Denoising method and device for discharge breakdown ultrasonic signal, computer equipment and storage medium

Abstract

The invention relates to a discharge breakdown ultrasonic signal denoising method and device, computer equipment and a storage medium. The method comprises the following steps: performing energy spectrum analysis on a collected discharge breakdown ultrasonic signal to obtain an energy spectrum density curve representing pulse energy and frequency correlation characteristics; determining the maximum layer number of wavelet decomposition in combination with the actual working condition demand of insulation monitoring of the high-voltage equipment; performing wavelet decomposition layer by layer according to a preset maximum decomposition layer number by taking a discharge breakdown ultrasonic signal as input to obtain an optimal wavelet of each decomposition layer; and integrating the optimal wavelets of each level to construct an optimal wavelet set, performing wavelet threshold de-noising processing on the discharge breakdown ultrasonic signal based on the set to obtain an approximation coefficient and a detail coefficient of each level, reconstructing the processed coefficients through wavelet inverse transformation, and finally outputting the de-noised discharge breakdown ultrasonic signal. By adopting the method, the de-noising effect can be remarkably improved while the integrity of the pulse waveform is ensured.

Description

Technical Field

[0001] This application relates to the field of signal denoising technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for denoising discharge breakdown ultrasonic signals. Background Technology

[0002] With the development of signal denoising technology, correlation-based wavelet selection (CBWS) and energy-based wavelet selection (EBWS) methods have emerged. The CBWS method selects the same wavelet function for all decomposition levels by calculating the overall correlation with an ideal reference pulse. However, this method ignores the differences in signal and noise performance at different scales, cannot achieve adaptive optimal filtering, and its performance is highly dependent on a clean reference pulse that is difficult to obtain in a noisy environment, thus limiting its effectiveness and leading to signal distortion after denoising. The EBWS method aims to maximize the energy of the approximate coefficients at each decomposition level. However, as the decomposition deepens, the energy of the discharge breakdown pulse naturally transfers to the detail coefficients. At this point, the EBWS selection criterion actively disperses the pulse energy instead of concentrating it, directly violating the basic principles of wavelet denoising and also resulting in signal distortion after denoising. Summary of the Invention

[0003] Based on this, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for denoising discharge breakdown ultrasonic signals that can accurately match the characteristics of different decomposition levels of discharge breakdown ultrasonic signals, avoid signal distortion caused by improper selection of wavelet basis, and adaptively determine the optimal number of decomposition levels.

[0004] In a first aspect, this application provides a method for denoising discharge breakdown ultrasonic signals, comprising:

[0005] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0006] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0007] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0008] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0009] In one embodiment, based on the discharge breakdown ultrasonic signal, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained, including:

[0010] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0011] In one embodiment, based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers for wavelet decomposition of the discharge breakdown ultrasonic signal is determined, including:

[0012] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0013] In one embodiment, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level, including:

[0014] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0015] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0016] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0017] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0018] In one embodiment, wavelet thresholding denoising is performed on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain approximation coefficients and detail coefficients after each level of processing, including:

[0019] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0020] Secondly, this application also provides a noise reduction device for discharge breakdown ultrasonic signals, comprising:

[0021] The analysis module is used to obtain the energy spectral density curve characterizing the relationship between pulse energy and frequency based on the discharge breakdown ultrasonic signal;

[0022] The maximum decomposition layer determination module is used to determine the maximum decomposition layer when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the actual needs of high-voltage equipment insulation monitoring.

[0023] The decomposition module is used to decompose the discharge breakdown ultrasonic signal as the current input signal according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level.

[0024] The denoising and reconstruction module is used to construct an optimal wavelet set based on the optimal wavelets at each decomposition level; based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level; wavelet reconstruction is performed based on the approximation coefficients and detail coefficients after processing at each level to obtain the denoised discharge breakdown ultrasonic signal.

[0025] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0026] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0027] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0028] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0029] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0030] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0031] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0032] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0033] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0034] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0035] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0036] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0037] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0038] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0039] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0040] The aforementioned method, apparatus, computer equipment, computer-readable storage medium, and computer program product for denoising discharge breakdown ultrasonic signals first perform energy spectrum analysis on the acquired discharge breakdown ultrasonic signal to obtain the energy spectral density curve characterizing the correlation between pulse energy and frequency. This yields the energy distribution profile at each frequency, clarifies the core frequency band range of the signal energy, and provides data support for subsequent wavelet parameter selection. Secondly, considering the actual operating conditions of high-voltage equipment insulation monitoring, the maximum number of wavelet decomposition levels is determined to ensure the decomposition scale matches the actual application scenario. Then, using the discharge breakdown ultrasonic signal as input, wavelet decomposition is performed layer by layer according to the preset maximum decomposition level to obtain the optimal wavelet at each decomposition level. Finally, the optimal wavelets at each level are integrated to construct an optimal wavelet set. Based on this set, wavelet threshold denoising processing is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients at each level. The processed coefficients are then reconstructed through inverse wavelet transform, ultimately outputting the denoised discharge breakdown ultrasonic signal. By using energy spectral density curves to assist in wavelet decomposition parameter decision-making and combining the actual needs of high-voltage equipment insulation monitoring to determine the maximum number of decomposition levels, this method overcomes the limitations of traditional wavelet denoising that relies on manual experience to set the decomposition scale. It effectively avoids computational redundancy caused by over-decomposition or signal feature loss caused by under-decomposition. The selection of optimal wavelets at each level and the construction of optimal wavelet sets break through the bottleneck of adapting a single wavelet basis to signals at all scales, significantly improving the adaptability of the wavelet basis to the characteristics of discharge breakdown ultrasonic signals at different scales. Without the need for a preset fixed wavelet basis, it achieves accurate separation of transient pulse features and broadband noise interference in discharge breakdown ultrasonic signals, significantly improving the signal-to-noise ratio and waveform fidelity of the denoised signal in a strong noise environment. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is an application environment diagram of a method for denoising discharge breakdown ultrasonic signals in one embodiment;

[0043] Figure 2 This is a flowchart illustrating a method for denoising ultrasonic signals during discharge breakdown in one embodiment.

[0044] Figure 3 This is a flowchart illustrating the steps of a wavelet selection method based on signal-to-noise ratio in one embodiment;

[0045] Figure 4This is a structural block diagram of a noise reduction device for discharge breakdown ultrasonic signals in one embodiment;

[0046] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0049] The noise reduction method for discharge breakdown ultrasonic signals provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on a cloud or other network server. First, the acquired discharge breakdown ultrasonic signal is subjected to energy spectrum analysis to obtain the energy spectral density curve characterizing the correlation between pulse energy and frequency. This yields the energy distribution profile at each frequency, clarifies the core frequency band range of the signal energy, and provides data support for subsequent wavelet parameter selection. Second, considering the actual operating conditions of high-voltage equipment insulation monitoring, the maximum number of wavelet decomposition levels is determined to ensure that the decomposition scale matches the actual application scenario. Then, using the discharge breakdown ultrasonic signal as input, wavelet decomposition is performed layer by layer according to the preset maximum decomposition level to obtain the optimal wavelet at each decomposition level. Finally, the optimal wavelets at each level are integrated to construct an optimal wavelet set. Based on this set, wavelet threshold denoising processing is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients at each level. The processed coefficients are then reconstructed through inverse wavelet transform, ultimately outputting the denoised discharge breakdown ultrasonic signal.

[0050] Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, and projection equipment. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0051] In one exemplary embodiment, such as Figure 2 As shown, a method for denoising discharge breakdown ultrasonic signals is provided, which is then applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 202 to 208. Wherein:

[0052] Step 202: Based on the discharge breakdown ultrasonic signal, obtain the energy spectral density curve characterizing the relationship between pulse energy and frequency;

[0053] Among them, the ultrasonic signal of electrical discharge breakdown refers to the ultrasonic signal radiated by the instantaneous mechanical vibration generated during the partial discharge process when the insulation of high-voltage equipment deteriorates (such as air gap discharge, surface discharge, etc.). This signal contains key information about the insulation fault of the equipment, but it is often interfered with by environmental noise, equipment operating vibration noise, etc. Energy Spectral Density (ESD, also known as energy spectrum) is a physical quantity that describes the energy distribution characteristics of non-periodic signals (such as electrical discharge breakdown ultrasonic pulses, which are transient non-periodic signals) in the frequency domain. Its core definition is the distribution function of the signal energy as a function of frequency. The horizontal axis of the energy spectral density curve is frequency (corresponding to different frequency components of the ultrasonic signal), and the vertical axis is energy density (corresponding to the signal energy carried by the corresponding frequency component). The peak position of the curve corresponds to the frequency at which the signal energy is most concentrated, and the frequency range covered by the curve corresponds to the effective frequency band of the signal.

[0054] Step 204: Based on the actual needs of high-voltage equipment insulation monitoring, determine the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal;

[0055] The maximum decomposition level refers to the highest number of decomposition levels that can be performed on the ultrasonic discharge breakdown signal under mathematical constraints and actual monitoring requirements when performing discrete wavelet transform. Each decomposition level breaks down the signal into approximate coefficients (low frequency, corresponding to the main features of the signal) and detail coefficients (high frequency, mostly noise or signal details). The higher the level, the more detailed the frequency domain division of the signal.

[0056] Specifically, the maximum number of decomposition levels is determined using the Number of Wavelet Decomposition Levels Selection (NWDLS) method.

[0057] Step 206: Take the discharge breakdown ultrasonic signal as the current input signal, decompose the current input signal according to the maximum decomposition level, and obtain the optimal wavelet of each decomposition level;

[0058] Here, each decomposition level refers to all wavelet decomposition levels from level 1 to the maximum decomposition level. The optimal wavelet refers to the wavelet basis with the strongest signal-noise separation capability selected by the signal-to-noise ratio based wavelet selection (SNRBWS) method for a certain level of input signal.

[0059] Step 208: Construct an optimal wavelet set based on the optimal wavelets of each decomposition level; perform wavelet threshold denoising on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain the approximation coefficients and detail coefficients after processing at each level; perform wavelet reconstruction based on the approximation coefficients and detail coefficients after processing at each level to obtain the denoised discharge breakdown ultrasonic signal.

[0060] Wavelet thresholding denoising refers to the process of processing the coefficients after wavelet decomposition based on preset threshold rules and threshold functions, removing small-amplitude coefficients corresponding to noise, and retaining effective signal coefficients. Wavelet reconstruction, also known as inverse discrete wavelet transform (IDWT), refers to the process of restoring the approximate coefficients and detail coefficients after processing at each level to the time-domain signal.

[0061] In the above-mentioned denoising method for discharge breakdown ultrasonic signals, the energy spectrum analysis of the acquired discharge breakdown ultrasonic signal is first performed to obtain the energy spectral density curve characterizing the correlation between pulse energy and frequency. This yields the energy distribution profile at each frequency, clarifies the core frequency band range of the signal energy, and provides data support for subsequent wavelet parameter selection. Secondly, considering the actual operating conditions of high-voltage equipment insulation monitoring, the maximum number of wavelet decomposition levels is determined to ensure that the decomposition scale matches the actual application scenario. Subsequently, using the discharge breakdown ultrasonic signal as input, wavelet decomposition is performed layer by layer according to the preset maximum decomposition level to obtain the optimal wavelet at each decomposition level. Finally, the optimal wavelets at each level are integrated to construct an optimal wavelet set. Based on this set, wavelet threshold denoising processing is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients at each level. Then, the processed coefficients are reconstructed through inverse wavelet transform, and finally, the denoised discharge breakdown ultrasonic signal is output. By using energy spectral density curves to assist in wavelet decomposition parameter decision-making and combining the actual needs of high-voltage equipment insulation monitoring to determine the maximum number of decomposition levels, this method overcomes the limitations of traditional wavelet denoising that relies on manual experience to set the decomposition scale. It effectively avoids computational redundancy caused by over-decomposition or signal feature loss caused by under-decomposition. The selection of optimal wavelets at each level and the construction of optimal wavelet sets break through the bottleneck of adapting a single wavelet basis to signals at all scales, significantly improving the adaptability of the wavelet basis to the characteristics of discharge breakdown ultrasonic signals at different scales. Without the need for a preset fixed wavelet basis, it achieves accurate separation of transient pulse features and broadband noise interference in discharge breakdown ultrasonic signals, significantly improving the signal-to-noise ratio and waveform fidelity of the denoised signal in a strong noise environment.

[0062] In one embodiment, based on the discharge breakdown ultrasonic signal, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained, including:

[0063] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0064] Specifically, in response to the characteristic analysis requirements of discharge breakdown ultrasonic signals, the reference pulse of the signal is used as the analysis object. The reference pulse is processed by frequency domain analysis to calculate the energy spectral density that can characterize the distribution law of pulse energy at different frequency components. Based on the energy spectral density data, an energy spectral density curve that intuitively reflects the correspondence between pulse energy and frequency is further generated.

[0065] In the above embodiments, by obtaining the energy spectral density and corresponding curve through frequency domain analysis of the reference pulse of the discharge breakdown ultrasonic signal, the distribution characteristics of signal energy in each frequency band can be clearly quantified, and the core energy frequency range of the discharge signal can be accurately located. This provides an objective and quantitative basis for determining the number of wavelet decomposition layers, effectively avoiding the signal feature loss or redundant noise introduction caused by empirical layer selection. At the same time, the energy spectral density curve can intuitively present the frequency distribution law of signal energy, laying the foundation for determining the maximum number of decomposition layers.

[0066] In one embodiment, based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers for wavelet decomposition of the discharge breakdown ultrasonic signal is determined, including:

[0067] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0068] Specifically, firstly, the energy spectral density of the discharge breakdown ultrasonic signal reference pulse is used as a reference. The energy spectral density provides a complete energy profile of the pulse across the entire frequency band. Secondly, an energy retention threshold is set, and a critical frequency is determined. This energy retention threshold is preset based on actual needs, aiming to define the "important signal portion" that needs to be retained. Subsequently, on the energy spectral density curve, energy is accumulated from the lowest frequency towards higher frequencies, precisely identifying the frequency point where the accumulated energy first reaches or exceeds the set energy retention threshold. This frequency is defined as the critical frequency F. min F min This represents the lowest frequency boundary that wavelet decomposition needs to process in order to retain the majority of pulse energy. Then, the sampling frequency F is determined based on the sampling theorem. S After determining the key frequency F min and the sampling frequency F of the initial input signal S Then, substituting this into the equation yields the maximum number of wavelet decomposition levels, J:

[0069]

[0070] In the formula, fix is ​​the floor function.

[0071] In the above embodiments, the key frequency is determined by energy accumulation and threshold determination, and then the maximum number of wavelet decomposition layers is calculated, which can accurately anchor the core energy frequency band of the discharge breakdown ultrasonic signal. The maximum number of layers determined in this way can avoid the inability to fully separate the effective features and noise in the signal due to too few decomposition layers, and can also prevent the introduction of redundant calculation and invalid high-frequency noise components due to too many layers. At the same time, this method provides a scientific and quantitative hierarchical basis for wavelet decomposition, improves the accuracy of subsequent wavelet decomposition in separating signals and noise, and ensures the integrity of key features of the discharge signal.

[0072] In one embodiment, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level, including:

[0073] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0074] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0075] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0076] Specifically, S4: Perform individual testing and single-level decomposition on the candidate wavelet library. At the current j-th level of decomposition, a systematic screening process is initiated to test and evaluate each candidate wavelet function contained in the wavelet library. For each candidate wavelet, a single-level discrete wavelet decomposition is performed, decomposing the current input signal into two components: approximation coefficients aj (reflecting the low-frequency trend components of the signal) and detail coefficients dj (reflecting the high-frequency detail components of the signal).

[0077] S5: Signal band identification based on peak comparison. After obtaining the approximation coefficients and detail coefficients, the maximum absolute value of these two sets of coefficients is compared. The principle of comparison is: the sub-band with the larger absolute value of its coefficients indicates that the sub-band contains stronger signal characteristics. Specifically, the judgment rule is: if the maximum value of the detail coefficients is greater than or equal to the maximum value of the approximation coefficients, the detail sub-band is judged as a "signal band" and the approximation sub-band as a "noise band"; otherwise, the approximation sub-band is judged as a "signal band" and the detail sub-band as a "noise band".

[0078] S6: Calculate Peak Signal-to-Noise Ratio (SNR). After identifying the signal band and noise band, it is necessary to quantitatively evaluate the performance of the current candidate wavelet. SNR is used as the evaluation metric, and its calculation formula is the ratio of the absolute maximum value of the signal band coefficients (signal) to the absolute maximum value of the noise band coefficients (noise):

[0079]

[0080] S7: Determine the optimal wavelet for the current level. After traversing all candidate wavelets in the wavelet library and calculating their respective peak signal-to-noise ratios (SNRs), compare all SNR values ​​and select the wavelet function that maximizes the SNR. The selected wavelet will be recorded as the optimal wavelet Ψj for the current decomposition level j.

[0081] S8: Update the signal and proceed to the next level. After completing the wavelet selection for the current level, input data needs to be prepared for the next level of processing. Following the standard wavelet decomposition process, the approximation coefficients aj of the current level are used as the input signal for the next level of decomposition. Simultaneously, the level index j is incremented by 1, and then a new round of wavelet selection begins.

[0082] In the above embodiments, by traversing and screening layer by layer, quantifying and determining signal and noise bands, and selecting wavelets with peak signal-to-noise ratio as the core indicator, the wavelet basis with the strongest signal and noise separation capability can be matched to each decomposition level, effectively improving the accuracy of signal decomposition in different frequency bands. During the hierarchical iteration process, the approximation coefficient of the previous level is used as the input signal of the next level, which conforms to the characteristics of wavelet multi-scale decomposition and ensures the targeting and coherence of each level of screening. The optimal wavelet set finally selected allows the effective signal energy to be more concentrated and the noise components to be more dispersed during subsequent signal decomposition, laying a good foundation for subsequent threshold denoising, reducing the loss of effective signal features and the residue of noise, and improving the overall effect of denoising of discharge breakdown ultrasonic signals.

[0083] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0084] Specifically, the Dobesi wavelet has good tight support and the higher the order, the stronger the regularity, and can adapt to the signal characteristics of different frequency bands; the Symlets wavelet is an improved version of the Dobesi wavelet with better symmetry, which can reduce the phase distortion after signal decomposition; the Coiflets wavelet has both good tight support and approximate symmetry, and the reconstructed signal has higher smoothness, making it suitable for preserving the waveform details of the discharge pulse.

[0085] In the above embodiments, the constructed multi-type, wide-order wavelet function library covers wavelet bases with different characteristics such as tight support, regularity, and symmetry, which can adapt to the frequency band feature differences of different decomposition levels of discharge breakdown ultrasonic signals. The rich selection of wavelet bases can ensure that the separation requirements of signals and noise at each level can be accurately matched during the optimal wavelet screening process, improve the ability of wavelet decomposition to extract signal features and the separation accuracy of noise, lay a good foundation for subsequent wavelet threshold denoising, and ensure the feature integrity of the discharge signal after denoising.

[0086] In one embodiment, wavelet thresholding denoising is performed on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain approximation coefficients and detail coefficients after each level of processing, including:

[0087] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0088] Specifically, using the pre-selected optimal wavelet set as the decomposition tool, the discharge breakdown ultrasonic signal is subjected to hierarchical wavelet decomposition to obtain the approximation coefficients and detail coefficients corresponding to each level. Based on the preset threshold rules and threshold functions, the detail coefficients of each level are processed: coefficients with amplitudes below the threshold are regarded as noise and are subjected to shrinkage (soft threshold) or zeroing (hard threshold) operations. The approximation coefficients remain unchanged or are only slightly adjusted, and finally the approximation coefficients and detail coefficients of each level are obtained.

[0089] In the above embodiments, the hierarchical decomposition is carried out by leveraging the adaptability of the optimal wavelet set, so that the signal and noise show clear amplitude differences at the coefficient level. Then, the coefficient components corresponding to noise are accurately distinguished and removed by threshold rules, which not only avoids the accidental deletion of effective signal coefficients, but also significantly reduces noise residue. The coefficients obtained after processing retain the core features of the discharge ultrasonic signal, providing a high-quality frequency domain carrier for subsequent wavelet reconstruction, and ultimately improving the signal-to-noise ratio of the denoised signal, ensuring the accuracy and reliability of insulation fault diagnosis of high-voltage equipment.

[0090] In one embodiment, such as Figure 3The image shows a method for denoising discharge breakdown ultrasonic signals in a specific embodiment, including:

[0091] S1: Constructing the wavelet function library and initializing parameters. First, a candidate library containing various types of wavelet functions needs to be constructed. After the complete wavelet library is established, initialization settings are performed, including setting the initial decomposition level j=1 and setting the discharge breakdown ultrasonic signal to be processed as the initial input signal S.

[0092] Wavelet libraries typically comprise three classical wavelet families: the Dobessi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family. Each family contains multiple wavelets of different orders. Specifically, the Dobessi wavelet family covers orders 1-40, the Symlets wavelet family covers orders 2-15, and the Coiflets wavelet family covers orders 1-5. This design aims to provide a sufficiently rich selection of wavelet basis functions to accommodate ultrasonic discharge breakdown signals with varying characteristics.

[0093] S2: Determine the maximum number of decomposition levels J. This is determined using the Number of Wavelet Decomposition Levels Selection (NWDLS) method.

[0094] First, the energy spectral density of the discharge breakdown ultrasonic signal reference pulse is used as a reference. The energy spectral density provides a complete energy profile of the pulse across the entire frequency band. Second, an energy retention threshold is set, and a critical frequency is determined. This energy retention threshold is preset based on actual needs, aiming to define the "important signal portion" that needs to be retained. Then, on the energy spectral density curve, energy is accumulated from the lowest frequency towards higher frequencies, precisely identifying the frequency point where the accumulated energy first reaches or exceeds the set energy retention threshold. This frequency is defined as the critical frequency F. min F min This represents the lowest frequency boundary that wavelet decomposition needs to process in order to retain the majority of pulse energy. Then, the sampling frequency F is determined based on the sampling theorem. S After determining the key frequency F min and the sampling frequency F of the initial input signal S Then, substituting this into the equation yields the maximum number of wavelet decomposition levels, J:

[0095] (1)

[0096] In the formula, fix is ​​the floor function. The theoretical basis of formula (1) is the characteristic of binary wavelet transform: each level of decomposition divides the frequency range into two halves. In addition to providing good wavelet decomposition by selecting an appropriate number of decomposition levels, it also avoids the need for maximum level decomposition of the signal, saving a lot of processing time in wavelet selection and denoising.

[0097] S3: Processing is performed cyclically according to the decomposition level. Starting from the first level, each level is processed sequentially until the preset maximum number of decomposition levels J is reached. The processing objective of each level is to select the optimal wavelet from the wavelet library that best suits the current scale features. When entering a new decomposition level, the relevant parameters for wavelet selection are reset to ensure that the selection process at each level is independent and fair.

[0098] S4: Perform individual testing and single-level decomposition on the candidate wavelet library. At the current j-th level of decomposition, a systematic screening process is initiated to test and evaluate each candidate wavelet function in the wavelet library. For each candidate wavelet, a single-level discrete wavelet decomposition is performed, decomposing the current input signal into two components: approximation coefficients aj (reflecting the low-frequency trend components of the signal) and detail coefficients dj (reflecting the high-frequency detail components of the signal). This step essentially tests the decomposition capability of each wavelet function for the current scale signal, providing a data foundation for subsequent signal band identification.

[0099] S5: Signal Band Identification Based on Peak Comparison. After obtaining the approximation coefficients and detail coefficients, the maximum absolute value of these two sets of coefficients is compared. The principle of comparison is: the sub-band with the larger absolute value of its coefficients indicates that the sub-band contains stronger signal characteristics. Specifically, if the maximum value of the detail coefficients is greater than or equal to the maximum value of the approximation coefficients, the detail sub-band is determined to be a "signal band," and the approximation sub-band is determined to be a "noise band"; conversely, if the maximum value is less than or equal to the maximum value of the approximation coefficients, the approximation sub-band is determined to be a "signal band," and the detail sub-band is determined to be a "noise band." This peak comparison-based determination method can effectively identify the frequency bands where the pulse energy of the discharge breakdown ultrasonic signal is mainly concentrated.

[0100] S6: Calculate the peak signal-to-noise ratio (SNR). After identifying the signal band and noise band, it is necessary to quantitatively evaluate the performance of the current candidate wavelet. The peak signal-to-noise ratio (SNR) is used as the evaluation index, and its calculation formula is Equation (2): the ratio of the absolute maximum value of the signal band coefficient (signal) to the absolute maximum value of the noise band coefficient (noise).

[0101] (2)

[0102] Formula (2) is relatively simple to calculate, but its physical meaning is clear: the larger the ratio, the better the wavelet can concentrate the signal energy into a specific sub-band during decomposition, while suppressing noise in another sub-band. This is exactly the characteristic that high-quality denoising is expected to achieve.

[0103] S7: Determine the optimal wavelet for the current level. After traversing all candidate wavelets in the wavelet library and calculating their respective peak signal-to-noise ratios (SNRs), compare all SNR values ​​and select the wavelet function that maximizes the SNR. The selected wavelet will be recorded as the optimal wavelet Ψj for the current decomposition level j. This selection process ensures that the wavelet basis functions used at each scale are those that best enhance the signal and suppress noise.

[0104] S8: Update the signal and proceed to the next level. After completing the wavelet selection for the current level, input data needs to be prepared for the next level of processing. Following the standard wavelet decomposition process, the approximation coefficients aj of the current level are used as the input signal for the next level of decomposition. Simultaneously, the level index j is incremented by 1, and then the process returns to S3 to begin a new round of wavelet selection. This iterative process ensures that the most suitable wavelet function is obtained at each scale.

[0105] S9: Output the final wavelet selection result. After all J decomposition levels have been processed, an optimal wavelet set Ψ={Ψ1,Ψ2,...,ΨJ} will be output. This set fully describes the optimal wavelet configuration at each decomposition level for a specific discharge breakdown ultrasonic signal, providing the best wavelet basis function selection scheme for subsequent wavelet threshold denoising processing.

[0106] S10: Using the pre-selected optimal wavelet set Ψ as the decomposition tool, hierarchical wavelet decomposition is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients corresponding to each level. Based on the preset threshold rules and threshold functions, the detail coefficients of each level are processed: coefficients with amplitudes below the threshold are considered noise and are either contracted (soft threshold) or zeroed (hard threshold); the approximation coefficients remain unchanged or are only slightly adjusted, finally obtaining the processed approximation coefficients and processed detail coefficients of each level. The inverse wavelet transform reconstructs the denoised signal. Taking all the processed approximation coefficients and detail coefficients as input, the discrete wavelet inverse transform is performed to output the denoised discharge breakdown ultrasonic signal.

[0107] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0108] Based on the same inventive concept, this application also provides a noise reduction device for discharge breakdown ultrasonic signals to implement the noise reduction method for discharge breakdown ultrasonic signals described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the noise reduction device for discharge breakdown ultrasonic signals provided below can be found in the limitations of the noise reduction method for discharge breakdown ultrasonic signals above, and will not be repeated here.

[0109] In one exemplary embodiment, such as Figure 4 As shown, a denoising device for discharge breakdown ultrasonic signals is provided, comprising: an analysis module 402, a maximum decomposition layer determination module 404, a decomposition module 406, and a denoising and reconstruction module 408, wherein:

[0110] Analysis module 402 is used to obtain the energy spectral density curve characterizing the relationship between pulse energy and frequency based on the discharge breakdown ultrasonic signal;

[0111] The maximum decomposition layer determination module 404 is used to determine the maximum decomposition layer when performing wavelet decomposition on the discharge breakdown ultrasonic signal according to the actual needs of high-voltage equipment insulation monitoring.

[0112] The decomposition module 406 is used to decompose the discharge breakdown ultrasonic signal as the current input signal according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level.

[0113] The denoising and reconstruction module 408 is used to construct an optimal wavelet set based on the optimal wavelets of each decomposition level; based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level; wavelet reconstruction is performed based on the approximation coefficients and detail coefficients after processing at each level to obtain the denoised discharge breakdown ultrasonic signal.

[0114] In one embodiment, the analysis module 402 is configured to:

[0115] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0116] In one embodiment, the maximum decomposition layer determination module 404 is used for:

[0117] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0118] In one embodiment, the decomposition module 406 is configured to:

[0119] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0120] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0121] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0122] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0123] In one embodiment, the denoising and reconstruction module 408 is used for:

[0124] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0125] Each module in the aforementioned noise reduction device for discharge breakdown ultrasonic signals can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0126] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When executed by the processor, the computer program implements a method for denoising ultrasonic signals caused by discharge breakdown. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0127] Those skilled in the art will understand that Figure 5 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 to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0128] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0129] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0130] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0131] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0132] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0133] In one embodiment, based on the discharge breakdown ultrasonic signal, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained, including:

[0134] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0135] In one embodiment, based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers for wavelet decomposition of the discharge breakdown ultrasonic signal is determined, including:

[0136] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0137] In one embodiment, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level, including:

[0138] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0139] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0140] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0141] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0142] In one embodiment, wavelet thresholding denoising is performed on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain approximation coefficients and detail coefficients after each level of processing, including:

[0143] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0144] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0145] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0146] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0147] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0148] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0149] In one embodiment, based on the discharge breakdown ultrasonic signal, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained, including:

[0150] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0151] In one embodiment, based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers for wavelet decomposition of the discharge breakdown ultrasonic signal is determined, including:

[0152] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0153] In one embodiment, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level, including:

[0154] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0155] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0156] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0157] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0158] In one embodiment, wavelet thresholding denoising is performed on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain approximation coefficients and detail coefficients after each level of processing, including:

[0159] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0160] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0161] Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained;

[0162] Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal.

[0163] Using the discharge breakdown ultrasonic signal as the current input signal, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level;

[0164] Based on the optimal wavelets at each decomposition level, an optimal wavelet set is constructed. Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level. Based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

[0165] In one embodiment, based on the discharge breakdown ultrasonic signal, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained, including:

[0166] Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse. The energy spectral density is used to describe the distribution of pulse energy at different frequency components. Based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

[0167] In one embodiment, based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers for wavelet decomposition of the discharge breakdown ultrasonic signal is determined, including:

[0168] Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

[0169] In one embodiment, the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level, including:

[0170] For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal.

[0171] Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band.

[0172] Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

[0173] In one embodiment, the wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; wherein the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

[0174] In one embodiment, wavelet thresholding denoising is performed on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain approximation coefficients and detail coefficients after each level of processing, including:

[0175] Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, while the coefficients corresponding to the effective signal are retained, thus obtaining the approximate coefficients and detail coefficients after processing at each level.

[0176] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0177] Those skilled in the art will understand 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 can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0178] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0179] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for denoising ultrasonic signals caused by discharge breakdown, characterized in that, The method includes: Based on the discharge breakdown ultrasonic signal, the energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained; Based on the actual needs of high-voltage equipment insulation monitoring, the maximum number of decomposition layers is determined when performing wavelet decomposition on the discharge breakdown ultrasonic signal. The discharge breakdown ultrasonic signal is used as the current input signal, and the current input signal is decomposed according to the maximum decomposition level to obtain the optimal wavelet of each decomposition level. Based on the optimal wavelets of each decomposition level, an optimal wavelet set is constructed; based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain the approximation coefficients and detail coefficients after processing at each level; based on the approximation coefficients and detail coefficients after processing at each level, wavelet reconstruction is performed to obtain the denoised discharge breakdown ultrasonic signal.

2. The method according to claim 1, characterized in that, The method of obtaining the energy spectral density curve characterizing the relationship between pulse energy and frequency based on the discharge breakdown ultrasonic signal includes: Frequency domain analysis is performed on the reference pulse of the discharge breakdown ultrasonic signal to obtain the energy spectral density of the reference pulse, which is used to describe the distribution of pulse energy at different frequency components; based on the energy spectral density, an energy spectral density curve characterizing the relationship between pulse energy and frequency is obtained.

3. The method according to claim 1, characterized in that, The determination of the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the actual needs of high-voltage equipment insulation monitoring includes: Determine the energy retention threshold; accumulate energy from the lowest frequency end to the high frequency end along the energy spectral density curve; when the accumulated energy first reaches or exceeds the energy retention threshold, record the corresponding frequency point as the key frequency; calculate the maximum number of decomposition layers when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the signal sampling frequency and the key frequency.

4. The method according to claim 1, characterized in that, The step of decomposing the current input signal according to the maximum decomposition level to obtain the optimal wavelet for each decomposition level includes: For the current level, each candidate wavelet function in the wavelet function library is traversed. The current input signal is decomposed into a single-level discrete wavelet function by the traversed candidate wavelet functions to obtain the approximate coefficients that reflect the low-frequency trend of the signal and the detail coefficients that reflect the high-frequency details of the signal. Calculate the maximum absolute value of the approximation coefficient and the detail coefficient respectively. If the maximum absolute value of the detail coefficient is not less than the maximum absolute value of the approximation coefficient, the detail sub-band is determined to be the signal band and the approximation sub-band is determined to be the noise band. If the maximum absolute value of the detail coefficient is less than the maximum absolute value of the approximation coefficient, the approximation sub-band is determined to be the signal band and the detail sub-band is determined to be the noise band. Calculate the peak signal-to-noise ratio (PSNR) based on the absolute maximum values ​​of the signal band coefficients and the noise band coefficients; compare the corresponding PSNR of all candidate wavelet functions at the current level, select the candidate wavelet function with the largest PSNR as the optimal wavelet for the current decomposition level, and use the approximation coefficients of the current level as the current input signal for the next level.

5. The method according to claim 4, characterized in that, The wavelet function library includes the Dobesi wavelet family, the Symlets wavelet family, and the Coiflets wavelet family; among them, the Dobesi wavelet family covers wavelets of orders 1-40, the Symlets wavelet family covers wavelets of orders 2-15, and the Coiflets wavelet family covers wavelets of orders 1-5.

6. The method according to claim 1, characterized in that, Based on the optimal wavelet set, wavelet threshold denoising is performed on the discharge breakdown ultrasonic signal to obtain approximation coefficients and detail coefficients after each level of processing, including: Based on the optimal wavelet set, the detail coefficients of each level are processed according to the preset threshold rules and threshold functions. Coefficients with amplitudes below the threshold are regarded as noise shrinkage or zeroed out, and the coefficients corresponding to the effective signals are retained to obtain the approximate coefficients and detail coefficients after processing at each level.

7. A noise reduction device for ultrasonic signals caused by discharge breakdown, characterized in that, The device includes: The analysis module is used to obtain the energy spectral density curve characterizing the relationship between pulse energy and frequency based on the discharge breakdown ultrasonic signal; The maximum decomposition layer determination module is used to determine the maximum decomposition layer when performing wavelet decomposition on the discharge breakdown ultrasonic signal based on the actual needs of high-voltage equipment insulation monitoring. The decomposition module is used to take the discharge breakdown ultrasonic signal as the current input signal, decompose the current input signal according to the maximum decomposition level, and obtain the optimal wavelet of each decomposition level. The denoising and reconstruction module is used to construct an optimal wavelet set based on the optimal wavelets of each decomposition level; perform wavelet threshold denoising on the discharge breakdown ultrasonic signal based on the optimal wavelet set to obtain the approximation coefficients and detail coefficients after processing at each level; and perform wavelet reconstruction based on the approximation coefficients and detail coefficients after processing at each level to obtain the denoised discharge breakdown ultrasonic signal.

8. A computer 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 method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images