Wind power transmission chain inter-stage decoupling anti-aliasing fmd fault feature extraction method and system

Abstract

This invention discloses a method and system for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind turbine drivetrains, relating to the field of wind turbine drivetrain fault diagnosis technology. The method includes: acquiring raw acoustic signals of the drivetrain, real-time turbine speed, and casing vibration acceleration signals, and preprocessing them to obtain a clean input signal; calculating the meshing frequency and rotational frequency of each stage based on the number of gear teeth and real-time speed of the drivetrain; performing inter-stage decoupling FMD decomposition on the clean input signal based on the meshing frequency and rotational frequency of each stage to obtain preliminary decoupling modal components, and performing anti-aliasing mode screening to remove aliasing modal components, obtaining non-aliasing decoupling modal components; and purifying the fault features of the non-aliasing decoupling modal components to construct a decoupling fault feature spectrum. This invention achieves complete decoupling of acoustic signals between drivetrain stages and thorough suppression of modal aliasing, making the fault features of each stage appear independently and clearly, effectively improving the early fault detection rate and diagnostic accuracy.

Description

Technical Field

[0001] This invention belongs to the field of wind turbine drivetrain fault diagnosis technology, specifically relating to a method and system for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind turbine drivetrains. Background Technology

[0002] Wind turbine drivetrains typically employ a three-stage composite transmission structure consisting of a two-stage planetary gear train and a single-stage parallel gear train. This structure, as the core load-bearing and power transmission link connecting the main shaft and generator, endures alternating loads, gust impacts, electromagnetic braking torque fluctuations, and frequent speed and load variations during continuous wind turbine operation. Mechanical failures such as broken gear teeth, cracked secondary sun gears, bearing wear, and loose couplings are frequent. If these faults are not detected and located in time, they can trigger a chain reaction of drivetrain failures or even damage to the entire turbine. Acoustic detection technology, with its inherent advantages such as non-contact installation, no alteration of the stress state of the measured component, immunity to multi-interface attenuation interference in the vibration transmission path, and significantly higher sensitivity to early, weak faults compared to traditional vibration acceleration signals, has gradually become one of the mainstream technologies for uninterrupted online monitoring of wind turbine drivetrains. Meanwhile, Eigenmode Decomposition (FMD), as a rapidly developing blind source separation and non-recursive signal decomposition method in recent years, can accurately capture the periodic and impulsive features in the signal without relying on preset basis functions. At the same time, it effectively suppresses the frequency aliasing and mode aliasing problems that are common in methods such as Empirical Mode Decomposition (EMD). Therefore, it is regarded as the preferred algorithm for extracting acoustic fault features of transmission chains.

[0003] However, in the engineering application of the above-mentioned methodological system, the inherent strong coupling characteristics of inter-stage vibrations in the multi-stage transmission structure of wind power transmission chains constitute a fundamental problem restricting the accuracy of fault diagnosis. Specifically, the first-stage planetary gear train, the second-stage planetary gear train, and the third-stage parallel gear train each generate meshing vibrations and acoustic radiation signals of different frequency bands during operation. These signals are superimposed and modulated through structural paths such as gearbox housing, connecting flanges, and transmission shafts, forming a complex coupled sound field with interlaced inter-stage meshing frequencies and series modulation of rotational frequencies. This makes the fault characteristics corresponding to each transmission stage intertwined in the frequency domain and difficult to distinguish. Under this strong coupling background, the acoustic detection signal is essentially a strongly non-stationary composite signal composed of multi-stage meshing frequencies, inter-stage series modulation components of rotational frequencies, and broadband environmental noise. The complexity of its spectral structure and the severity of mode mixing far exceed the fault diagnosis scenarios of conventional rotating machinery such as fixed-axis gearboxes.

[0004] Faced with the combined challenges of interstage coupling and mode aliasing, existing mainstream decomposition algorithms have revealed varying degrees of shortcomings in engineering practice. EMD and its improved algorithms employ recursive extreme point envelope and sieving strategies. Without strict mathematical orthogonality support, energy leakage easily occurs between intrinsic mode functions at each stage, causing coupling components from different transmission stages to become mixed within the same mode component and unable to be separated. Even with the introduction of noise-assisted strategies, the inherent problem of interstage coupling components being split into multiple spurious modes cannot be fundamentally solved. Although Variational Mode Decomposition (VMD) improves the orthogonality of modes through a variational framework, its decomposition effect heavily relies on the pre-set number of modes. However, the number of interstage coupling frequency bands in wind turbine transmission chains changes in real time with operating conditions. A fixed number of modes cannot adaptively match the dynamic evolution of multi-stage transmission coupling characteristics. Over-decomposition and under-decomposition frequently alternate in actual wind field data, causing the decomposition results to lose physical consistency. Although the standard FMD improves the robustness of the decomposition to some extent through non-recursive filtering and period estimation strategies, its filter design is not specifically optimized for the spectral structure of multi-stage series coupling in the transmission chain. The filter channels cannot effectively distinguish the dense sideband components generated by frequency conversion series modulation. After decomposition, there are still non-negligible residual mode aliasing. The fault features are significantly distorted in the envelope demodulation spectrum. Under low signal-to-noise ratio conditions, the fault features are completely submerged in the residual coupling components and noise background.

[0005] In summary, existing decomposition methods generally lack in-depth modeling and dedicated decoupling mechanisms for the coupling mechanism between stages of the wind turbine drivetrain. They cannot reveal the frequency composition law of multi-stage series synthesis modulation at the physical level, nor can they achieve effective separation of coupled components and complete suppression of aliasing components at the algorithm level. As a result, the accuracy of fault identification is difficult to meet the actual needs of engineering, and false and false diagnoses occur frequently, which seriously restricts the large-scale engineering deployment of acoustic detection technology in the field of wind turbine drivetrain fault early warning. Summary of the Invention

[0006] The purpose of this invention is to overcome the above-mentioned problems and provide a method and system for extracting fault features of wind power transmission chain decoupling and anti-aliasing FMD, so as to solve the technical problem that fault features cannot be effectively extracted due to strong acoustic signal coupling and modal aliasing between transmission chain stages in the prior art.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains, comprising the following steps: Acquire raw acoustic signals of the drive train, real-time speed of the unit, and vibration acceleration signals of the housing. Preprocess the acquired signals to obtain clean input signals. Calculate the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed; Based on the meshing frequencies and rotation frequencies of each stage, the pure input signal is subjected to interstage decoupling FMD decomposition to obtain preliminary decoupled mode components; The preliminary decoupled mode components are subjected to anti-aliasing mode screening to remove aliasing mode components, resulting in non-aliasing decoupled mode components. Fault features are purified from the non-aliased decoupled mode components to construct a decoupled fault feature spectrum.

[0008] A further improvement of the present invention is that the preprocessing of the acquired signal to obtain a clean input signal includes: Synchronously acquire raw acoustic signals of the drive train, real-time speed of the unit and vibration acceleration signals of the housing, set the sampling frequency, and use a unified timestamp to complete the timing alignment of multi-source signals; Wavelet soft thresholding and Grubbs criterion are used to remove outliers and obtain a clean input signal; The expression for the pure input signal is:

[0009] In the formula, For a clean input signal; This is a wavelet soft thresholding denoising function; The original acoustic signal of the transmission chain; The average temporal amplitude of the acoustic signal; This represents the standard deviation of the timing amplitude of the acoustic signal.

[0010] A further improvement of this invention is that the calculation of the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed includes: Calculate the first-stage meshing frequency, second-stage meshing frequency, third-stage meshing frequency, low-speed shaft frequency, first-stage sun gear frequency, and second-stage planetary gear frequency based on the number of teeth and real-time rotational speed of the transmission chain gears; Constructing an inter-stage cascaded synthesis coupling model:

[0011] In the formula, This is the interstage coupling modulation frequency; For the first Stage meshing frequency, ; The modulation order; This is the rotational frequency of a first-order solar rotor; The rotation frequency of a second-stage planetary gear; It is a low-speed shaft rotation frequency.

[0012] A further improvement of this invention is that, based on the meshing frequencies and transition frequencies of each stage, the pure input signal is subjected to inter-stage decoupling FMD decomposition to obtain preliminary decoupled modal components, including: Load the clean input signal, and set the FIR filter length, initial decomposition mode number, and maximum number of iterations; Using an FIR filter bank:

[0013] In the formula, For the first The second iteration FIR filter coefficients; For filter coefficient index; This is the length of the FIR filter; The initial decomposition mode number; Using the meshing frequencies and frequency transitions at each stage as the filtering centers, the clean input signal is convolved and decomposed with the FIR filter bank:

[0014] In the formula, For the first The second iteration One decoupled modal component; For a clean input signal; This is a convolution operation; Calculate the autocorrelation function of each decoupled mode component, take the local maximum point after crossing zero as the fault period estimate, and update the FIR filter coefficients based on the fault period estimate:

[0015] In the formula, For the first The second iteration Fault cycle estimates for each mode; For the first Autocorrelation function of each mode; This is the filter coefficient update function; These are the updated FIR filter coefficients; After iterating to the maximum number of iterations, the initial decoupling decomposition is completed, and the initial decoupling modal components are obtained.

[0016] A further improvement of the present invention is that the step of performing anti-aliasing mode screening on the preliminary decoupled mode components, eliminating aliasing mode components, and obtaining non-aliasing decoupled mode components includes: Calculate the correlation coefficient between any two initially decoupled modal components and construct the correlation coefficient matrix:

[0017] In the formula, For the first The modal component and the first The correlation coefficient between the modal components; It is the covariance function; It is the variance function; For the first The, the One initial decoupled modal component; For the modal component pairs with the highest correlation coefficients, calculate the correlation kurtosis of each modal component:

[0018] In the formula, For the first The correlation kurtosis of each modal component; This represents the number of signal sampling points. The cumulative order of the period; The sampling period; For the first The modal component in the th ... The amplitude of each sampling point; The modal components with higher correlation kurtosis are retained, while the aliased modal components with correlation coefficients higher than a preset threshold and lower correlation kurtosis are removed. This process is repeated until the number of modes reaches the target value, resulting in aliased decoupled modal components.

[0019] A further improvement of this invention is that the step of refining the fault features of the non-aliasing decoupled mode components to construct a decoupled fault feature spectrum includes: Perform Hilbert transform on each non-aliased decoupled mode component to extract the fault envelope amplitude of each mode component:

[0020] In the formula, For Hilbert transform operators; For the first Analytical signals of each modal component; For the first One non-aliased decoupled mode component; For integration variables; For the first The fault envelope amplitude of each modal component; Based on the fault envelope amplitude of each modal component, the independent fault features and modulation features at each level are integrated to construct a decoupled fault feature spectrum:

[0021] In the formula, To decouple the fault characteristic spectrum; For the first Stage meshing frequency; This is the rotational frequency of a first-order solar rotor; The rotation frequency of a second-stage planetary gear; For the first Fault envelope amplitude spectrum of each modal component.

[0022] A further improvement of this invention is that it also includes an adaptive iterative calibration step based on operating conditions: When the unit speed changes, the meshing frequency and rotation frequency of each stage are recalculated based on the real-time speed after the change, and the filter center frequency of the inter-stage decoupling FMD decomposition is updated. When the wind turbine is in a low-load idling condition, the health status reference signal under the current condition is collected, and the correlation coefficient threshold of the anti-aliasing mode screening is reset according to the reference signal. During long-term operation, acoustic signals of the transmission chain are periodically collected and fault features are extracted. The drift trend of the acoustic features with service time is fitted, and the long-period drift correction of the decoupled fault feature spectrum is performed based on the drift trend.

[0023] Secondly, this invention provides a wind power transmission chain inter-stage decoupling and anti-aliasing FMD fault feature extraction system, comprising the following modules: The signal acquisition and preprocessing module is used to acquire the original acoustic signals of the drive chain, the real-time speed of the unit and the vibration acceleration signals of the housing, and to preprocess the acquired signals to obtain a clean input signal. The interstage frequency calibration module is used to calculate the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed. The interstage decoupling FMD decomposition module is used to perform interstage decoupling FMD decomposition on the pure input signal based on the meshing frequency and the rotation frequency of each stage to obtain preliminary decoupling modal components; An anti-aliasing mode filtering module is used to perform anti-aliasing mode filtering on the preliminary decoupled mode components, remove aliasing mode components, and obtain non-aliasing decoupled mode components; The fault feature purification and decoupling spectrum construction module is used to purify the fault features of the non-aliased decoupled mode components and construct the decoupled fault feature spectrum.

[0024] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-mentioned wind power transmission chain inter-stage decoupling anti-aliasing FMD fault feature extraction method.

[0025] Fourthly, the present invention provides a storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method for extracting fault features of wind power transmission chain inter-stage decoupling and anti-aliasing FMD.

[0026] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a method for extracting FMD (Fault Characteristics) fault features for decoupling and anti-aliasing in wind turbine drivetrain stages. By simultaneously acquiring and preprocessing acoustic, rotational speed, and vibration signals, the method first ensures the consistency of input signal quality from the source, providing a reliable signal foundation for subsequent decomposition. Then, based on the number of gear teeth and real-time rotational speed, the method calculates the meshing frequency and rotational frequency of each stage, giving the subsequent decomposition steps a clear physical direction rather than blind searching, ensuring a physical correspondence between the decomposition results and the transmission characteristics of each stage of the drivetrain. Based on this, interstage decoupling is performed using the calculated meshing frequency and rotational frequency of each stage as the filtering center. FMD decomposition separates the originally intertwined coupled components at each stage into independent preliminary decoupled modal components, achieving decoupling and separation of signals between stages from the source. Subsequently, anti-aliasing modal filtering removes aliased modal components, further eliminating spurious components and cross-interference remaining from the decomposition process. This ensures that the remaining non-aliased decoupled modal components truly reflect the independent damage information of each transmission stage. Finally, the purified modal components undergo fault feature extraction, and a decoupled fault feature spectrum is constructed, allowing the fault features of each stage of the transmission to be presented independently and clearly in the spectrum, directly usable for fault diagnosis without secondary processing. These steps are sequentially linked and progressively purified, ultimately separating and revealing the fault features at each stage that were previously submerged in the background of strong coupling and modal aliasing. This effectively solves the technical problem of ineffective fault feature extraction caused by strong acoustic signal coupling and modal aliasing between transmission stages. Attached Figure Description

[0027] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of the invention in any way. Furthermore, the shapes and proportions of the components in the drawings are merely illustrative to aid in understanding the invention and do not specifically limit the shapes and proportions of the components of the invention.

[0028] Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the system of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0030] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0031] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0032] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper," "lower," "horizontal," or "inner" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0033] Furthermore, the use of the term "horizontal" does not imply that the component must be absolutely horizontal, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0034] In the description of the embodiments of the present invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention according to the specific circumstances.

[0035] The present invention will now be described in further detail with reference to the accompanying drawings: The first objective of this invention is to provide a method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains, such as... Figure 1 As shown, it includes the following steps: S1, Acoustic signal acquisition and refined preprocessing of the transmission chain.

[0036] Synchronous acquisition of raw acoustic signals from the drive train Real-time speed of the unit Box vibration acceleration The sampling frequency was set to 32kHz, and time stamps were unified to achieve timing alignment of multi-source signals. Wavelet soft thresholding denoising and Grubbs' criterion were used to remove outliers, resulting in a clean input signal.

[0037]

[0038] In the formula, The average temporal amplitude of the acoustic signal; The standard deviation of the timing amplitude of the acoustic signal. It is a wavelet soft thresholding denoising function that can effectively remove interference components such as wind noise and electromagnetic clutter.

[0039] The Grubbs criterion is used to eliminate occasional outlier values ​​that may occur during field data acquisition: when the amplitude of a sampling point deviates from the signal mean by more than 3 times the standard deviation, the point is determined to be an outlier and is removed, thereby ensuring the statistical consistency of the input signal.

[0040] S2, Frequency calibration and coupling model construction between transmission chain stages.

[0041] The core frequencies of each stage are calculated based on the number of teeth in the transmission chain gears and the real-time rotational speed. Specifically, the meshing frequency of the first stage is calculated. Secondary meshing frequency Third-stage meshing frequency Low-speed shaft frequency First-level solar wheel rotation frequency and the rotation frequency of the second-level planetary cycle Constructing an inter-stage cascaded synthesis coupling model:

[0042] In the formula, This is the interstage coupling modulation frequency; For the first Stage meshing frequency, ; is the modulation order (a non-negative integer); This is the rotational frequency of a first-order solar rotor; The rotation frequency of a second-stage planetary gear; The low-speed shaft rotation frequency is used. This model can fully characterize the frequency distribution law of multi-stage coupling in the transmission chain. That is, under the joint modulation of its own rotation frequency and the rotation frequency of adjacent transmission stages, the meshing frequency of each stage forms a sideband cluster in the spectrum with the meshing frequency as the center and the rotation frequency of each stage as the interval. The superposition of these sideband clusters constitutes the complete frequency structure of inter-stage series synthetic coupling.

[0043] S3, inter-stage decoupling and anti-aliasing FMD decomposition.

[0044] Loading preprocessed signals Set the FIR filter length Initial decomposition mode number Maximum number of iterations Using an FIR filter bank:

[0045] In the formula, For the first The second iteration FIR filter coefficients; filter length It can balance filtering accuracy and signal smoothness. An embedded inter-stage decoupling unit performs convolution decomposition with the meshing frequencies and turning frequencies of each stage calculated in S2 as the filtering center.

[0046] In the formula, For the first The second iteration One decoupled modal component; For convolution operations, precise frequency domain matching filtering of signals and filters can be achieved. That is, the meshing frequency and the turn frequency of each stage are used as the passband center frequency of each filter channel to ensure that each channel extracts the signal component of the corresponding transmission stage.

[0047] Calculate modal autocorrelation The local maximum point after crossing zero is taken as the fault cycle estimate. And complete the FIR filter coefficient update:

[0048] In the formula, For the first Autocorrelation function of each mode; This is the filter coefficient update function. The initial decoupling is completed after iterating to the maximum number of iterations.

[0049] This step binds the center frequency of the filter in the standard FMD algorithm to the meshing frequency and rotation frequency of each stage of the transmission chain, so that each filter channel is always optimized around the frequency band where the transmission characteristics of each stage are located during the iteration process, thereby achieving effective separation of interstage coupling components at the decomposition source.

[0050] S4, Anti-aliasing mode screening.

[0051] Calculate the correlation coefficient between any two preliminary decoupled modal components. , build 3D correlation coefficient matrix:

[0052] In the formula, It is the covariance function; Let be the variance function, and the range of the correlation coefficient is . A higher value indicates a greater degree of aliasing between the two modal components. The correlation kurtosis is calculated for the mode pair with the highest correlation. :

[0053] In the formula, This represents the number of signal sampling points. This is the periodic cumulative order (usually 3 to 7). The sampling period is defined as the correlation kurtosis value. A larger correlation kurtosis value indicates that the fault impact characteristics contained in the modal component are more significant. Modal components with larger correlation kurtosis are retained, while weak feature modes with severe aliasing are removed. That is, modal components with correlation coefficients higher than a preset threshold (e.g., 0.85) and relatively smaller correlation kurtosis are identified as aliased modal components and removed. This process is iterated until the number of modes reaches the target value, resulting in non-aliased decoupled modal components.

[0054] This step identifies the degree of aliasing between modes through correlation coefficients, evaluates the fault information content of each mode through correlation kurtosis, and iteratively filters modes based on the criterion of selecting one of the two with high correlation and retaining high kurtosis. This ensures the independence of each modal component while maximizing the retention of fault impact information.

[0055] S5, Fault Feature Purification and Decoupling Spectrum Construction.

[0056] Hilbert transform is applied to the non-aliased decoupled modal components to extract the fault envelope amplitude and enhance the periodic impact characteristics.

[0057] In the formula, For Hilbert transform operators; For the first Analytical signals of each modal component; For the first One non-aliased decoupled mode component; For integration variables; For the first The fault envelope amplitude of each modal component can effectively amplify the impact characteristics of faults such as broken gear teeth and cracked sun gears, because these faults generate periodic transient impacts. Hilbert envelope demodulation can convert these impact components into amplitude-modulated signals, thus forming a highly recognizable spectral line in the envelope spectrum with fault characteristic frequencies and their harmonics as peaks. By integrating the independent fault characteristics and modulation characteristics of each stage, a clustered-amplitude two-dimensional form of the transmission chain decoupled fault characteristic spectrum is constructed.

[0058] In the formula, To decouple the fault characteristic spectrum; For the first Stage meshing frequency; This is the rotational frequency of a first-order solar rotor; The rotation frequency of a second-stage planetary gear; For the first Fault envelope amplitude spectrum of each modal component.

[0059] The decoupling characteristic spectrum can achieve complete separation of interstage coupling components; the characteristics of the first-stage transmission only appear in the first stage. Within the frequency band centered on the second-order transmission characteristic, it only appears in the frequency band centered on the second-order transmission characteristic. Within the frequency band centered on the three-stage transmission characteristics, the three-stage transmission characteristics only appear in the frequency band centered on the three-stage transmission characteristics. Within the central frequency band, there is no cross-interference between the characteristics at each level. Under normal operating conditions, the decoupling characteristic spectrum only shows the fundamental frequency of the high-speed shaft rotation and its low-order harmonic modulation; under fault conditions, the unique single-sided modulation characteristic of the rotational coupling between the first-stage sun gear and the second-stage planetary gear (i.e., the sideband only appears on one side of the meshing frequency) can be clearly shown. This characteristic is a typical sign of cracks in the second-stage sun gear and can directly identify this type of fault. After processing by this method, mode aliasing is completely eliminated, the fault impact amplitude is increased by 3 to 5 times, weak early fault characteristics can be directly identified, noise components are completely eliminated, the frequency band signal-to-noise ratio is increased to over 20 dB, and the fault characteristic identification rate reaches 100%.

[0060] S6, adaptive iterative calibration under operating conditions.

[0061] When the unit speed changes, the interstage frequency and filter center frequency are recalibrated to complete the real-time update of decoupling parameters. Specifically, when the speed changes from Change to At that time, the meshing frequency and rotation frequency of each stage change synchronously and proportionally, the coupling model in S2 is updated accordingly, and the filter center frequency in S3 is adjusted synchronously. When the wind turbine is in a low-load idling condition, a health status reference signal is collected and the decoupling threshold is reset (i.e., the correlation coefficient screening threshold is recalculated) to achieve reference calibration. Because the transmission chain load is small and the background noise is low under low load conditions, the signal collected at this time can be used as the "fingerprint" reference of the unit to correct system deviations caused by sensor aging and environmental changes in subsequent detection. During long-term operation, the acoustic feature drift trend is fitted (such as the slow decrease of meshing frequency amplitude, the gradual increase of sideband amplitude, and other long-term evolution trends), the accumulated error within the service cycle is eliminated, and long-term drift correction is completed to ensure the stability of feature extraction under variable operating conditions and long-term operation.

[0062] The second objective of this invention is to provide a wind power transmission chain inter-stage decoupling and anti-aliasing FMD fault feature extraction system, such as... Figure 2 As shown, it includes the following modules: Signal acquisition and preprocessing module: used to acquire raw acoustic signals of the drive train, real-time speed of the unit and vibration acceleration signals of the housing, and preprocess the acquired signals to obtain clean input signals; Interstage frequency calibration module: used to calculate the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed; Interstage decoupling FMD decomposition module: used to perform interstage decoupling FMD decomposition on the pure input signal based on the meshing frequency and the rotation frequency of each stage to obtain preliminary decoupling modal components; Anti-aliasing mode screening module: used to perform anti-aliasing mode screening on the preliminary decoupled mode components, remove aliasing mode components, and obtain non-aliasing decoupled mode components; Fault Feature Purification and Decoupling Spectrum Construction Module: Used to purify fault features of the non-aliased decoupled modal components and construct a decoupled fault feature spectrum.

[0063] The system is implemented entirely through algorithms without requiring hardware modifications, is compatible with existing acoustic acquisition systems, and is suitable for mass deployment in various onshore and offshore wind farms.

[0064] In one embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or 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. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions from the computer storage medium to achieve a corresponding method flow or function. The processor described in this embodiment of the present invention can be used in the operation of a wind turbine blade lightning damage detection method based on infrared-electromagnetic coupling.

[0065] This invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the wind turbine blade lightning damage detection method based on infrared-electromagnetic coupling in the above embodiments.

[0066] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0067] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0068] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0069] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains, characterized in that, Includes the following steps: Acquire raw acoustic signals of the drive train, real-time speed of the unit, and vibration acceleration signals of the housing. Preprocess the acquired signals to obtain clean input signals. Calculate the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed; Based on the meshing frequencies and rotation frequencies of each stage, the pure input signal is subjected to interstage decoupling FMD decomposition to obtain preliminary decoupled mode components; The preliminary decoupled mode components are subjected to anti-aliasing mode screening to remove aliasing mode components, resulting in non-aliasing decoupled mode components. Fault features are purified from the non-aliased decoupled mode components to construct a decoupled fault feature spectrum.

2. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 1, characterized in that, The preprocessing of the acquired signal to obtain a clean input signal includes: Synchronously acquire raw acoustic signals of the drive train, real-time speed of the unit and vibration acceleration signals of the housing, set the sampling frequency, and use a unified timestamp to complete the timing alignment of multi-source signals; Wavelet soft thresholding and Grubbs criterion are used to remove outliers and obtain a clean input signal; The expression for the pure input signal is: In the formula, For a clean input signal; This is a wavelet soft thresholding denoising function; The original acoustic signal of the transmission chain; The average temporal amplitude of the acoustic signal; This represents the standard deviation of the timing amplitude of the acoustic signal.

3. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 1, characterized in that, The calculation of the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed includes: Calculate the first-stage meshing frequency, second-stage meshing frequency, third-stage meshing frequency, low-speed shaft frequency, first-stage sun gear frequency, and second-stage planetary gear frequency based on the number of teeth and real-time rotational speed of the transmission chain gears; Constructing an inter-stage cascaded synthesis coupling model: In the formula, This is the interstage coupling modulation frequency; For the first Stage meshing frequency, ; The modulation order; This is the rotational frequency of a first-order solar detonator; The rotation frequency of a second-stage planetary gear; It is a low-speed shaft rotation frequency.

4. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 3, characterized in that, The step of performing interstage decoupling FMD decomposition on the pure input signal based on the meshing frequencies and transition frequencies of each stage to obtain preliminary decoupled modal components includes: Load the clean input signal, and set the FIR filter length, initial decomposition mode number, and maximum number of iterations; Using an FIR filter bank: In the formula, For the first The second iteration FIR filter coefficients; For filter coefficient index; This is the length of the FIR filter; The initial decomposition mode number; Using the meshing frequencies and frequency transitions at each stage as the filtering centers, the clean input signal is convolved and decomposed with the FIR filter bank: In the formula, For the first The second iteration One decoupled modal component; For a clean input signal; This is a convolution operation; Calculate the autocorrelation function of each decoupled mode component, take the local maximum point after crossing zero as the fault period estimate, and update the FIR filter coefficients based on the fault period estimate: In the formula, For the first The second iteration Fault cycle estimates for each mode; For the first Autocorrelation function of each mode; This is the filter coefficient update function; These are the updated FIR filter coefficients; After iterating to the maximum number of iterations, the initial decoupling decomposition is completed, and the initial decoupling modal components are obtained.

5. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 4, characterized in that, The step of performing anti-aliasing mode screening on the preliminary decoupled mode components, removing aliasing mode components, and obtaining non-aliasing decoupled mode components includes: Calculate the correlation coefficient between any two initially decoupled modal components and construct the correlation coefficient matrix: In the formula, For the first The modal component and the first The correlation coefficient between the modal components; It is the covariance function; It is the variance function; For the first The, the One initial decoupled modal component; For the modal component pairs with the highest correlation coefficients, calculate the correlation kurtosis of each modal component: In the formula, For the first The correlation kurtosis of each modal component; This represents the number of signal sampling points. The cumulative order of the period; The sampling period; For the first The modal component in the ... The amplitude of each sampling point; The modal components with higher correlation kurtosis are retained, while the aliased modal components with correlation coefficients higher than a preset threshold and lower correlation kurtosis are removed. This process is repeated until the number of modes reaches the target value, resulting in aliased decoupled modal components.

6. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 5, characterized in that, The step of refining fault features of the non-aliasing decoupled modal components and constructing a decoupled fault feature spectrum includes: Perform Hilbert transform on each non-aliased decoupled mode component to extract the fault envelope amplitude of each mode component: In the formula, For Hilbert transform operators; For the first Analytical signals of each modal component; For the first One non-aliased decoupled mode component; For integration variables; For the first The fault envelope amplitude of each modal component; Based on the fault envelope amplitude of each modal component, the independent fault features and modulation features at each level are integrated to construct a decoupled fault feature spectrum: In the formula, To decouple the fault characteristic spectrum; For the first Stage meshing frequency; This is the rotational frequency of a first-order solar detonator; The rotation frequency of a second-stage planetary gear; For the first Fault envelope amplitude spectrum of each modal component.

7. The method for extracting FMD fault features of inter-stage decoupling and anti-aliasing in wind power transmission chains according to claim 1, characterized in that, It also includes an adaptive iterative calibration step based on operating conditions: When the unit speed changes, the meshing frequency and rotation frequency of each stage are recalculated based on the real-time speed after the change, and the filter center frequency of the inter-stage decoupling FMD decomposition is updated. When the wind turbine is in a low-load idling condition, the health status reference signal under the current condition is collected, and the correlation coefficient threshold of the anti-aliasing mode screening is reset according to the reference signal. During long-term operation, acoustic signals of the transmission chain are periodically collected and fault features are extracted. The drift trend of the acoustic features with service time is fitted, and the long-period drift correction of the decoupled fault feature spectrum is performed based on the drift trend.

8. A system for extracting fault features of inter-stage decoupling and anti-aliasing FMD in wind power transmission chains, characterized in that, Includes the following modules: The signal acquisition and preprocessing module is used to acquire the original acoustic signals of the drive chain, the real-time speed of the unit and the vibration acceleration signals of the housing, and to preprocess the acquired signals to obtain a clean input signal. The interstage frequency calibration module is used to calculate the meshing frequency and rotational frequency of each stage based on the number of teeth of the transmission chain gears and the real-time rotational speed. The interstage decoupling FMD decomposition module is used to perform interstage decoupling FMD decomposition on the pure input signal based on the meshing frequency and the rotation frequency of each stage to obtain preliminary decoupling modal components; An anti-aliasing mode filtering module is used to perform anti-aliasing mode filtering on the preliminary decoupled mode components, remove aliasing mode components, and obtain non-aliasing decoupled mode components; The fault feature purification and decoupling spectrum construction module is used to purify the fault features of the non-aliased decoupled mode components and construct the decoupled fault feature spectrum.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the wind power transmission chain inter-stage decoupling anti-aliasing FMD fault feature extraction method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the wind power transmission chain decoupling anti-aliasing FMD fault feature extraction method according to any one of claims 1 to 7.

Images