Blade wind sweeping audio noise reduction method and system based on empirical mode decomposition
By decomposing and screening blade sweeping audio signals using empirical mode decomposition, the problems of low detection efficiency and severe noise interference in traditional methods are solved, achieving efficient noise suppression and fault feature preservation, and improving the accuracy of fault diagnosis.
Patent Information
- Application Number
- CN202511410569.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-29
- Publication Date
- 2026-01-09
AI Technical Summary
Traditional blade crack detection methods are inefficient and have a high false detection rate. Furthermore, the ambient noise (especially low-frequency wind noise) in the blade sweeping audio signal causes significant interference, leading to a decrease in the accuracy of fault diagnosis.
The empirical mode decomposition method is used to decompose the blade sweeping audio signal into intrinsic mode function (IMF) components and residual components. Useful IMF components that represent fault characteristics are selected, low-frequency wind noise components are removed, and the noise-reduced audio signal is reconstructed.
It significantly improves the signal-to-noise ratio, enhances the accuracy and recognition rate of fault diagnosis, provides cleaner and more reliable input data, and offers high-quality data support for the intelligent diagnostic model of blade cracks.
Smart Images
Figure CN121306160A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of audio signal noise reduction technology, and in particular to a method and system for audio noise reduction of blade sweeping based on empirical mode decomposition. Background Technology
[0002] With the large-scale commissioning of wind turbine generators, blades, as one of the most expensive components, are exposed to complex outdoor environments for extended periods, making them highly susceptible to cracks and other faults, leading to decreased power generation efficiency and reduced lightning protection performance. Traditional crack detection relies primarily on manual methods, which suffer from low efficiency, high cost, and a high risk of missed or false detections. While using wind turbine blade sweeping audio signals for fault diagnosis offers advantages such as convenient acquisition, no impact on unit operation, and good data consistency, in practical applications, the acquired audio signals often contain significant amounts of environmental noise, especially low-frequency wind noise interference. Furthermore, signal clarity is significantly affected by factors such as acquisition method and wind speed, severely reducing the accuracy of fault diagnosis and the model's classification ability. Therefore, there is an urgent need for an audio noise reduction method that can effectively suppress noise, improve the signal-to-noise ratio, and preserve fault characteristic frequencies. Summary of the Invention
[0003] This invention provides a method and system for denoising blade sweeping audio based on empirical mode decomposition, which solves the problems of low efficiency and high false detection rate of traditional manual blade crack detection, as well as the serious interference of environmental noise (especially low-frequency wind noise) in the blade sweeping audio signal, which leads to a decrease in the accuracy of fault diagnosis.
[0004] The objective of this invention can be achieved through the following technical solutions: The first aspect of this invention is to provide a blade sweeping audio noise reduction method based on empirical mode decomposition, comprising: S1: Collect the raw audio signal X(t) of blade sweeping during wind turbine operation; S2: Perform empirical mode decomposition on the original audio signal X(t), and adaptively decompose it into several intrinsic mode function components and a residual component r(t), wherein the higher-order IMF components with larger decomposition scale contain low-frequency signals, and the lower-order IMF components with smaller decomposition scale contain high-frequency signals. S3: Based on the needs of fault diagnosis, select useful IMF components that represent fault characteristics from the decomposed IMF components, and remove noise IMF components that are mainly low-frequency wind noise. S4: Reconstruct the filtered useful IMF components with the residual component r(t) to obtain the noise-reduced audio signal.
[0005] Furthermore, the specific sub-steps for performing empirical mode decomposition include: S21: Identify all local maxima and local minima in the original audio signal X(t); S22: Perform interpolation fitting on the maximum and minimum points respectively to obtain the upper envelope U(t) and the lower envelope L(t); S23: Calculate the average value of the upper and lower envelopes to obtain the mean envelope m(t); the specific process of obtaining the mean envelope m(t) is expressed by the formula:
[0006] In the formula, This represents a moment in the calculation process, based on the time in the first... The value of the envelope at time t and the first The mean of the values of the envelope at time t is used as the first time. The value of the time-mean envelope; S24: Subtract the mean envelope m(t) from the original audio signal X(t) to obtain the intermediate signal; S25: Determine whether the intermediate signal h(t) satisfies the conditions of IMF; if it does, then h(t) is taken as the first IMF component; if it does not, then h(t) is taken as the new X(t), and steps S21-S25 are repeated until the conditions are met; wherein, the first IMF component is taken as the current component IMF1; S26: Separate the current component IMF1 from the original audio signal X(t) to obtain the first-order residual signal. ; S27: The first-order residual signal As a new original signal, repeat steps S21-S26 to extract subsequent IMF components in sequence until the decomposition ends when the residual signal is a monotonic function or a constant function.
[0007] Furthermore, the conditions for the IMF in S25 include: (a) Throughout the entire signal segment, the number of extreme points is equal to or differs from the number of zero-crossing points by at most one; (b) At any time, the mean of the upper envelope defined by the local maximum point and the lower envelope defined by the local minimum point is zero.
[0008] Furthermore, the current component IMF1 is separated from the original audio signal X(t) to obtain a first-order residual signal, which is specifically expressed by the formula:
[0009] In the formula, This represents the first-order residual signal.
[0010] Furthermore, the intermediate signal is obtained by subtracting the mean envelope m(t) from the original audio signal X(t), and the intermediate signal is specifically expressed by the formula:
[0011] In the formula, This indicates an intermediate signal.
[0012] Furthermore, according to the needs of fault diagnosis, useful IMF components representing fault characteristics are screened from the decomposed IMF components, and noise IMF components dominated by low-frequency wind noise are removed, including: The criteria for selecting useful IMF components are: retaining low-order IMF components containing high-frequency signals of blade crack characteristics and removing high-order IMF components containing low-frequency signals of swept wind noise.
[0013] Further, the step of reconstructing the filtered useful IMF components and the residual component r(t) to obtain the denoised audio signal includes:
[0014] In the formula, Indicates the filtered first One useful IMF component This represents the final residual component. This indicates the number of all useful IMF components after filtering. This represents the audio signal after noise reduction.
[0015] A second aspect of the present invention is to provide a blade sweeping audio noise reduction system based on empirical mode decomposition, comprising: Data acquisition module: used to acquire the raw audio signal of blade sweeping during wind turbine operation; Audio decomposition module: used to perform empirical mode decomposition on the original audio signal, adaptively decomposing it into several intrinsic mode function components and a residual component, wherein the higher-order IMF components with larger decomposition scale contain low-frequency signals, and the lower-order IMF components with smaller decomposition scale contain high-frequency signals. Fault diagnosis module: Used to filter out useful IMF components that represent fault characteristics from the decomposed IMF components according to the fault diagnosis requirements, and to remove noise IMF components that are mainly low-frequency wind noise. Noise reduction module: used to reconstruct the filtered useful IMF components and the residual components to obtain the noise-reduced audio signal.
[0016] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the aforementioned blade sweeping audio noise reduction method based on empirical mode decomposition.
[0017] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the blade sweeping audio noise reduction method based on empirical mode decomposition.
[0018] Compared with existing technologies, the beneficial effects of this invention are as follows: It acquires the raw audio signal of the blade sweeping during wind turbine operation; performs empirical mode decomposition (IMF) on the raw audio signal, adaptively decomposing it into several intrinsic mode function (IMF) components and a residual component. The higher-order IMF components with larger decomposition scales contain low-frequency signals, while the lower-order IMF components with smaller decomposition scales contain high-frequency signals. This adaptive decomposition overcomes the limitations of traditional filtering methods. Based on the needs of fault diagnosis, useful IMF components representing fault characteristics are selected from the decomposed IMF components, while noise IMF components dominated by low-frequency wind noise are removed. This effectively removes noise while maximizing the retention of fault characteristic information in the signal, resulting in minimal signal distortion and ensuring the accuracy of subsequent fault diagnosis. The selected useful IMF components are reconstructed with the residual component to obtain the denoised audio signal. This method has a clear process, high computational efficiency, and is suitable for online or near-real-time monitoring and analysis of wind turbine status in engineering sites. It significantly improves the signal-to-noise ratio of the blade sweeping audio signal, providing cleaner and more reliable input data for the intelligent blade crack diagnosis model based on audio analysis, thereby improving the fault identification rate.
[0019] In summary, this invention solves the problems of low efficiency and high false detection rate of traditional manual blade crack detection, as well as the serious interference of environmental noise (especially low-frequency wind noise) in the blade sweeping audio signal, which leads to a decrease in the accuracy of fault diagnosis. This invention also improves the accuracy of blade sweeping audio noise reduction and reduces the interference of noise on the audio signal. Attached Figure Description
[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0021] Figure 1This invention provides a step-by-step flowchart of a blade sweeping audio noise reduction method based on empirical mode decomposition. Figure 2 This invention provides a schematic diagram of the module flow of a blade sweeping audio noise reduction system based on empirical mode decomposition. Detailed Implementation
[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0024] To address the problems existing in the background technology, a method and system for denoising blade sweeping audio based on empirical mode decomposition has been developed, which has significant practical implications.
[0025] like Figure 1 As shown, the first aspect of the present invention is to provide a method for denoising blade sweeping audio based on empirical mode decomposition, comprising the following steps: Step S1: Collect the raw audio signal X(t) of blade sweeping during wind turbine operation; Step S2: Perform Empirical Mode Decomposition (EMD) on the original audio signal X(t), adaptively decomposing it into several intrinsic mode function (IMF) components and a residual component r(t), wherein the higher-order IMF components with larger decomposition scale contain low-frequency signals, and the lower-order IMF components with smaller decomposition scale contain high-frequency signals.
[0026] The specific sub-steps for performing empirical mode decomposition include: S21: Identify all local maxima and local minima in the original audio signal X(t); S22: Perform interpolation fitting on the maximum and minimum points respectively to obtain the upper envelope U(t) and the lower envelope L(t); S23: Calculate the average value of the upper and lower envelopes to obtain the mean envelope m(t); the specific process of obtaining the mean envelope m(t) is expressed by the formula:
[0027] In the formula, This represents a moment in the calculation process, based on the time in the first... The value of the envelope at time t and the first The mean of the values of the envelope at time t is used as the first time. The value of the time-mean envelope; S24: Subtract the mean envelope m(t) from the original audio signal X(t) to obtain the intermediate signal; S25: Determine whether the intermediate signal h(t) satisfies the conditions of IMF; if it does, then h(t) is taken as the first IMF component; if it does not, then h(t) is taken as the new X(t), and steps S21-S25 are repeated until the conditions are met; wherein, the first IMF component is taken as the current component IMF1; S26: Separate the current component IMF1 from the original audio signal X(t) to obtain the first-order residual signal. ; S27: The first-order residual signal As a new original signal, repeat steps S21-S26 to extract subsequent IMF components in sequence until the decomposition ends when the residual signal is a monotonic function or a constant function.
[0028] Among them, the conditions for the IMF in S25 include: (a) Throughout the entire signal segment, the number of extreme points is equal to or differs from the number of zero-crossing points by at most one; (b) At any time, the mean of the upper envelope defined by the local maximum point and the lower envelope defined by the local minimum point is zero.
[0029] Specifically, the current component IMF1 is separated from the original audio signal X(t) to obtain the first-order residual signal, which is specifically expressed by the formula:
[0030] In the formula, This represents the first-order residual signal.
[0031] The intermediate signal is obtained by subtracting the mean envelope m(t) from the original audio signal X(t). The intermediate signal is specifically expressed by the following formula:
[0032] In the formula, This indicates an intermediate signal.
[0033] It should be noted that the core purpose of this operation is to achieve adaptive separation of noise and useful signals. Its function is as follows: EMD can automatically decompose the complex original audio signal into a series of intrinsic mode functions (IMFs) arranged in an ordered manner from high to low frequencies, based on the signal's local time-varying characteristics. Through this decomposition, sweeping noise concentrated in the low-frequency band and blade crack fault feature signals distributed in the high-frequency band are separated into different IMF components. This lays a solid foundation for subsequent precise noise removal and fault feature preservation and enhancement by screening IMF components, overcoming the poor adaptability of traditional fixed basis function filtering methods when processing such non-stationary signals.
[0034] Step S3: Based on the needs of fault diagnosis, select useful IMF components that represent fault characteristics from the decomposed IMF components, and remove noise IMF components that are mainly low-frequency wind noise.
[0035] The criteria for selecting useful IMF components are as follows: retaining low-order IMF components containing high-frequency signals of blade crack characteristics and removing high-order IMF components containing low-frequency signals of swept wind noise.
[0036] It should be noted that the core purpose of this operation is to achieve precise separation of signal and noise, thereby maximizing the signal-to-noise ratio of fault features. Its function is to selectively retain low-order IMF components containing high-frequency crack features while eliminating high-order IMF components dominated by low-frequency wind noise. This directly removes the noise that most strongly interferes with diagnosis, thus highlighting previously submerged, weak fault features related to blade cracks. This provides a clean and reliable signal source for subsequent fault identification and diagnosis, significantly improving the accuracy and reliability of the diagnostic model.
[0037] Step S4: Reconstruct the filtered useful IMF components with the residual component r(t) to obtain the noise-reduced audio signal.
[0038] The noise-reduced audio signal is specifically expressed by the formula:
[0039] In the formula, Indicates the filtered first One useful IMF component This represents the final residual component. This indicates the number of all useful IMF components after filtering. This represents the audio signal after noise reduction.
[0040] It should be noted that the core purpose of this operation is to synthesize a complete and high-fidelity denoised signal. Its function is to reconstruct a clean audio signal by linearly superimposing the useful high-frequency IMF components representing fault characteristics (after filtering) with the residual component r(t) representing the overall signal trend. This removes most noise interference while fully preserving the useful components and basic structure of the original signal. This not only eliminates the influence of low-frequency wind noise but also ensures the integrity of the fault characteristics in both the time and frequency domains, providing a high-quality data foundation for subsequent accurate fault diagnosis and analysis.
[0041] This concludes the embodiment.
[0042] like Figure 2 As shown, a second aspect of the present invention is to provide a blade sweeping audio noise reduction system based on empirical mode decomposition, comprising: Data acquisition module 101: used to acquire the raw audio signal of blade sweeping during wind turbine operation; Audio decomposition module 102: used to perform empirical mode decomposition on the original audio signal, adaptively decomposing it into several intrinsic mode function components and a residual component, wherein the high-order IMF component with a large decomposition scale contains low-frequency signals, and the low-order IMF component with a small decomposition scale contains high-frequency signals. Fault diagnosis module 103: Used to filter out useful IMF components representing fault characteristics from the decomposed IMF components according to the fault diagnosis requirements, and remove noise IMF components dominated by low-frequency wind noise. Noise reduction module 104: used to reconstruct the filtered useful IMF components and the residual components to obtain the noise-reduced audio signal.
[0043] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a blade sweeping audio noise reduction method based on empirical mode decomposition.
[0044] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements a blade sweeping audio noise reduction method based on empirical mode decomposition.
[0045] 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, optical storage, etc.) containing computer-usable program code.
[0046] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, 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.
[0047] 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 1 The function specified in one or more boxes.
[0048] 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.
[0049] 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 protection scope of the present invention.
Claims
1. A method for denoising blade-sweeping audio based on empirical mode decomposition, characterized in that, include: S1: Collect the raw audio signal X(t) of blade sweeping during wind turbine operation; S2: Perform empirical mode decomposition on the original audio signal X(t), and adaptively decompose it into several intrinsic mode function components and a residual component r(t), wherein the higher-order IMF components with larger decomposition scale contain low-frequency signals, and the lower-order IMF components with smaller decomposition scale contain high-frequency signals. S3: Based on the needs of fault diagnosis, select useful IMF components that represent fault characteristics from the decomposed IMF components, and remove noise IMF components that are mainly low-frequency wind noise. S4: Reconstruct the filtered useful IMF components with the residual component r(t) to obtain the noise-reduced audio signal.
2. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 1, characterized in that, The specific sub-steps for performing empirical mode decomposition include: S21: Identify all local maxima and local minima in the original audio signal X(t); S22: Perform interpolation fitting on the maximum and minimum points respectively to obtain the upper envelope U(t) and the lower envelope L(t); S23: Calculate the average value of the upper and lower envelopes to obtain the mean envelope m(t); the specific process of obtaining the mean envelope m(t) is expressed by the formula: In the formula, This represents a moment in the calculation process, based on the time in the first... The value of the envelope at time t and the first The mean of the values of the envelope at time t is used as the first time. The value of the time-mean envelope; S24: Subtract the mean envelope m(t) from the original audio signal X(t) to obtain the intermediate signal; S25: Determine whether the intermediate signal h(t) satisfies the conditions of IMF; if it does, then h(t) is taken as the first IMF component; if it does not, then h(t) is taken as the new X(t), and steps S21-S25 are repeated until the conditions are met; wherein, the first IMF component is taken as the current component IMF1; S26: Separate the current component IMF1 from the original audio signal X(t) to obtain the first-order residual signal. ; S27: The first-order residual signal As a new original signal, repeat steps S21-S26 to extract subsequent IMF components in sequence until the decomposition ends when the residual signal is a monotonic function or a constant function.
3. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 2, characterized in that, The conditions for the IMF in S25 include: (a) Throughout the entire signal segment, the number of extreme points is equal to or differs from the number of zero-crossing points by at most one; (b) At any time, the mean of the upper envelope defined by the local maximum point and the lower envelope defined by the local minimum point is zero.
4. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 2, characterized in that, The current component IMF1 is separated from the original audio signal X(t) to obtain a first-order residual signal, which is specifically expressed by the formula: In the formula, This represents the first-order residual signal.
5. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 2, characterized in that, The intermediate signal is obtained by subtracting the mean envelope m(t) from the original audio signal X(t). The intermediate signal is specifically expressed by the following formula: In the formula, This indicates an intermediate signal.
6. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 1, characterized in that, According to the requirements of fault diagnosis, useful IMF components representing fault characteristics are screened from the decomposed IMF components, and noise IMF components dominated by low-frequency wind noise are removed, including: The criteria for selecting useful IMF components are: retaining low-order IMF components containing high-frequency signals of blade crack characteristics and removing high-order IMF components containing low-frequency signals of swept wind noise.
7. The blade sweeping audio noise reduction method based on empirical mode decomposition according to claim 1, characterized in that, The step of reconstructing the filtered useful IMF components and the residual component r(t) to obtain the denoised audio signal includes: In the formula, Indicates the filtered first One useful IMF component This represents the final residual component. This indicates the number of all useful IMF components after filtering. This represents the audio signal after noise reduction.
8. A blade-sweeping audio noise reduction system based on empirical mode decomposition, characterized in that, include: Data acquisition module: used to acquire the raw audio signal of blade sweeping during wind turbine operation; Audio decomposition module: used to perform empirical mode decomposition on the original audio signal, adaptively decomposing it into several intrinsic mode function components and a residual component, wherein the higher-order IMF components with larger decomposition scale contain low-frequency signals, and the lower-order IMF components with smaller decomposition scale contain high-frequency signals. Fault diagnosis module: Used to filter out useful IMF components that represent fault characteristics from the decomposed IMF components according to the fault diagnosis requirements, and to remove noise IMF components that are mainly low-frequency wind noise. Noise reduction module: used to reconstruct the filtered useful IMF components and the residual components to obtain the noise-reduced audio signal.
9. An electronic device, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the blade sweeping audio noise reduction method based on empirical mode decomposition as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the blade sweeping audio noise reduction method based on empirical mode decomposition as described in any one of claims 1-7.