERGY INVESTMENT CO LTD
- Filing Date
- 2025-11-26
- Publication Date
- 2026-05-01
AI Technical Summary
Traditional wind turbine vibration monitoring systems suffer from excessively high sampling frequencies, leading to high data storage costs and transmission pressures, and making it difficult to achieve efficient data compression without affecting the accuracy of condition assessment.
The original high-frequency vibration signal was downsampled to 2560Hz using a downsampling method. Combined with bandpass filtering and neural network analysis, the integrity of the key fault characteristic frequency band was ensured. Signal processing and feature extraction were performed using the VDI3834 standard.
It significantly reduces data storage requirements while maintaining the accuracy of vibration energy assessment, providing a cost-effective data storage solution and improving the level of intelligent operation and maintenance in the wind power industry.
Smart Images
Figure CN121952804A_ABST
Abstract
Description
A method for testing and analyzing the vibration of wind turbine drive trains using downsampling. Technical Field
[0001] This invention relates to the field of vibration testing and analysis technology for wind turbine drivetrains, and mainly to a method for vibration testing and analysis of wind turbine drivetrains using downsampling processing. Background Technology
[0002] With the large-scale development of my country's wind power industry, the number of wind turbines under the clustered operation and maintenance model of wind farms is growing exponentially, and the amount of vibration monitoring data in the transmission chain is expanding rapidly, posing a severe challenge to the data storage system. Traditional vibration monitoring systems typically use sampling frequencies of 10kHz or higher for high-speed components (such as high-speed shafts of gearboxes and generator bearings), resulting in an average daily data volume of over 5GB per unit. Large-scale wind farms face enormous storage costs and data transmission pressure.
[0003] Data storage optimization is of significant practical importance for achieving large-scale and economical operation of wind turbine condition monitoring. However, achieving efficient data compression while ensuring monitoring accuracy is a technical challenge, requiring both precise analysis of the characteristic frequency bands of vibration signals and systematic verification of downsampling algorithms. Existing research indicates that the effective fault characteristics of high-speed shaft vibration signals from domestically produced wind turbines are mainly concentrated in the 0-1280Hz range. Reasonable downsampling processing can significantly reduce the data volume without affecting the accuracy of condition assessment. Summary of the Invention
[0004] To address the storage redundancy problem caused by excessively high sampling rates in high-speed intervals of wind turbine drivetrain vibration monitoring, this invention, based on a first aspect, proposes a vibration data optimization method based on standardized downsampling. By uniformly downsampling the original high-frequency vibration signal to 2560Hz, while ensuring coverage of the main fault characteristic frequency band (0–1280Hz, satisfying the sampling theorem), the data storage volume is significantly reduced, while maintaining the accuracy of vibration energy assessment. The specific steps are as follows:
[0005] Vibration data were collected at selected measurement points;
[0006] Vibration data of high-speed components are downsampled;
[0007] According to the VDI3834 standard, the vibration data collected at each measuring point are processed by bandpass filtering.
[0008] Calculate the characteristic parameters of the filtered vibration signal;
[0009] If the calculated value exceeds the VDI3834 standard reference value, proceed to the next step of time domain analysis and frequency domain analysis.
[0010] Furthermore, the sampling frequency of the downsampling process is 2560Hz.
[0011] Furthermore, the bandpass filtering process employs a bandpass filter, where κ(t,τ) is a Gaussian bandpass kernel, and the specific formula includes:
[0012]
[0013] Where, σ K denoted by f0, which represents the filter width parameter; f0 represents the center frequency; t represents the time variable; and τ represents the time integral variable.
[0014] The filtered signal is Where T represents the total signal acquisition time.
[0015] Furthermore, the calculation of the feature index specifically includes: taking the indicators of different factors as vectors to form a vector matrix, using mean processing to obtain a parameter feature matrix, and using the parameter feature matrix as the input layer of the test analysis model constructed based on the neural network for training.
[0016] Furthermore, the time-domain analysis observes whether the vibration waveform exhibits periodic impact phenomena.
[0017] Furthermore, the frequency domain analysis includes: whether sidebands appear in the gear's spectrum and whether characteristic frequencies appear in its envelope spectrum.
[0018] According to a second aspect of the present invention, a computer program product is provided, on which one or more computer programs are stored, which, when executed by a computer processor, implement the method described above.
[0019] The above-described one or more technical solutions in the embodiments of this application have at least one of the following technical effects:
[0020] This invention, based on measured vibration data from 70 wind turbine units in the coastal mountainous areas of Fujian, systematically studied the impact of different downsampling rates on fault feature extraction using a combination of signal processing and engineering practice. Ultimately, 2560Hz was determined to be the optimal sampling frequency. This frequency setting not only fully preserves key fault features of the transmission chain (such as gear meshing frequency and bearing fault characteristic frequency), but also significantly reduces the amount of data stored.
[0021] This invention not only provides wind farm operators with an economical and efficient data storage solution, but also provides equipment manufacturers with a technical basis for optimizing the design of condition monitoring systems, which is of positive significance for promoting the improvement of intelligent operation and maintenance level in my country's wind power industry. Attached Figure Description
[0022] The accompanying drawings are included to provide a further understanding of the embodiments and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments and, together with the description, serve to explain the principles of the invention. Other embodiments and many anticipated advantages of the embodiments will be readily recognized as they become better understood through reference to the following detailed description. Elements in the drawings are not necessarily to scale. The same reference numerals refer to corresponding similar parts.
[0023] Figure 1 shows a schematic flowchart of a wind turbine drivetrain vibration test and analysis method with downsampling processing according to an embodiment of the present invention.
[0024] Figure 2 is a schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application. Detailed Implementation
[0025] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.
[0026] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.
[0027] Figure 1 shows a flowchart of a wind turbine drivetrain vibration test and analysis method with downsampling processing according to an embodiment of the present invention, as shown in Figure 1:
[0028] S1. Collect vibration data at the selected measuring points;
[0029] S2. Downsampling is performed on the vibration data of high-speed components;
[0030] First, select a downsampling factor, which refers to the percentage reduction in the sampling rate. For example, if a downsampling factor of 2 is selected, then one out of every two sampling points will be retained, reducing the sample data by half.
[0031] Before downsampling, to avoid aliasing caused by the reduced sampling rate, the data is usually low-pass filtered. The filter's function is to remove high-frequency components while retaining low-frequency components, ensuring that the downsampled data does not lose important information.
[0032] Downsampling is achieved by reducing the number of sampling points. In some embodiments, the original signal has a sampling rate of 1000Hz, and with a downsampling factor of 2, the downsampled sampling rate is 500Hz.
[0033] After downsampling, check whether the signal retains sufficient characteristics, especially for vibration data, to accurately reflect the dynamic characteristics of the system. Spectral analysis or time-domain analysis can be used to check whether the downsampled data is distorted or has lost important information.
[0034] In the final step, the correctness of the downsampling is verified by comparing the signal characteristics before and after downsampling. For example, time-domain or frequency-domain metrics (such as peak value, spectral amplitude, etc.) are used to compare changes in the signal to ensure that no critical information is lost during the downsampling process. Feedback optimization is used to improve downsampling efficiency.
[0035] S3. According to the VDI3834 standard, bandpass filtering is performed on the vibration data collected at each measuring point.
[0036] Bandpass filtering is used, with κ(t,τ) as the Gaussian bandpass kernel. Specific formulas include:
[0037]
[0038] Where, σ K denoted by f0, which represents the filter width parameter; f0 represents the center frequency; t represents the time variable; and τ represents the time integral variable.
[0039] The filtered signal is Where T represents the total signal acquisition time.
[0040] A bandpass filter allows a specific frequency range (centered at the center frequency f0, determined by the filter width parameter σ) to be used. K Vibration signals within a defined range are allowed to pass through, while signals outside that range are suppressed. This allows the vibration signal of the frequency band of interest to be separated from the original signal containing various frequency components, eliminating irrelevant frequency interference and facilitating the analysis of vibration characteristics in a specific frequency band.
[0041] Among them, the Gaussian bandpass kernel Part of it exhibits Gaussian function characteristics, which smooth the signal. As (t-τ) increases, the function value decays rapidly, making the influence of signal components far from the current time t on the filtering result smaller. This effectively suppresses high-frequency noise and other interference signals, improving the signal-to-noise ratio.
[0042] The filtering method of this invention combines time and frequency information, and performs integration operations in the time domain. Signal processing can preserve the local temporal characteristics of the signal to a certain extent, while focusing on the frequency band near the center frequency f0 in the frequency domain. This allows the filtered signal to retain the vibration characteristics of specific frequency components and reflect the temporal changes of the signal, which is beneficial for subsequent analysis work such as fault diagnosis and condition monitoring based on vibration signals.
[0043] S4. Calculate the characteristic indicators of the filtered vibration signal;
[0044] Different factor indicators are used as vectors to form a vector matrix. The mean is used to obtain the parameter feature matrix. The test analysis model based on the neural network uses the parameter feature matrix as the input layer and inputs it into the neural network model for training.
[0045] After filtering, the vibration signal is typically low-pass or band-pass filtered. This ensures the signal does not contain excessively high-frequency noise or distortion; therefore, the filtered signal will be used for feature extraction.
[0046] Feature metrics typically include time-domain features, frequency-domain features, and time-frequency-domain features:
[0047] Time-domain characteristics: such as mean, standard deviation, peak value, kurtosis, skewness, maximum value, minimum value, RMS value, etc.
[0048] Frequency domain characteristics: These include the results of the signal's spectral analysis. Common characteristics include power spectral density, center frequency, bandwidth, and peaks in the frequency distribution.
[0049] Time-frequency domain characteristics: Features obtained through time-frequency analysis methods (such as wavelet transform, short-time Fourier transform, etc.), such as instantaneous frequency and energy distribution.
[0050] Multiple feature indicators extracted from the signal (such as time-domain and frequency-domain features) are used as different dimensions to construct a feature vector. For each set of vibration data (such as the signal at each moment), a feature vector is generated.
[0051] For example: In some embodiments, the feature matrix X is an N*10 matrix:
[0052]
[0053] The feature matrix will be averaged, that is, the mean of each column (each feature) will be standardized, and the processed data will be input into the neural network model.
[0054] The mean is standardized as follows:
[0055]
[0056] Where, μ i and σ i Let x represent the i-th feature respectively. i The mean and standard deviation.
[0057] The neural network is trained using supervised learning to learn the mapping relationship between features and the target. It includes an input layer, hidden layers, and an output layer. The input layer uses the mean-processed feature matrix as its input. The number of nodes in the input layer equals the dimension of the features (e.g., 10 features). Depending on the complexity of the problem, one or more hidden layers can be designed; the number of nodes in each hidden layer can be determined experimentally. Generally, the number of hidden layers and the number of nodes in each layer significantly impact the model's expressive power. The number of nodes in the output layer can vary depending on the task; for example, in classification problems, the number of output layer nodes equals the number of categories, while in regression problems, the number of output layer nodes is 1.
[0058] The neural network training process includes: dividing the dataset into training, validation, and test sets; selecting an appropriate loss function; using mean squared error (MSE) to solve regression problems and cross-entropy to solve classification problems; and using gradient descent or the Adam optimizer to minimize the loss function. The weights and biases of the neural network are then adjusted using the backpropagation algorithm until the network's predictive ability reaches the expected accuracy.
[0059] S5. If the calculated values exceed the VDI3834 standard reference values, proceed to the next step of time-domain and frequency-domain analysis. Time-domain analysis should be used to observe whether periodic impact phenomena appear in the vibration waveform; frequency-domain analysis should be used to observe whether sidebands appear in the gear's spectrum or whether characteristic frequencies appear in the envelope spectrum.
[0060] Referring now to FIG2, a schematic diagram of a computer system 200 suitable for implementing an electronic device according to an embodiment of the present application is shown. The electronic device shown in FIG2 is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present application.
[0061] As shown in Figure 2, the computer system 200 includes a central processing unit (CPU) 201, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 202 or programs loaded from storage section 208 into random access memory (RAM) 203. The RAM 203 also stores various programs and data required for the operation of the system 200. The CPU 201, ROM 202, and RAM 203 are interconnected via a bus 204. An input / output (I / O) interface 205 is also connected to the bus 204.
[0062] The following components are connected to I / O interface 205: an input section 206 including a keyboard, mouse, etc.; an output section 207 including a liquid crystal display (LCD) and speakers, etc.; a storage section 208 including a hard disk, etc.; and a communication section 209 including a network interface card such as a LAN card and a modem, etc. The communication section 209 performs communication processing via a network such as the Internet. A drive 210 is also connected to I / O interface 205 as needed. A removable medium 211, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 210 as needed so that computer programs read from it can be installed into storage section 208 as needed.
[0063] Specifically, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 209, and / or installed from removable medium 211. When the computer program is executed by central processing unit (CPU) 201, it performs the functions defined in the methods of this application. It should be noted that the computer-readable storage medium of this application can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable storage medium other than a computer-readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. Program code contained on a computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.
[0064] Computer program code for performing the operations of this application can be written in one or more programming languages or a combination thereof. Programming languages include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0065] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0066] The modules described in the embodiments of this application can be implemented in software or in hardware.
[0067] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: collect vibration data at selected measurement points; perform downsampling processing on the vibration data of high-speed components; perform bandpass filtering processing on the vibration data collected at each measurement point according to the VDI3834 standard; calculate characteristic indicators on the filtered vibration signal; and if the calculated values exceed the reference values of the VDI3834 standard, perform further time-domain and frequency-domain analysis.
[0068] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.
Claims
1. A method for testing and analyzing the vibration of a wind turbine drivetrain using downsampling processing, characterized in that, include: Vibration data were collected at selected measurement points; vibration data of high-speed components were downsampled. According to the VDI3834 standard, the vibration data collected at each measuring point are processed by bandpass filtering. Calculate the characteristic parameters of the filtered vibration signal; If the calculated value exceeds the VDI3834 standard reference value, proceed to the next step of time domain analysis and frequency domain analysis.
2. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 1, characterized in that, The downsampling process uses a sampling frequency of 2560Hz.
3. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 1, characterized in that, The bandpass filtering process employs a bandpass filter, where κ(t,τ) is a Gaussian bandpass kernel. The specific formula includes: Where, σ K denoted by f0, which represents the filter width parameter; f0 represents the center frequency; t represents the time variable; and τ represents the time integral variable.
4. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 3, characterized in that, The bandpass filtering process produces a filtered signal that is... Where T represents the total signal acquisition time.
5. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 1, characterized in that, The calculation of the feature index specifically includes: taking the indexes of different factors as vectors to form a vector matrix, using mean processing to obtain a parameter feature matrix, and using the parameter feature matrix as the input layer of a test analysis model constructed based on a neural network for training.
6. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 1, characterized in that, The time-domain analysis observes whether the vibration waveform exhibits periodic impact phenomena.
7. The method for testing and analyzing the vibration of the wind turbine drive train according to claim 1, characterized in that, The frequency domain analysis includes whether sidebands appear in the gear's spectrum and whether characteristic frequencies appear in its envelope spectrum.
8. A computer program product, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1-7.
9. A computing system, characterized in that, It includes a processor and a memory, the processor being configured to perform the method as described in any one of claims 1-7.