Signal processing method and system combining singular value decomposition and wavelet threshold denoising

Abstract

The invention discloses a singular value decomposition and wavelet threshold denoising combined signal processing method and system. The method comprises the following steps: acquiring a noise-containing fault signal; performing matrix conversion and singular value decomposition on the fault signal, and calculating an energy contribution value of each singular value; and processing the wavelet decomposition coefficient for the singular value of which the energy contribution value is higher than a set threshold value by adopting an improved threshold value and a threshold value function, and reconstructing an original signal through inverse wavelet transformation. By implementing the method provided by the invention, more flexible and effective denoising can be realized, and the accuracy and reliability of fault detection are improved.

Description

Technical Field

[0001] This invention relates to the field of low-voltage cable fault diagnosis technology, and in particular to a signal processing method and system that combines singular value decomposition and wavelet threshold denoising. Background Technology

[0002] In the field of fault detection for low-voltage power cables, signal quality plays a decisive role in the accuracy and reliability of fault feature extraction. However, in practice, the widespread noise interference, such as impulse noise, periodic interference, and non-stationary noise, greatly limits the accuracy of signal processing. Especially under the requirements of high-precision fault detection, traditional noise reduction methods often fall short.

[0003] Traditionally, Singular Value Decomposition (SVD) denoising methods rely on human experience to determine the number of singular values ​​and pre-set a fixed truncation threshold based on this experience. This method is not ideal for time-varying and non-stationary cable fault signals because the fixed threshold cannot adapt to changes in signal characteristics under different operating environments, easily leading to the loss of key signal features or ineffective removal of residual noise. Furthermore, although SVD can capture the main components of the signal through matrix decomposition, its ability to preserve local abrupt changes such as discharge pulses is limited, making it poorly suited for handling cable partial discharge faults with obvious transient characteristics. Wavelet transform, as a commonly used denoising technique, suppresses noise by setting different thresholds; however, single wavelet thresholding denoising algorithms have inherent limitations. Hard thresholding functions may cause signal distortion, while soft thresholding functions may cause loss of detailed information. In mixed noise environments such as white noise superimposed with power frequency interference, simultaneously preserving global features and reconstructing local details becomes a challenge. Traditional methods typically use a fixed threshold for processing, but as the number of wavelet decomposition layers increases, this may lead to excessive filtering of useful signals in low-frequency coefficients and residual noise in high-frequency coefficients.

[0004] In summary, existing technologies face numerous challenges in addressing cable fault detection in complex noise environments.

[0005] Therefore, it is necessary to design a new method to achieve more flexible and effective noise reduction and improve the accuracy and reliability of fault detection. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a signal processing method and system that combines singular value decomposition and wavelet threshold denoising.

[0007] To achieve the above objectives, the present invention employs the following technical solution: a signal processing method combining singular value decomposition and wavelet thresholding denoising, comprising: Acquire noisy fault signals; The fault signal is subjected to matrix transformation and singular value decomposition, and the energy contribution value of each singular value is calculated. For singular values ​​whose energy contribution exceeds a set threshold, an improved threshold and threshold function are used to process the wavelet decomposition coefficients, and the original signal is reconstructed through inverse wavelet transform.

[0008] The further technical solution is as follows: performing matrix transformation and singular value decomposition on the fault signal, and calculating the energy contribution value of each singular value, includes: The fault signal is decomposed by SVD and converted into a matrix according to the Hankel matrix construction principle; The matrix is ​​subjected to singular value decomposition to obtain a set of singular values ​​describing different components of the fault signal, and the energy contribution value corresponding to each singular value is calculated.

[0009] The further technical solution is as follows: the formula for calculating the energy contribution value is: ,in, ; ; It is a singular value.

[0010] The further technical solution is as follows: after performing matrix transformation and singular value decomposition on the fault signal, and calculating the energy contribution value of each singular value, it further includes: Set singular values ​​whose energy contribution is not higher than the set threshold to zero.

[0011] The further technical solution is as follows: For singular values ​​whose energy contribution values ​​exceed a set threshold, an improved threshold and threshold function are used to process the wavelet decomposition coefficients, and the original signal is reconstructed through inverse wavelet transform, including: Based on the signal characteristics corresponding to the singular values ​​whose energy contribution values ​​are higher than a set threshold, wavelet basis functions and decomposition levels are selected, and wavelet transform is performed on the signal to obtain wavelet decomposition coefficients. The wavelet decomposition coefficients are processed using an improved threshold and threshold function to obtain the processed wavelet decomposition coefficients. The original signal is reconstructed by inverse wavelet transform of the processed wavelet decomposition coefficients.

[0012] The further technical solution is: the improved threshold is ; The number of decomposition layers; For the improved threshold.

[0013] Its further technical solution is: the threshold function is expressed as a and b are function adjustment factors; a is an adjustable constant. .

[0014] The further technical solution is as follows: the wavelet basis function is the cofi5 wavelet basis function.

[0015] The further technical solution is that the number of decomposition layers is 5.

[0016] This invention also provides a signal processing system combining singular value decomposition and wavelet thresholding denoising, comprising: Acquisition unit, used to acquire fault signals containing noise; The calculation unit is used to perform matrix transformation and singular value decomposition on the fault signal, and to calculate the energy contribution value of each singular value. The reconstruction unit is used to process the wavelet decomposition coefficients of singular values ​​whose energy contribution values ​​are higher than a set threshold using an improved threshold and threshold function, and to reconstruct the original signal through inverse wavelet transform.

[0017] The advantages of this invention compared to existing technologies are as follows: This invention acquires noisy fault signals, first performs matrix transformation and singular value decomposition (SVD) to calculate the energy contribution of each singular value, and then filters out high-energy-contribution singular values ​​based on a set threshold to retain the main characteristic parts of the signal. Next, the signal components corresponding to these filtered singular values ​​are optimized using improved wavelet thresholding and threshold functions to optimize the wavelet decomposition coefficients, effectively suppressing residual noise while preserving important signal features. Finally, the original signal is reconstructed through inverse wavelet transform, achieving more flexible and effective noise removal in complex noisy environments, significantly improving the accuracy and reliability of fault detection. This method not only utilizes SVD to optimize signal representation from a global perspective but also leverages the advantages of wavelet transform in local details, enabling accurate extraction and recovery of key features of the fault signal even under strong noise interference.

[0018] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the signal processing method combining singular value decomposition and wavelet thresholding denoising provided in an embodiment of the present invention. Figure 2 A schematic diagram of the variation curve of singular value energy contribution value provided in an embodiment of the present invention; Figure 3 This is a comparison diagram of the fault signal before denoising and the fault signal after SVD denoising provided in an embodiment of the present invention; Figure 4 A comparison chart of the improved threshold functions provided in the embodiments of the present invention; Figure 5 A graph showing the processing of the original signal using MATLAB software provided in this embodiment of the invention; Figure 6 This is a schematic diagram comparing the signal after preliminary denoising and the signal after wavelet denoising provided in an embodiment of the present invention; Figure 7 A schematic block diagram of a signal processing system combining singular value decomposition and wavelet threshold denoising provided in an embodiment of the present invention; Figure 8 A schematic block diagram of a computer device provided for an embodiment of the present invention. Detailed Implementation

[0021] 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, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0023] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0024] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] Please see Figure 1 , Figure 1This is a flowchart illustrating a signal processing method combining singular value decomposition (SVD) and wavelet thresholding denoising, as provided in an embodiment of the present invention. This method is applied in servers. By combining SVD and wavelet thresholding denoising techniques, the noisy fault signal is first converted into a matrix, and key features are extracted using SVD. The energy contribution value of each singular value is calculated to identify and remove the main noise components while retaining important signal features. For the signal portion corresponding to singular values ​​with energy contribution values ​​higher than a set threshold, wavelet transform is performed using an improved wavelet basis function and an adaptively selected decomposition level. The wavelet coefficients are then processed using an optimized threshold and threshold function to achieve a more accurate denoising process. This method not only improves the flexibility and targeting of signal denoising but also enhances the accuracy and reliability of fault detection, making it particularly suitable for signal processing in complex noisy environments.

[0026] Figure 1 This is a schematic flowchart of the signal processing method combining singular value decomposition and wavelet thresholding denoising provided in an embodiment of the present invention. Figure 1 As shown, the method includes the following steps S110 to S130.

[0027] S110, Obtain the fault signal containing noise.

[0028] In this embodiment, the noisy fault signal refers to the electrical signal collected from the low-voltage power supply cable in actual operation, which is subject to interference from environmental noise and the electrical noise of the equipment itself. These signals contain characteristic information generated when the cable fails, but due to various noise interferences (such as electromagnetic interference, thermal noise, etc.), it is difficult to accurately extract fault features directly from these signals.

[0029] Specifically, acquiring noisy fault signals may involve using sensors or data acquisition systems to capture voltage or current changes in the cable's operating state. Because the operating environment of low-voltage power cables is complex and variable, the signals not only contain fault-related information but also a significant amount of irrelevant noise. For example, during power transmission, external electromagnetic interference can cause signal distortion; simultaneously, factors such as cable material aging and poor contact at joints can also generate additional noise.

[0030] Therefore, before processing these fault signals, it is essential to first clarify that the obtained original signal contains the aforementioned various noise components. This step is crucial for subsequent effective signal denoising using singular value decomposition combined with wavelet thresholding. By identifying and quantifying the forms and intensities of these noises, denoising algorithms can be designed more specifically to improve the accuracy and reliability of fault detection. This process lays the foundation for subsequent steps, including converting the noisy fault signal into a form suitable for SVD processing and optimizing the selection of the number of singular values ​​using the energy contribution principle, thereby achieving more accurate signal reconstruction and noise suppression.

[0031] S120. Perform matrix transformation and singular value decomposition on the fault signal, and calculate the energy contribution value of each singular value.

[0032] In this embodiment, singular values ​​refer to a set of values ​​obtained by decomposing a matrix using the Singular Value Decomposition (SVD) technique. These values ​​represent the weight or importance of the original signal (in this case, a noisy fault signal) on different components. Specifically, SVD decomposes an m×n matrix A into three matrices U, Σ, and V. T The product of, i.e., A = UΣV T , where Σ is a diagonal matrix containing singular values ​​arranged in descending order, and their squares represent the variance of the original matrix data in the corresponding direction.

[0033] The energy contribution value refers to the proportion of energy corresponding to each singular value to the total energy, used to assess the importance of that singular value in the representation of the original signal features. A higher energy contribution value indicates that the corresponding singular value is more critical in describing the original signal. Its calculation formula is usually based on the sum of squares of the singular values.

[0034] In one embodiment, step S120 described above may include steps S121 to S122.

[0035] S121. Perform SVD decomposition on the fault signal and convert the fault signal into a matrix according to the Hankel matrix construction principle.

[0036] This step involves converting the one-dimensional time series fault signal into a two-dimensional matrix form suitable for Singular Value Decomposition (SVD). Following the Hankel matrix construction principle, the time series of the fault signal is rearranged into an m×n matrix for subsequent Singular Value Decomposition.

[0037] S122. Perform singular value decomposition on the matrix to obtain a set of singular values ​​describing different components of the fault signal, and calculate the energy contribution value corresponding to each singular value.

[0038] The singular value vector obtained from singular value decomposition is: The formula for calculating the energy contribution value is as follows: ,in, ; ; It is a singular value.

[0039] In this step, the SVD technique is applied to decompose the matrix formed in the previous step, yielding a set of singular values. Then, using the aforementioned energy contribution formula, the energy contribution ratio of each singular value is calculated. This process helps identify which singular values ​​carry the most important signal information, thus laying the foundation for selecting appropriate singular values ​​for signal reconstruction.

[0040] This also includes setting singular values ​​whose energy contribution is below a set threshold to zero. This means retaining only singular values ​​that significantly contribute to the signal characteristics while removing unimportant singular values, thereby achieving signal denoising and feature extraction. For example, this invention mentions setting a contribution threshold. The value is 0.2%, meaning that only singular values ​​whose energy contribution rate is greater than or equal to 0.2% are considered valid; other singular values ​​are set to zero and excluded from signal reconstruction. This is done to remove noise components while preserving key signal characteristics.

[0041] Singular Value Decomposition (SVD) can transform the original signal into a set of singular value diagonal matrices describing different components of the fault signal. The energy contribution rate of each singular value after SVD decomposition is calculated, reflecting its contribution to the signal characteristics, and a contribution threshold is set. Singular values ​​below a threshold are set to zero, while high-energy singular values ​​are retained. A new diagonal matrix is ​​constructed using the remaining non-zero singular values, and an inverse transform operation is performed through this new matrix to achieve noise suppression and feature reconstruction of the signal.

[0042] S130. For singular values ​​whose energy contribution values ​​are higher than the set threshold, the wavelet decomposition coefficients are processed using an improved threshold and threshold function, and the original signal is reconstructed through inverse wavelet transform.

[0043] In one embodiment, step S130 described above may include steps S131 to S133.

[0044] S131. Select wavelet basis functions and decomposition levels based on the signal characteristics corresponding to singular values ​​whose energy contribution values ​​are higher than a set threshold, and perform wavelet transform on the signal to obtain wavelet decomposition coefficients.

[0045] In this embodiment, wavelet decomposition coefficients refer to the coefficients obtained by performing wavelet transform on the signal at different frequency bands. These coefficients represent the information of the signal in different frequency ranges. Wavelet transform decomposes the signal into a low-frequency (approximate) part and a high-frequency (detail) part.

[0046] In this embodiment, the coif5 wavelet basis function is selected. This is a wavelet basis function with good orthogonality and symmetry, which is suitable for signal analysis and processing in power systems.

[0047] The number of decomposition layers is set to 5, which means that the signal will be decomposed into one low-frequency component and four high-frequency components, each of which contains information within a specific frequency range.

[0048] S132. The wavelet decomposition coefficients are processed using an improved threshold and threshold function to obtain the processed wavelet decomposition coefficients.

[0049] An improved threshold design is used to more accurately distinguish between noise and useful signals. The improved threshold is... ; The number of decomposition layers; This is an improved threshold. The formula decreases as the number of decomposition levels increases, consistent with the characteristic that the amplitude of noise signals gradually decreases with increasing decomposition levels. The improved threshold increases with the number of decomposition levels... The wavelet coefficients gradually decrease, a trend consistent with the variation of wavelet coefficients in noise signals. This adjustment allows for a more effective distinction between the original signal and noise.

[0050] like Figure 4 As shown, by addressing the signal distortion inherent in the hard thresholding function and the detail loss inherent in the soft thresholding function, the problem of losing key abrupt change information during denoising is solved. The thresholding function is expressed as... a and b are function adjustment factors; a is an adjustable constant. .

[0051] Combining the advantages of hard and soft thresholding, this function is continuous at the threshold point, avoiding the discontinuity problem of the hard thresholding function and the bias caused by the soft thresholding function, thereby improving the quality of the reconstructed signal. Specifically, the improved thresholding function is continuous at the threshold point... At this point, the function is continuous, which compensates for the shortcomings of hard functions, such as the existence of discontinuities that cause signal oscillations and distortions. By adjusting the value of parameter 'a', the inherent bias of soft threshold functions is avoided, thus improving the noise reduction effect of the reconstructed signal.

[0052] Parameters a and b serve as adjustment factors, allowing the behavior of the threshold function to be modified according to actual needs. For example, in this embodiment, parameter a=5 is used to fine-tune the sensitivity of the threshold function.

[0053] S133. Perform inverse wavelet transform on the processed wavelet decomposition coefficients to reconstruct the original signal.

[0054] The wavelet decomposition coefficients, after being processed by the improved threshold and threshold function, will be used for inverse wavelet transform (IWT) to reconstruct the denoised original signal.

[0055] The inverse wavelet transform is the reverse process of the wavelet transform. It recombines the wavelet coefficients of each layer into the form of the original signal. However, the signal has already undergone denoising, effectively removing most of the noise interference, while retaining important fault characteristics and abrupt change information.

[0056] Through the above steps, the method proposed in this invention can not only effectively remove noise from low-voltage power supply cable signals, but also significantly improve the accuracy and reliability of subsequent fault feature extraction. Combining SVD and an improved wavelet thresholding denoising method, this technical solution demonstrates excellent performance improvements, particularly in key evaluation indicators such as signal-to-noise ratio, root mean square error, and waveform similarity coefficient.

[0057] In this embodiment, firstly, an orthogonal wavelet basis is used to perform a multi-scale wavelet decomposition algorithm on the noisy signal to obtain the spectral distribution of detail components and approximate components at each level. On the MATLAB platform, the decomposition level is optimized through spectral energy analysis. As the decomposition level progresses, the bandwidth of each frequency band exhibits an exponential decay characteristic. When the energy concentration of the characteristic signal within the target frequency band exceeds a preset threshold at a specific decomposition level, while the noise floor energy is below the tolerance value, that level is determined as the optimal decomposition scale. This means that at this decomposition level, noise can be effectively represented in the detail components.

[0058] This embodiment combines the advantages of Singular Value Decomposition (SVD) and wavelet thresholding denoising, processing signal noise from both global and local perspectives. Compared to traditional single denoising algorithms, this method can better restore the local variation characteristics of the signal and has better denoising performance.

[0059] First, the number of singular values ​​in SVD is optimized using the energy contribution spectrum principle to remove most of the noise from the original signal. SVD decomposition is then performed on the noisy signal. Based on the Hankel matrix construction principle, the noisy fault signal is transformed into an m×n matrix, and the singular values ​​in the matrix are extracted. The energy contribution value corresponding to each singular value is calculated, a contribution threshold is set, and singular values ​​with contribution rates less than the threshold are set to zero, while high-energy singular values ​​are retained for signal reconstruction. The non-zero singular values ​​are then combined to form a new optimized matrix, and an inverse transformation operation is performed to achieve noise suppression and feature reconstruction of the signal, resulting in a preliminarily denoised fault signal.

[0060] Wavelet thresholding: Subsequently, the residual noise is further processed by combining the improved wavelet thresholding function to clarify the local variation characteristics of the fault signal.

[0061] Based on the signal characteristics, an appropriate wavelet basis function and decomposition level are selected, and wavelet coefficients for different frequency band characteristics are obtained through wavelet transform. For example, in this embodiment, the coif5 wavelet basis function is selected, and the decomposition level is 5. An improved threshold and threshold function are applied to process the wavelet decomposition coefficients to obtain the processed wavelet decomposition coefficients. The improved threshold design decreases as the decomposition level increases, which conforms to the characteristic that the amplitude of the noise signal gradually decreases. The improved threshold function is continuous at the threshold point, avoiding the discontinuity problem of the hard threshold function and the deviation caused by the soft threshold function. The processed wavelet decomposition coefficients are then subjected to inverse wavelet transform to reconstruct the original signal, thereby achieving signal denoising.

[0062] Please see Figure 2 The energy contribution rate of the fault signal after SVD decomposition begins to be less than 0.2% at the 13th singular value. Therefore, it is believed that the first 12 singular values ​​retain the original characteristics of the signal. The remaining singular values ​​are set to zero and the signal is reconstructed.

[0063] like Figure 3 As shown, the original signal and the reconstructed signal after denoising in the above steps are compared. After SVD processing, most of the noise interference has been removed, and the temporal changes of the fault signal are clearly visible. The reconstructed signal after SVD denoising can retain most of the fault features, but some white noise still remains in the signal, which will affect the accuracy of subsequent fault detection and identification of the signal. It is necessary to perform secondary wavelet threshold denoising on the processed signal to further clarify the detailed features of the signal.

[0064] This embodiment employs an improved wavelet threshold and wavelet threshold function to process the SVD-denoised signal, suppressing residual noise interference and reconstructing the original signal. Specifically, appropriate wavelet basis functions and decomposition levels are selected based on signal characteristics, and wavelet coefficients for different frequency band characteristics are obtained through wavelet transform. The signal is decomposed into low-frequency and high-frequency signals, with further decomposition of the low-frequency components to refine the frequency band distribution until the desired denoising requirements are met. The decomposed wavelet coefficients are then processed using an improved threshold and threshold function. The main components of the real signal include power frequency signals and low-frequency injection signals, which have relatively low bandwidth and large wavelet coefficients; noise components are high-frequency, with smaller wavelet coefficients. High-amplitude useful signals are retained, while high-frequency, low-amplitude noise signals are filtered out. The wavelet coefficients after threshold function optimization are subjected to inverse transform processing to reconstruct the signal, achieving denoising.

[0065] like Figure 5As shown, the original signal was processed using MATLAB software. The spectral characteristics of the components after wavelet decomposition reveal that the detail component d1 has a wide spectral range and a uniform amplitude distribution, indicating that it contains a significant amount of noise. As the decomposition level increases, the spectral width gradually decreases until the spectrum of the d5 component mainly contains a 180Hz characteristic signal with very little noise interference. This indicates that at this decomposition level, noise can be effectively represented in the detail components. Then, an improved threshold function is used to process each detail component to remove noise interference, thus completing the denoising of the fault signal.

[0066] like Figure 6 As shown, by comparing the signal after preliminary denoising with the signal after wavelet denoising, it can be seen that this method can effectively suppress high-intensity noise interference, accurately reconstruct the original waveform change trend, and clearly see the local change characteristics of the signal.

[0067] like Figure 6 As shown, this embodiment combines the advantages of singular value decomposition and wavelet thresholding denoising, processing signal noise from both global and local perspectives. Compared to traditional single denoising algorithms, it can better restore the local variation characteristics of the signal and has better denoising performance. This method combines the advantages of singular value decomposition and wavelet thresholding denoising, not only effectively improving the signal denoising effect but also enhancing the noise resistance of the joint algorithm under strong noise interference, effectively preserving the temporal variation characteristics of the fault signal.

[0068] Through the above methods, this embodiment not only effectively improves the signal denoising effect but also enhances the noise resistance of the joint algorithm under strong noise interference, effectively preserving the temporal variation characteristics of the fault signal. Experimental results show that in the simulated fault signal with 10dB of noise added, the signal-to-noise ratio (SNR) improvement after SVD-WT denoising is the highest, reaching 38.14 dB, and it outperforms traditional denoising algorithms in all three evaluation indicators: SNR, root mean square error, and waveform similarity coefficient.

[0069] The aforementioned signal processing method combining singular value decomposition (SVD) and wavelet thresholding for denoising first acquires a noisy fault signal, performs matrix transformation and SVD, calculates the energy contribution of each singular value, and filters out high-energy-contribution singular values ​​based on a set threshold to preserve the main features of the signal. Next, the wavelet decomposition coefficients are optimized using an improved wavelet threshold and threshold function for the signal components corresponding to these filtered singular values, effectively suppressing residual noise while preserving important signal features. Finally, the original signal is reconstructed through inverse wavelet transform, achieving more flexible and effective noise removal in complex noisy environments, significantly improving the accuracy and reliability of fault detection. This method not only utilizes SVD to optimize signal representation from a global perspective but also leverages the advantages of wavelet transform in local details, enabling accurate extraction and recovery of key features of the fault signal even under strong noise interference.

[0070] Figure 7 This is a schematic block diagram of a signal processing system 300 combining singular value decomposition and wavelet thresholding denoising, provided in an embodiment of the present invention. Figure 7 As shown, corresponding to the above-described signal processing method combining singular value decomposition and wavelet thresholding denoising, the present invention also provides a signal processing system 300 combining singular value decomposition and wavelet thresholding denoising. This signal processing system 300 includes units for performing the above-described signal processing method combining singular value decomposition and wavelet thresholding denoising, and the system can be configured in a server. Specifically, please refer to... Figure 7 The signal processing system 300, which combines singular value decomposition and wavelet threshold denoising, includes an acquisition unit 301, a calculation unit 302, and a reconstruction unit 303.

[0071] The acquisition unit 301 is used to acquire a noisy fault signal; the calculation unit 302 is used to perform matrix transformation and singular value decomposition on the fault signal, and calculate the energy contribution value of each singular value; the reconstruction unit 303 is used to process the wavelet decomposition coefficients of singular values ​​with energy contribution values ​​higher than a set threshold using an improved threshold and threshold function, and reconstruct the original signal through inverse wavelet transform.

[0072] In one embodiment, the acquisition unit 301 includes: The transformation subunit is used to perform SVD decomposition on the fault signal and convert the fault signal into a matrix according to the Hankel matrix construction principle; the decomposition subunit is used to perform singular value decomposition on the matrix to obtain a set of singular values ​​describing different components of the fault signal and calculate the energy contribution value corresponding to each singular value.

[0073] In one embodiment, the system further includes a zeroing unit for zeroing singular values ​​whose energy contribution values ​​are not higher than a set threshold.

[0074] In one embodiment, the reconstruction unit 303 includes: The transform subunit is used to select wavelet basis functions and decomposition levels based on the signal characteristics corresponding to singular values ​​whose energy contribution values ​​are higher than a set threshold, and to perform wavelet transform on the signal to obtain wavelet decomposition coefficients; the processing subunit is used to process the wavelet decomposition coefficients using an improved threshold and threshold function to obtain processed wavelet decomposition coefficients; the inverse transform subunit is used to reconstruct the original signal by performing inverse wavelet transform on the processed wavelet decomposition coefficients.

[0075] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the signal processing system 300 combining singular value decomposition and wavelet threshold denoising and each unit can be referred to the corresponding description in the foregoing method embodiments. For the sake of convenience and brevity, it will not be repeated here.

[0076] The aforementioned signal processing system 300, which combines singular value decomposition and wavelet thresholding for denoising, can be implemented as a computer program. This computer program can be used in various applications such as... Figure 8 It runs on the computer device shown.

[0077] Please see Figure 8 , Figure 8 This is a schematic block diagram of a computer device provided in an embodiment of this application. The computer device 500 can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers.

[0078] See Figure 8 The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0079] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform a signal processing method that combines singular value decomposition and wavelet thresholding denoising.

[0080] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0081] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can perform a signal processing method that combines singular value decomposition and wavelet threshold denoising.

[0082] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0083] The processor 502 is used to run a computer program 5032 stored in the memory to implement all the steps of the signal processing method combining singular value decomposition and wavelet thresholding denoising.

[0084] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0085] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0086] Therefore, the present invention also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein when executed by a processor, the computer program causes the processor to perform all the steps of the signal processing method combining singular value decomposition and wavelet thresholding denoising.

[0087] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0088] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0089] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of each unit is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0090] The steps in the method of this invention can be adjusted, merged, or reduced in order according to actual needs. The units in the system of this invention can be merged, divided, or reduced according to actual needs. Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0091] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0092] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A signal processing method combining singular value decomposition and wavelet thresholding denoising, characterized in that, include: Acquire noisy fault signals; The fault signal is subjected to matrix transformation and singular value decomposition, and the energy contribution value of each singular value is calculated. For singular values ​​whose energy contribution exceeds a set threshold, an improved threshold and threshold function are used to process the wavelet decomposition coefficients, and the original signal is reconstructed through inverse wavelet transform.

2. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 1, characterized in that, The step of performing matrix transformation and singular value decomposition on the fault signal, and calculating the energy contribution value of each singular value, includes: The fault signal is decomposed by SVD and converted into a matrix according to the Hankel matrix construction principle; The matrix is ​​subjected to singular value decomposition to obtain a set of singular values ​​describing different components of the fault signal, and the energy contribution value corresponding to each singular value is calculated.

3. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 2, characterized in that, The formula for calculating the energy contribution value is as follows: ,in, ; ; It is a singular value.

4. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 3, characterized in that, After performing matrix transformation and singular value decomposition on the fault signal, and calculating the energy contribution value of each singular value, the method further includes: Set singular values ​​whose energy contribution is not higher than the set threshold to zero.

5. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 1, characterized in that, The process involves processing the wavelet decomposition coefficients of singular values ​​whose energy contribution exceeds a set threshold using an improved threshold and threshold function, and reconstructing the original signal through inverse wavelet transform, including: Based on the signal characteristics corresponding to the singular values ​​whose energy contribution values ​​are higher than a set threshold, wavelet basis functions and decomposition levels are selected, and wavelet transform is performed on the signal to obtain wavelet decomposition coefficients. The wavelet decomposition coefficients are processed using an improved threshold and threshold function to obtain the processed wavelet decomposition coefficients. The original signal is reconstructed by inverse wavelet transform of the processed wavelet decomposition coefficients.

6. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 5, characterized in that, The improved threshold is ; The number of decomposition layers; For the improved threshold.

7. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 6, characterized in that, The threshold function is expressed as follows: a and b are function adjustment factors; a is an adjustable constant. .

8. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 4, characterized in that, The wavelet basis function is the cofi5 wavelet basis function.

9. The signal processing method combining singular value decomposition and wavelet thresholding denoising according to claim 4, characterized in that, The number of decomposition layers is 5.

10. A signal processing system combining singular value decomposition and wavelet thresholding denoising, characterized in that, include: Acquisition unit, used to acquire fault signals containing noise; The calculation unit is used to perform matrix transformation and singular value decomposition on the fault signal, and to calculate the energy contribution value of each singular value. The reconstruction unit is used to process the wavelet decomposition coefficients of singular values ​​whose energy contribution values ​​are higher than a set threshold using an improved threshold and threshold function, and to reconstruct the original signal through inverse wavelet transform.

Images