Partial discharge sound signal de-noising processing method of power equipment and electronic equipment

Abstract

The invention discloses a partial discharge sound signal denoising processing method of power equipment and electronic equipment. The method comprises the following steps: acquiring original partial discharge sound signals corresponding to a plurality of channels in a microphone array; based on the original partial discharge sound signals corresponding to the plurality of channels, obtaining initial complex wavelet coefficients corresponding to the plurality of channels; optimizing the initial complex wavelet coefficients respectively corresponding to the plurality of channels to obtain optimized complex wavelet coefficients respectively corresponding to the plurality of channels; performing denoising processing on the optimized complex wavelet coefficients corresponding to the plurality of channels to obtain target complex wavelet coefficients corresponding to the plurality of channels; and obtaining a target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to the plurality of channels. According to the method and the device, the technical problem of insufficient denoising precision of the partial discharge sound signal caused in a complex noise environment is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of smart grid, in particular to a partial discharge sound signal denoising processing method of power equipment and electronic equipment. BACKGROUND

[0002] In modern power systems, power equipment such as transformers, gas insulated switchgear, power cables, etc. play a crucial role. Their stability and reliability directly affect the safe and efficient operation of the entire power grid. Partial discharge, as an early warning signal, can reflect the degree of deterioration of the insulation condition of the equipment. Although the partial discharge sound signal is weak and short-lived, its continuous existence often indicates a potential risk of insulation failure, which may lead to equipment damage and even power grid failure, even large-scale power outages. Therefore, real-time monitoring and accurate identification of partial discharge have become one of the key technologies to prevent major accidents in power systems and ensure the long-term stable operation of power equipment. The related technology has obvious deficiencies in the denoising processing of partial discharge sound signals, mainly in the following aspects:

[0003] The time-frequency analysis method, such as short-time Fourier transform, used in related technology has inherent time-frequency resolution limitations that limit its ability to capture the transient characteristics of partial discharge sound signals, especially when dealing with complex characteristics of "narrow pulse + wide frequency band", which can easily lead to loss of signal details or misjudgment of noise as signal; secondly, the fusion method of multi-channel signals in related technology relies too much on artificially designed rules, such as energy weighting and phase alignment, which lacks robustness when faced with complex noise in the substation environment, such as the mixture of mechanical vibration and electromagnetic interference, making it difficult to ensure the purity of the signal; in addition, the generation of denoising mask in related technology highly depends on human experience, which limits its generalization ability in different scenarios. Although the related partial discharge sound signal denoising method can provide a certain degree of signal purification within a certain range, when faced with complex noise environment, due to the failure to effectively solve the problems of adaptive capture of signal characteristics and intelligent suppression of noise, the denoising precision of partial discharge sound signal is insufficient.

[0004] At present, there is no effective solution to the above problems. SUMMARY

[0005] The embodiments of the present application provide a partial discharge sound signal denoising processing method of power equipment and electronic equipment to at least solve the technical problem of insufficient denoising precision of partial discharge sound signal caused by complex noise environment.

[0006] According to one aspect of the present invention, a method for denoising partial discharge sound signals of power equipment is provided, comprising: acquiring original partial discharge sound signals corresponding to multiple channels in a microphone array, wherein the original partial discharge sound signals represent partial discharge sound signals containing noise interference acquired from the power equipment; obtaining initial complex wavelet coefficients corresponding to multiple channels based on the original partial discharge sound signals corresponding to multiple channels, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals; optimizing the initial complex wavelet coefficients corresponding to multiple channels to obtain optimized complex wavelet coefficients corresponding to multiple channels, wherein the optimized complex wavelet coefficients represent complex wavelet coefficients obtained after optimizing the partial discharge features in the initial complex wavelet coefficients; denoising the optimized complex wavelet coefficients corresponding to multiple channels to obtain target complex wavelet coefficients corresponding to multiple channels, wherein the target complex wavelet coefficients represent complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge features from the optimized complex wavelet coefficients; and obtaining a target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to multiple channels.

[0007] According to another aspect of the present invention, a partial discharge sound signal denoising processing device for power equipment is also provided, comprising: an original partial discharge sound signal acquisition module, configured to acquire original partial discharge sound signals corresponding to each of multiple channels in a microphone array, wherein the original partial discharge sound signals represent partial discharge sound signals containing noise interference acquired from the power equipment; an initial complex wavelet coefficient acquisition module, configured to obtain initial complex wavelet coefficients corresponding to each of the multiple channels based on the original partial discharge sound signals corresponding to each of the multiple channels, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals; and an optimized complex wavelet coefficient acquisition module, configured to optimize the multiple channels... The initial complex wavelet coefficients corresponding to each channel are optimized to obtain optimized complex wavelet coefficients for each channel. The optimized complex wavelet coefficients represent the complex wavelet coefficients obtained after optimizing the partial discharge characteristics in the initial complex wavelet coefficients. The target complex wavelet coefficient acquisition module is used to perform noise denoising processing on the optimized complex wavelet coefficients corresponding to each channel to obtain target complex wavelet coefficients for each channel. The target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge characteristics from the optimized complex wavelet coefficients. The target partial discharge sound signal determination module is used to obtain the target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to each channel.

[0008] According to another aspect of the present invention, a non-volatile storage medium is also provided, the non-volatile storage medium storing a plurality of instructions, the instructions being adapted to be loaded by a processor and executed by any one of the partial discharge sound signal denoising processing methods for power equipment.

[0009] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the partial discharge sound signal denoising processing method for any of the power devices described in the present invention.

[0010] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the partial discharge sound signal denoising processing method for any of the power equipment described in the present invention.

[0011] In this embodiment of the invention, the original partial discharge sound signals corresponding to multiple channels in a microphone array are acquired. These original partial discharge sound signals represent partial discharge sound signals containing noise interference acquired from power equipment. Based on these original partial discharge sound signals, initial complex wavelet coefficients are obtained for each channel. These initial complex wavelet coefficients characterize the time-frequency domain features of the corresponding partial discharge sound signals. The initial complex wavelet coefficients for each channel are then optimized to obtain optimized complex wavelet coefficients for each channel. These optimized complex wavelet coefficients represent complex wavelet coefficients obtained after optimizing the partial discharge features in the initial complex wavelet coefficients. The coefficients are denoised to obtain target complex wavelet coefficients for each of the multiple channels. The target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge characteristics from the optimized complex wavelet coefficients. Based on the target complex wavelet coefficients for each of the multiple channels, the target partial discharge sound signal of the power equipment is obtained. This achieves the goal of determining the target partial discharge sound signal by acquiring the original partial discharge sound signal and optimizing, denoising, and recombining the multiple initial complex wavelet coefficients obtained from the decomposition. This achieves the technical effect of improving the denoising accuracy of the partial discharge sound signal in complex noise environments, and solves the technical problem of insufficient denoising accuracy of the partial discharge sound signal in complex noise environments. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0013] Figure 1 This is a flowchart of a method for denoising the partial discharge sound signal of power equipment according to an embodiment of the present invention;

[0014] Figure 2 This is a flowchart of an optional method for denoising the partial discharge sound signal of a power device according to an embodiment of the present invention;

[0015] Figure 3 This is a schematic diagram of a partial discharge sound signal denoising processing device for power equipment according to an embodiment of the present invention. Detailed Implementation

[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] According to an embodiment of the present invention, a method for denoising partial discharge sound signals of power equipment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0019] Figure 1 This is a flowchart of a method for denoising the partial discharge sound signal of power equipment according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0020] Step S102: Obtain the original partial discharge sound signals corresponding to each of the multiple channels in the microphone array, wherein the original partial discharge sound signals represent the partial discharge sound signals containing noise interference collected from the power equipment.

[0021] Optionally, by deploying multiple microphones around the power equipment, the partial discharge sound signal emitted by the equipment, along with the accompanying environmental noise, can be captured simultaneously. Multiple channels refer to each microphone in the microphone array independently acquiring signals, forming different signal channels. The raw partial discharge sound signal is the sound signal directly acquired from the power equipment, containing characteristic signals generated by partial discharge as well as various environmental noises (including but not limited to wind noise, mechanical vibration noise, human voice, and electromagnetic interference). Partial discharge activity in power equipment is usually very weak, with complex signal patterns and short durations; therefore, the raw signal often contains a high amount of noise, with signal and noise coexisting, making it difficult to directly analyze and diagnose the specific situation of the partial discharge.

[0022] Step S104: Based on the original partial discharge sound signals corresponding to each of the multiple channels, the initial complex wavelet coefficients corresponding to each of the multiple channels are obtained. The initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals.

[0023] Optionally, in the context of processing partial discharge sound signals, a dual-tree complex wavelet transform method can be used to extract the signal features corresponding to each of the multiple channels. The initial complex wavelet coefficients obtained by this method not only reflect the intensity changes of the partial discharge sound signal at different time scales, but also carry details about the phase of the partial discharge sound signal, which is particularly important for analyzing transient signals with complex frequency components, such as partial discharge.

[0024] In one optional embodiment, the initial complex wavelet coefficients for each of the multiple channels are obtained based on the original partial discharge sound signals corresponding to each channel, including: performing bandpass filtering on the original partial discharge sound signal of any channel to obtain the filtered partial discharge sound signal corresponding to any channel; normalizing the filtered partial discharge sound signal to obtain the normalized partial discharge sound signal corresponding to any channel; obtaining the initial complex wavelet coefficients for any channel based on the normalized partial discharge sound signal; and obtaining the initial complex wavelet coefficients for each of the multiple channels based on the original partial discharge sound signals corresponding to each channel, using the method of obtaining the initial complex wavelet coefficients for any channel.

[0025] Optionally, the purpose of bandpass filtering the raw partial discharge sound signal for each channel is to remove frequency components unrelated to partial discharge, including but not limited to low-frequency mechanical vibration noise and high-frequency quantization noise. The filtered partial discharge sound signal for any channel can be obtained as follows: ,in, This represents the instantaneous value of the filtered partial discharge sound signal at any sampling point. This represents the instantaneous value of the original partial discharge sound signal of any channel at any sampling point, where i represents the index of any channel. This represents any filter coefficient at any sampling point. , This represents any filter coefficient at the next sampling point after any sampling point. This represents the learning rate of the filter. Let e[n] represent the input partial discharge sound signal at any sampling point, e[n] represent the error partial discharge sound signal at any sampling point, k represent the index of any filter coefficient, L represent the order of the filter, and n represent the index of any sampling point. Next, the partial discharge sound signal after bandpass filtering needs to be normalized to facilitate the comparison and fusion of different partial discharge sound signals in subsequent processing. Normalization using the median absolute deviation can standardize the intensity of the partial discharge sound signal while maintaining the time-frequency characteristics of the signal. The normalized partial discharge sound signal corresponding to any channel is obtained as follows: ,in, This represents the instantaneous value of the normalized partial discharge sound signal of any channel at any sampling point. This represents the absolute deviation of the median of the instantaneous value of the filtered partial discharge sound signal at any sampling point. This represents the median of the instantaneous value of the filtered partial discharge sound signal at any sampling point. Finally, the normalized partial discharge sound signal is subjected to a dual-tree complex wavelet transform to obtain the initial complex wavelet coefficients for each channel. This process is repeated for all channels in the microphone array to obtain the initial complex wavelet coefficients for each channel. The initial complex wavelet coefficients obtained through this method not only enhance the understanding of the partial discharge sound signal but also provide a solid foundation for subsequent partial discharge sound signal enhancement and feature extraction.

[0026] In one optional embodiment, when there are multiple initial complex wavelet coefficients for any channel, the initial complex wavelet coefficients for any channel are obtained based on the normalized partial discharge sound signal, including: performing frame processing on the normalized partial discharge sound signal according to a predetermined sampling duration to obtain multiple frame signals for any channel; using a dual-tree complex wavelet transform method to decompose the multiple frame signals to obtain the initial complex wavelet coefficients corresponding to each of the multiple frame signals; and obtaining the initial complex wavelet coefficients for any channel based on the initial complex wavelet coefficients corresponding to each of the multiple frame signals.

[0027] Optionally, the normalized partial discharge sound signal can be framed according to a predetermined sampling duration. This framing facilitates subsequent time-frequency domain analysis because dual-tree complex wavelet transform performs better on signal segments of fixed length, avoiding boundary effects and improving computational efficiency. Any frame signal from any channel can be obtained as follows: ,in, This represents the instantaneous value of any frame signal from any channel at any sampling point. Represents the index of any sampling point. The number of sampling points is represented by t, the index of any frame signal is represented by t, the number of frame signals is represented by G, and M represents the frame shift between the starting sampling points of two adjacent frame signals. Next, a dual-tree complex wavelet transform is used to decompose each frame signal, generating complex wavelet coefficients at multiple time scales. The time-frequency analysis capability of the complex wavelet coefficients helps distinguish partial discharge sound signals from environmental noise.

[0028] In an optional embodiment, when any frame signal is decomposed into multiple scales, and each scale corresponds one-to-one with multiple initial complex wavelet coefficients, a dual-tree complex wavelet transform method is used to decompose the multiple frame signals of any channel to obtain the initial complex wavelet coefficients corresponding to each of the multiple frame signals. This includes: using a dual-tree complex wavelet transform method to decompose any frame signal of any channel in the following manner to obtain any initial complex wavelet coefficient of any frame signal:

[0029] ;

[0030] in, Denotes any initial complex wavelet coefficients. Let represent the first wavelet coefficient at any scale, where any scale is the scale corresponding to any initial complex wavelet coefficient. This represents the function value at any sampling point. This represents the function value of the first wavelet basis function at any sampling point for any scale. This represents the translation of the first wavelet basis function in the time domain. This represents the second real wavelet coefficients at any scale. This represents the function value of the second wavelet basis function at any sampling point for any scale. The second wavelet basis function is represented by its translation in the time domain, where i represents the imaginary unit, j represents the index of any scale, k represents the index of the position at any scale, n represents the index of any sampling point, and N represents the number of sampling points. Multiple initial complex wavelet coefficients corresponding to each of the multiple frame signals are obtained by using any initial complex wavelet coefficients.

[0031] Optionally, dual-tree complex wavelet transform is a two-dimensional wavelet transform technique that decomposes the signal using two wavelet basis functions that are Hilbert transforms of each other. This decomposes any frame of the signal into complex wavelet coefficients at different scales, each corresponding to a local characteristic of the signal at a specific time scale. Since the energy distribution and frequency characteristics of partial discharge sound signals are often highly localized, this enhancement of time-frequency characteristics is crucial for subsequent signal recognition and classification. The frequency band corresponding to any scale is... ,in, The sampling frequency is represented by , and J represents the number of scales. By decomposing the partial discharge sound signal into different scales, the wavelet transform at each scale can focus on a specific frequency band of the partial discharge sound signal, thereby achieving comprehensive frequency analysis of the partial discharge sound signal. The frequency band division at different scales also helps to more effectively separate the partial discharge sound signal from noise in the time-frequency domain.

[0032] Step S106: Optimize the initial complex wavelet coefficients corresponding to each of the multiple channels to obtain the optimized complex wavelet coefficients corresponding to each of the multiple channels. The optimized complex wavelet coefficients represent the complex wavelet coefficients obtained after optimizing the partial discharge characteristics in the initial complex wavelet coefficients.

[0033] Optionally, through dual-tree complex wavelet transform, the partial discharge sound signal is decomposed into a series of scale- and location-dependent complex coefficients, which contain amplitude and phase information of the signal. However, these coefficients not only contain the actual signal components of the partial discharge but may also be mixed with environmental noise and other interference factors. By processing the initial complex wavelet coefficients with specific strategies, the complex coefficients after highlighting the partial discharge sound signal and reducing the influence of noise can be achieved. The optimization goal is to enhance the characteristics of the partial discharge sound signal, making the partial discharge activity more obvious, while minimizing noise components to ensure signal purity and the integrity of the partial discharge characteristics.

[0034] In an optional embodiment, where any channel includes multiple frame signals, any frame signal is decomposed into multiple scales, and multiple scales correspond one-to-one with multiple optimized complex wavelet coefficients, the initial complex wavelet coefficients corresponding to each of the multiple channels are optimized to obtain the optimized complex wavelet coefficients corresponding to each of the multiple channels. This includes optimizing any initial complex wavelet coefficient to obtain any optimized wavelet coefficient by: obtaining the frequency domain attention weight of any initial complex wavelet coefficient, wherein the frequency domain attention weight is used to quantify the importance of the target frequency domain energy of the corresponding initial complex wavelet coefficient at any frequency point relative to the total target frequency domain energy at multiple frequency points. Frequency domain energy represents the frequency domain energy associated with the partial discharge characteristics of power equipment, selected from the initial frequency domain energy. The initial frequency domain energy represents the energy distribution of the corresponding initial complex wavelet coefficients in the frequency domain. Scale attention weights at any scale are obtained, where any scale represents the scale corresponding to the initial complex wavelet coefficients. The scale attention weights are used to quantify the importance of the amplitude energy at the corresponding scale compared to the total amplitude energy across multiple scales. Based on any initial complex wavelet coefficient, the scale attention weights, and the frequency domain attention weights, any optimized complex wavelet coefficient is obtained. Using the method for obtaining any optimized complex wavelet coefficient, optimized complex wavelet coefficients corresponding to each of the multiple channels are obtained.

[0035] Optionally, firstly, for any initial complex wavelet coefficient, its frequency domain attention weight is calculated. This can identify and highlight the characteristic frequencies of the partial discharge sound signal, accurately focusing on the key frequency bands of the signal even in complex noise environments. The target frequency domain energy refers to the frequency domain energy highly correlated with the partial discharge characteristics, while the initial frequency domain energy refers to the entire energy distribution of the coefficient in the frequency domain. The calculation of the frequency domain attention weight is based on an in-depth analysis of the spectral characteristics of the partial discharge sound signal. By identifying and highlighting the frequency bands most relevant to the characteristics of the partial discharge sound signal, the ability to identify partial discharge activity in the signal can be significantly enhanced. Next, for any scale, its scale attention weight is calculated. The scale attention weight reflects the importance of the amplitude energy of the corresponding scale relative to the total amplitude energy of all scales. The amplitude energy of a scale can be obtained by calculating the sum of squares of the amplitudes of all complex wavelet coefficients at that scale. The introduction of the scale attention weight allows for adaptive adjustment of the contribution of each scale in the final signal reconstruction. For example, scales that are more prominent in the characteristic frequency range of the partial discharge sound signal will be assigned higher weights. Furthermore, for the initial complex wavelet coefficients of the same scale in each channel, multiple amplitude covariance matrices can be determined. Each element of any amplitude covariance matrix is ​​obtained by calculating the covariance of the amplitudes of the initial complex wavelet coefficients at the same scale in multiple frame signals corresponding to any two channels. Any element of any amplitude covariance matrix can be obtained as follows: ,in, Let represent any element of any magnitude covariance matrix. These represent the indices of any channel. Indicates the number of channels. and These represent the amplitudes of any initial complex wavelet coefficients corresponding to the same scale. For elements in the amplitude covariance matrix whose covariance values ​​are greater than a predetermined threshold, it indicates that the corresponding initial complex wavelet coefficients have strong consistency with the partial discharge sound signal, and can be used as the focus of subsequent optimization. Finally, by multiplying any initial complex wavelet coefficient by the corresponding frequency domain attention weight and scale attention weight, the optimized complex wavelet coefficients (optimized complex wavelet coefficients) can be obtained. This optimization not only considers the characteristics of the partial discharge sound signal in the frequency domain, but also combines the performance of the partial discharge sound signal at different time scales to achieve a comprehensive enhancement of the characteristics of the partial discharge sound signal. Any optimized complex wavelet coefficient can be obtained as follows: first, multiply any initial complex wavelet coefficient W... DT (j,k) and scale attention weights Multiply to obtain the scale-enhancing coefficient. Next, regarding Performing a Fast Fourier Transform yields the frequency domain representation F of the scale enhancement coefficients. 1 (j,f); further, F 1 (j,f) and frequency domain attention weights Perform element-wise multiplication to obtain the optimized frequency domain signal F. 2 (j,f)=F 1 (j,f)· Finally, regarding F 2 Perform an inverse fast Fourier transform on (j,f) to obtain the optimized complex wavelet coefficients. The above operation is repeated on all channels to obtain the optimized complex wavelet coefficients for each of the multiple channels.

[0036] In one optional embodiment, obtaining the frequency domain attention weight of any initial complex wavelet coefficient includes: obtaining the initial frequency domain energy of any initial complex wavelet coefficient by using the fast Fourier transform method; performing bandpass filtering on the initial frequency domain energy to obtain the target frequency domain energy; and normalizing the target frequency domain energy to obtain the frequency domain attention weight.

[0037] Optionally, by obtaining the frequency domain attention weights of any initial complex wavelet coefficients, the frequency bands related to the partial discharge characteristics in the partial discharge sound signal can be highlighted, while suppressing the influence of irrelevant noise frequency bands. First, by using a Fast Fourier Transform, any initial complex wavelet coefficient is transformed to the frequency domain, obtaining the energy distribution of any initial complex wavelet coefficient in the frequency domain, i.e., the initial frequency domain energy. The initial frequency domain energy of any initial complex wavelet coefficient can be obtained as follows: E(j,f)=|F(j,f)| 2 Where E(j,f) represents the initial frequency domain energy, and F(j,f) = FFT(W DT (j,k)) represents the frequency domain distribution of any initial complex wavelet coefficients. Next, bandpass filtering is applied to the initial frequency domain energy to obtain the target frequency domain energy, retaining the frequency band energy related to the characteristic frequency range of the partial discharge sound signal while filtering out energy from other frequency bands. Finally, the target frequency domain energy is normalized to obtain the frequency domain attention weights. Normalization ensures that the frequency domain attention weights are between 0 and 1, allowing them to be directly used in subsequent signal enhancement or denoising operations. A larger weight value indicates a higher importance of the corresponding frequency point in the signal. The frequency domain attention weights can be obtained as follows: = ,in, Represents the frequency domain attention weights. This represents the target frequency domain energy at any frequency point. This represents the total target frequency domain energy at multiple frequency points. Indicates the index of any frequency point.

[0038] In one optional embodiment, obtaining the scale attention weight at any scale includes: determining the phase deviation corresponding to each of multiple initial complex wavelet coefficients at any scale; filtering the multiple initial complex wavelet coefficients under a preset condition to obtain filtered complex wavelet coefficients, wherein the preset condition indicates that the phase deviation of any initial complex wavelet coefficient is less than a preset phase deviation threshold, and the filtered complex wavelet coefficients represent complex wavelet coefficients that are filtered from multiple initial complex wavelet coefficients and are associated with the partial discharge characteristics of power equipment; determining the amplitude energy at any scale based on the filtered complex wavelet coefficients; obtaining the amplitude energy corresponding to each of the multiple scales using the method of obtaining the amplitude energy at any scale; and determining the scale attention weight at any scale based on the amplitude energy corresponding to each of the multiple scales.

[0039] Optionally, by obtaining scale attention weights at any scale, the scale most closely associated with partial discharge features in the partial discharge sound signal can be quantified to facilitate subsequent signal optimization. First, the phase deviations corresponding to multiple initial complex wavelet coefficients are calculated at any scale. This is done by comparing the phases of adjacent complex wavelet coefficients at the same scale to evaluate the phase stability of the signal at that scale. The phase deviation of any initial complex wavelet coefficient can be obtained as follows: ,in, This represents the phase deviation of any initial complex wavelet coefficient. This represents the phase of any initial complex wavelet coefficient. This represents the phase of the previous initial complex wavelet coefficient for any given initial complex wavelet coefficient. Then, for each initial complex wavelet coefficient, if its phase deviation is less than a preset phase deviation threshold... The initial complex wavelet coefficients are determined to be those dominated by the partial discharge signal and are retained for subsequent amplitude energy calculations; otherwise, they are considered noise-dominated coefficients. Further, based on the filtered complex wavelet coefficients, the total amplitude energy of these coefficients at the corresponding scale is statistically analyzed. The amplitude energy at any scale can be obtained as follows: ,in, Let K(j) represent the amplitude energy at any scale, and let K(j) represent the set of initial complex wavelet coefficients that satisfy a phase deviation less than a preset phase deviation threshold. Repeat the above process to calculate the amplitude energy for each scale, forming a complete scale energy distribution. Finally, compare the amplitude energy of each scale with the sum of the amplitude energies of all scales, and obtain the attention weight for each scale through normalization. A larger scale attention weight indicates that it is more critical in the signal feature optimization process. The scale attention weight for any scale can be obtained as follows: = ,in, Represents the scale attention weights at any scale. (j)= J represents the scale attention weight at any unnormalized scale, and J represents the number of scales.

[0040] Step S108: Denoise the optimized complex wavelet coefficients corresponding to each of the multiple channels to obtain the target complex wavelet coefficients corresponding to each of the multiple channels. The target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference that is unrelated to the partial discharge characteristics from the optimized complex wavelet coefficients.

[0041] Optionally, in the partial discharge sound signal processing, denoising the optimized complex wavelet coefficients corresponding to each of the multiple channels to obtain the target complex wavelet coefficients is a key step in ensuring that the partial discharge features in the partial discharge sound signal are clearly distinguishable and reducing noise interference. The target complex wavelet coefficients obtained through denoising processing are a manifestation that further enhances the partial discharge features and reduces noise interference in the multiple channel signals.

[0042] In one optional embodiment, where any channel includes multiple frame signals, any frame signal is decomposed into multiple scales, and multiple scales correspond one-to-one with multiple target complex wavelet coefficients, the optimized complex wavelet coefficients corresponding to each of the multiple channels are denoised to obtain the target complex wavelet coefficients corresponding to each of the multiple channels. This includes: denoising any optimized complex wavelet coefficient to obtain any target complex wavelet coefficient in the following manner: determining the phase deviation of any optimized complex wavelet coefficient; masking any optimized complex wavelet coefficient based on a preset phase deviation and the phase deviation of any optimized complex wavelet coefficient to obtain any masked complex wavelet coefficient; using an adaptive threshold denoising method based on any masked complex wavelet coefficient to obtain any target complex wavelet coefficient; and obtaining the target complex wavelet coefficients corresponding to each of the multiple channels using the same method as obtaining any target complex wavelet coefficient.

[0043] Optionally, firstly, the phase deviation of any optimized complex wavelet coefficient can be determined using the same calculation method as for determining the phase deviation of any initial complex wavelet coefficient. Based on a preset phase deviation threshold, the optimized complex wavelet coefficients are masked. If the phase deviation of the optimized complex wavelet coefficient is less than the threshold, it is determined to be a component related to the characteristics of the partial discharge sound signal and is retained; if the phase deviation is greater than or equal to the threshold, it may be considered a noise component and is set to zero through masking. Any masked complex wavelet coefficient can be obtained as follows: ,in, Denotes the complex wavelet coefficients of any mask. This represents the phase deviation of any optimized complex wavelet coefficient. Next, an adaptive threshold denoising method is applied to any masked complex wavelet coefficient. Adaptive threshold denoising dynamically determines the threshold based on the statistical characteristics of the partial discharge sound signal, enabling more accurate identification and removal of noise while preserving the key features of the partial discharge sound signal. This method avoids the problems of over-denoising or noise residue that may occur with a fixed threshold. Finally, after masking and adaptive threshold denoising, the obtained target complex wavelet coefficients are purer and more focused on the characteristics of partial discharge. Any target complex wavelet coefficient can be obtained as follows: ,in, Denotes the complex wavelet coefficients of any target. = This represents a preset adaptive threshold for any scale. This represents the number of optimized complex wavelet coefficients at any scale. The optimized complex wavelet coefficients of each channel undergo the same denoising process, ensuring the consistency and comparability of all channel signals after denoising, laying the foundation for subsequent multi-channel target complex wavelet coefficient fusion and analysis.

[0044] Step S110: Based on the target complex wavelet coefficients corresponding to each of the multiple channels, the target partial discharge sound signal of the power equipment is obtained.

[0045] Optionally, the target partial discharge sound signal of the power equipment can be obtained through the target complex wavelet coefficients. This process enables high-quality conversion of the partial discharge sound signal from the time-frequency domain to the time domain, ensuring clear partial discharge characteristics and suppressing noise components while maintaining the continuity and integrity of the partial discharge sound signal. This denoised partial discharge sound signal (target partial discharge sound signal) is suitable for real-time monitoring and analysis of partial discharge in power equipment, improving the reliability and accuracy of partial discharge detection. This allows for the timely detection of potential equipment problems, which is of great value for preventing power equipment failures and ensuring the safe operation of the power grid.

[0046] In one optional embodiment, where any channel includes multiple frame signals, any frame signal is decomposed into multiple scales, and multiple scales correspond one-to-one with multiple target complex wavelet coefficients, the target partial discharge sound signal of the power equipment is obtained based on the target complex wavelet coefficients corresponding to each of the multiple channels. This includes: using a dual-tree complex wavelet inverse transform method to reassemble the multiple target complex wavelet coefficients in any frame signal to obtain the reassembled partial discharge sound signal of any frame signal; performing overlapping and splicing processing on the reassembled partial discharge sound signals corresponding to each of the multiple frame signals to obtain the reassembled partial discharge sound signal of any channel; and integrating the reassembled partial discharge sound signals corresponding to each of the multiple channels to obtain the target partial discharge sound signal.

[0047] Optionally, based on the target complex wavelet coefficients corresponding to multiple channels, the target partial discharge sound signal of the power equipment is obtained. This process aims to extract a clear partial discharge sound signal from a complex environment by integrating time-frequency domain analysis of the partial discharge sound signal with multi-channel information. First, the dual-tree complex wavelet inverse transform method is used to process multiple target complex wavelet coefficients within any frame of the signal, reconstructing them from the time-frequency domain back to the time domain to generate a single frame of the partial discharge sound signal. This inverse transform process is performed on those complex wavelet coefficients that remove noise and retain partial discharge characteristics; therefore, the reconstructed signal is purer and its characteristics are more prominent. The reconstructed partial discharge sound signal of any frame of the signal can be obtained as follows: ,in, This represents the partial discharge sound signal after recombination of any frame of signal. This represents the index of any sampling point in the complex wavelet coefficients of any target. This indicates the number of sampling points. Next, considering the frame shift overlap during partial discharge sound signal acquisition, it is necessary to perform overlapping splicing processing on the partial discharge sound signals between adjacent frames to smooth the signal transition between frames and eliminate signal redundancy or breakpoints caused by overlap. This process is usually accomplished by calculating the average value of the signal in the overlapping region, which ensures the continuity and integrity of the signal on the time axis. For example, for two consecutive frames, assuming the overlapping region of the first frame signal is […]. -M+1, The overlapping region of the second frame signal is [1, M]. The synthesized signal of the overlapping portion of two adjacent frames in any channel can be obtained as follows: ,in, The composite signal of the overlapping portion of two adjacent frames in any channel, where M represents the frame shift between the starting sampling points of the two adjacent frames. For non-overlapping regions, the original reconstructed partial discharge sound signal value is directly retained. Finally, the reconstructed partial discharge sound signals of all channels are integrated, and the spliced ​​signals of each channel are arranged according to their channel indices to obtain the final multi-channel clean partial discharge sound signal, i.e., the target partial discharge sound signal. After obtaining the target partial discharge sound signal, the denoising effect of the target partial discharge sound signal relative to the original partial discharge sound signal can be comprehensively evaluated by calculating three indicators: Signal-to-Noise Ratio (SNR), Root Mean Square Error (RMSE), and Peak Retention Rate (PRR). The SNR of the target partial discharge sound signal compared to the original partial discharge sound signal can be obtained as follows: SNR = Wherein, SNR represents the signal-to-noise ratio of the target partial discharge sound signal to the original partial discharge sound signal. This represents the total energy of the multi-channel raw partial discharge sound signal at any sampling point. This represents the total energy of the target partial discharge sound signal at any sampling point, where n represents the index of any sampling point. This indicates the number of sampling points. The root mean square error (RMSE) between the target partial discharge sound signal and the original partial discharge sound signal can be obtained as follows: RMSE = Where RMSE represents the root mean square error between the target partial discharge sound signal and the original partial discharge sound signal. The peak retention rate (PRR) between the target partial discharge sound signal and the original partial discharge sound signal can be obtained as follows: PRR = Wherein, PRR represents the peak retention rate between the target partial discharge sound signal and the original partial discharge sound signal. This represents the maximum value in the total energy of the multi-channel raw partial discharge sound signal from multiple sampling points. This represents the maximum value among the total energy of the target partial discharge sound signal from multiple sampling points. The signal-to-noise ratio (SNR) is used to quantify the noise suppression effect, the root mean square error (RMSE) is used to measure the accuracy of denoising, and the peak retention rate focuses on the preservation of partial discharge characteristics.

[0048] Through the above steps S102 to S110, the goal of optimizing, denoising, and recombining multiple initial complex wavelet coefficients obtained by decomposing the original partial discharge sound signal can be achieved to determine the target partial discharge sound signal. This achieves the technical effect of improving the denoising accuracy of the partial discharge sound signal in complex noise environments, thereby solving the technical problem of insufficient denoising accuracy of the partial discharge sound signal caused by complex noise environments.

[0049] Based on the above embodiments and optional embodiments, the present invention proposes an optional implementation method. Figure 2 This is a flowchart of an optional method for denoising the partial discharge sound signal of power equipment according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes:

[0050] S1: Obtain the original partial discharge sound signals corresponding to each of the multiple channels in the multi-channel microphone array. The specific implementation process is the same as in the aforementioned embodiment, and will not be repeated here.

[0051] S2: Bandpass filtering and normalization are performed on the original partial discharge sound signals corresponding to each of the multiple channels to obtain the normalized partial discharge sound signals corresponding to each of the multiple channels. The specific implementation process is the same as that in the previous embodiment, and will not be repeated here.

[0052] S3: The dual-tree complex wavelet transform method is used to decompose the normalized partial discharge sound signals corresponding to each of the multiple channels to obtain the initial complex wavelet coefficients corresponding to each of the multiple channels. The specific implementation process is the same as the previous embodiment, and will not be repeated here.

[0053] S4: The initial complex wavelet coefficients corresponding to each of the multiple channels are optimized by using scale attention weighting and frequency domain attention weighting methods to obtain the optimized complex wavelet coefficients corresponding to each of the multiple channels. The specific implementation process is the same as the previous embodiment, and will not be repeated here.

[0054] S5: Perform adaptive threshold denoising on the optimized complex wavelet coefficients corresponding to each of the multiple channels to obtain the target complex wavelet coefficients corresponding to each of the multiple channels. The specific implementation process is the same as the previous embodiment, and will not be repeated here.

[0055] S6: The dual-tree complex wavelet inverse transform method is used to reorganize the target complex wavelet coefficients corresponding to each of the multiple channels to obtain the reorganized partial discharge sound signals corresponding to each of the multiple channels. The specific implementation process is the same as the aforementioned embodiment, and will not be repeated here.

[0056] S7: Integrate the recombined partial discharge sound signals corresponding to each of the multiple channels to obtain the target partial discharge sound signal. The specific implementation process is the same as in the aforementioned embodiment, and will not be repeated here.

[0057] This embodiment can achieve at least one of the following effects: (1) Time-frequency optimization application of dual-tree complex wavelets: constructing dual-tree complex wavelets to solve the translation sensitivity of traditional wavelets; covering a wider frequency band through multi-layer decomposition, fully including the core frequency band of partial discharge. (2) Adaptive frequency domain attention mechanism: dynamically generating attention weights based on sub-band signal-to-noise ratio, strengthening the weights of sub-bands with significant partial discharge characteristics, and weakening the weights of noise-dominated sub-bands, replacing manual weighting rules. (3) Time-frequency domain joint adaptive threshold: combining the time-domain noise standard deviation and the frequency-domain noise power spectrum, constructing a dynamic threshold function, which improves both the noise suppression rate and the retention rate of partial discharge characteristics. (4) Full-dimensional quality assessment system: establishing a two-dimensional assessment index of "time domain + frequency domain" to ensure that the pre-denoised signal meets the needs of subsequent diagnosis.

[0058] This embodiment also provides a noise reduction device for partial discharge sound signals of power equipment. This device is used to implement the above embodiments and preferred embodiments, and will not be repeated for details already described. As used below, the terms "module" and "device" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0059] According to an embodiment of the present invention, an apparatus embodiment for implementing the above-described method for denoising partial discharge sound signals of power equipment is also provided. Figure 3 This is a schematic diagram of a partial discharge sound signal denoising processing device for power equipment according to an embodiment of the present invention. Figure 3 As shown, the partial discharge sound signal denoising processing device for the aforementioned power equipment includes: an original partial discharge sound signal acquisition module 300, an initial complex wavelet coefficient acquisition module 302, an optimized complex wavelet coefficient acquisition module 304, a target complex wavelet coefficient acquisition module 306, and a target partial discharge sound signal determination module 308, wherein:

[0060] The original partial discharge sound signal acquisition module 300 is used to acquire the original partial discharge sound signals corresponding to each of the multiple channels in the microphone array. The original partial discharge sound signal represents the partial discharge sound signal containing noise interference acquired from the power equipment.

[0061] The initial complex wavelet coefficient acquisition module 302 is connected to the original partial discharge sound signal acquisition module 300. It is used to obtain the initial complex wavelet coefficients corresponding to each of the multiple channels based on the original partial discharge sound signals corresponding to each channel. The initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals.

[0062] The optimized complex wavelet coefficient acquisition module 304 is connected to the initial complex wavelet coefficient acquisition module 302. It is used to optimize the initial complex wavelet coefficients corresponding to multiple channels to obtain the optimized complex wavelet coefficients corresponding to multiple channels. The optimized complex wavelet coefficients represent the complex wavelet coefficients obtained after optimizing the partial discharge characteristics in the initial complex wavelet coefficients.

[0063] The target complex wavelet coefficient acquisition module 306 is connected to the optimized complex wavelet coefficient acquisition module 304. It is used to perform noise reduction processing on the optimized complex wavelet coefficients corresponding to multiple channels to obtain the target complex wavelet coefficients corresponding to multiple channels. The target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference that is unrelated to the partial discharge characteristics from the optimized complex wavelet coefficients.

[0064] The target partial discharge sound signal determination module 308 is connected to the target complex wavelet coefficient acquisition module 306, and is used to obtain the target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to multiple channels.

[0065] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0066] It should be noted that the original partial discharge sound signal acquisition module 300, the initial complex wavelet coefficient acquisition module 302, the optimized complex wavelet coefficient acquisition module 304, the target complex wavelet coefficient acquisition module 306, and the target partial discharge sound signal determination module 308 mentioned above correspond to steps S102 to S110 in the embodiments. The examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in the above embodiments. It should be noted that the above modules, as part of the device, can run in a computer terminal.

[0067] It should be noted that the optional or preferred implementation methods of this embodiment can be found in the relevant descriptions in the embodiments, and will not be repeated here.

[0068] The aforementioned partial discharge sound signal denoising device for power equipment may further include a processor and a memory. The original partial discharge sound signal acquisition module 300, the initial complex wavelet coefficient acquisition module 302, the optimized complex wavelet coefficient acquisition module 304, the target complex wavelet coefficient acquisition module 306, and the target partial discharge sound signal determination module 308 are all stored in the memory as program modules. The processor executes the aforementioned program modules stored in the memory to realize the corresponding functions.

[0069] The processor contains a core that retrieves the corresponding program modules from memory. One or more cores may be configured. Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory includes at least one memory chip.

[0070] According to an embodiment of this application, an embodiment of a non-volatile storage medium is also provided. Optionally, in this embodiment, the non-volatile storage medium includes a stored program, wherein, when the program runs, it controls the device where the non-volatile storage medium is located to execute any of the aforementioned partial discharge sound signal denoising processing methods for power equipment.

[0071] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals, and the non-volatile storage medium includes stored programs.

[0072] Optionally, during program execution, the device containing the non-volatile storage medium is controlled to perform the following functions: acquire the original partial discharge sound signals corresponding to multiple channels in the microphone array, wherein the original partial discharge sound signals represent the partial discharge sound signals containing noise interference acquired from the power equipment; based on the original partial discharge sound signals corresponding to multiple channels, obtain the initial complex wavelet coefficients corresponding to multiple channels, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals; optimize the initial complex wavelet coefficients corresponding to multiple channels to obtain optimized complex wavelet coefficients corresponding to multiple channels, wherein the optimized complex wavelet coefficients represent the complex wavelet coefficients obtained after optimizing the partial discharge features in the initial complex wavelet coefficients; perform denoising processing on the optimized complex wavelet coefficients corresponding to multiple channels to obtain the target complex wavelet coefficients corresponding to multiple channels, wherein the target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge features from the optimized complex wavelet coefficients; and obtain the target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to multiple channels.

[0073] According to an embodiment of this application, an embodiment of a processor is also provided. Optionally, in this embodiment, the processor is used to run a program, wherein the program executes any of the above-described methods for denoising partial discharge sound signals from power equipment.

[0074] According to an embodiment of this application, an embodiment of a computer program product is also provided. Optionally, in this embodiment, the computer program product includes a computer program that, when executed by a processor, implements the steps of the partial discharge sound signal denoising processing method for any of the above-described power equipment.

[0075] Optionally, when the aforementioned computer program product is executed on a data processing device, it is suitable to execute an initialization program having the following method steps: acquiring the original partial discharge sound signals corresponding to multiple channels in the microphone array, wherein the original partial discharge sound signals represent partial discharge sound signals containing noise interference acquired from the power equipment; obtaining the initial complex wavelet coefficients corresponding to multiple channels based on the original partial discharge sound signals corresponding to multiple channels, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals; optimizing the initial complex wavelet coefficients corresponding to multiple channels to obtain optimized complex wavelet coefficients corresponding to multiple channels, wherein the optimized complex wavelet coefficients represent complex wavelet coefficients obtained after optimizing the partial discharge features in the initial complex wavelet coefficients; performing denoising processing on the optimized complex wavelet coefficients corresponding to multiple channels to obtain target complex wavelet coefficients corresponding to multiple channels, wherein the target complex wavelet coefficients represent complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge features from the optimized complex wavelet coefficients; and obtaining the target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to multiple channels.

[0076] This invention provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring original partial discharge sound signals corresponding to multiple channels in a microphone array, wherein the original partial discharge sound signals represent partial discharge sound signals containing noise interference acquired from a power device; obtaining initial complex wavelet coefficients corresponding to multiple channels based on the original partial discharge sound signals corresponding to each channel, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals; optimizing the initial complex wavelet coefficients corresponding to each channel to obtain optimized complex wavelet coefficients corresponding to each channel, wherein the optimized complex wavelet coefficients represent complex wavelet coefficients obtained after optimizing the partial discharge features in the initial complex wavelet coefficients; performing denoising processing on the optimized complex wavelet coefficients corresponding to each channel to obtain target complex wavelet coefficients corresponding to each channel, wherein the target complex wavelet coefficients represent complex wavelet coefficients obtained after removing noise interference unrelated to the partial discharge features from the optimized complex wavelet coefficients; and obtaining the target partial discharge sound signal of the power device based on the target complex wavelet coefficients corresponding to each channel.

[0077] The order of the above embodiments of the present invention is merely for description and does not represent the superiority or inferiority of the embodiments.

[0078] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0079] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of modules described above can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between modules, and may be electrical or other forms.

[0080] The modules described above as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0081] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0082] If the aforementioned integrated modules are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable non-volatile storage medium. Based on this understanding, the technical solution of this 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 non-volatile storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned non-volatile storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0083] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for denoising partial discharge sound signals of power equipment, characterized in that, include: Acquire the original partial discharge sound signals corresponding to each of the multiple channels in the microphone array, wherein the original partial discharge sound signals represent the partial discharge sound signals containing noise interference collected from the power equipment; Based on the original partial discharge sound signals corresponding to each of the multiple channels, the initial complex wavelet coefficients corresponding to each of the multiple channels are obtained, wherein the initial complex wavelet coefficients are used to characterize the time-frequency domain features of the corresponding partial discharge sound signals. The initial complex wavelet coefficients corresponding to each of the multiple channels are optimized to obtain the optimized complex wavelet coefficients corresponding to each of the multiple channels, wherein the optimized complex wavelet coefficients represent the complex wavelet coefficients obtained after optimizing the partial discharge characteristics in the initial complex wavelet coefficients; The optimized complex wavelet coefficients corresponding to each of the multiple channels are denoised to obtain the target complex wavelet coefficients corresponding to each of the multiple channels. The target complex wavelet coefficients represent the complex wavelet coefficients obtained after removing noise interference that is unrelated to the partial discharge characteristics from the optimized complex wavelet coefficients. Based on the target complex wavelet coefficients corresponding to each of the multiple channels, the target partial discharge sound signal of the power equipment is obtained.

2. The method according to claim 1, characterized in that, The process of obtaining the initial complex wavelet coefficients for each of the multiple channels based on their respective original partial discharge sound signals includes: The original partial discharge sound signal of any channel is subjected to bandpass filtering to obtain the filtered partial discharge sound signal corresponding to any channel. The filtered partial discharge sound signal is normalized to obtain the normalized partial discharge sound signal corresponding to any channel. Based on the normalized partial discharge sound signal, the initial complex wavelet coefficients of any channel are obtained; Based on the original partial discharge sound signals corresponding to each of the multiple channels, the initial complex wavelet coefficients corresponding to each of the multiple channels are obtained by using the method of obtaining the initial complex wavelet coefficients of any one of the channels.

3. The method according to claim 2, characterized in that, When there are multiple initial complex wavelet coefficients for any channel, obtaining the initial complex wavelet coefficients for any channel based on the normalized partial discharge sound signal includes: The normalized partial discharge sound signal is framed according to a predetermined sampling duration to obtain multiple frame signals for any channel. The dual-tree complex wavelet transform method is used to decompose the multiple frame signals to obtain the initial complex wavelet coefficients corresponding to each of the multiple frame signals; The initial complex wavelet coefficients of any channel are obtained based on the initial complex wavelet coefficients corresponding to each of the multiple frame signals.

4. The method according to claim 3, characterized in that, When any frame signal is decomposed into multiple scales, and each scale corresponds one-to-one with multiple initial complex wavelet coefficients, the method of using a dual-tree complex wavelet transform to decompose the multiple frame signals of any channel to obtain the initial complex wavelet coefficients corresponding to each of the multiple frame signals includes: Using the dual-tree complex wavelet transform method, any frame signal of any channel is decomposed in the following manner to obtain any initial complex wavelet coefficients of any frame signal: ； in, Denotes any initial complex wavelet coefficients. Let represent the first wavelet coefficient at any scale, where any scale is the scale corresponding to any initial complex wavelet coefficient. This represents the function value at any sampling point. This represents the function value of the first wavelet basis function at any sampling point for any given scale. This represents the translation of the first wavelet basis function in the time domain. This represents the second real wavelet coefficient at any of the stated scales. The value of the second wavelet basis function at any sampling point represents the value of the function at any scale. The second wavelet basis function is represented by its translation in the time domain, where i represents the imaginary unit, j represents the index of any scale, k represents the index of the position at any scale, n represents the index of any sampling point, and N represents the number of sampling points. The method of obtaining any one of the initial complex wavelet coefficients is used to obtain the multiple initial complex wavelet coefficients corresponding to each of the multiple frame signals.

5. The method according to claim 1, characterized in that, In a scenario where any channel includes multiple frame signals, and any frame signal is decomposed into multiple scales, and the multiple scales correspond one-to-one with multiple optimized complex wavelet coefficients, the optimization of the initial complex wavelet coefficients corresponding to each of the multiple channels to obtain the optimized complex wavelet coefficients corresponding to each of the multiple channels includes: Any optimized complex wavelet coefficients can be obtained by optimizing any initial complex wavelet coefficients in the following way: Obtain the frequency domain attention weight of any initial complex wavelet coefficient, wherein the frequency domain attention weight is used to quantify the importance of the target frequency domain energy of the corresponding initial complex wavelet coefficient at any frequency point relative to the total target frequency domain energy at multiple frequency points, the target frequency domain energy represents the frequency domain energy that is filtered from the initial frequency domain energy and is associated with the partial discharge characteristics of the power equipment, and the initial frequency domain energy represents the energy distribution of the corresponding initial complex wavelet coefficient in the frequency domain; Obtain the scale attention weight at any scale, where any scale represents the scale corresponding to the initial complex wavelet coefficients, and the scale attention weight is used to quantify the importance of the amplitude energy at the corresponding scale compared to the total amplitude energy at the multiple scales; Based on the initial complex wavelet coefficients, the scale attention weights, and the frequency domain attention weights, the optimized complex wavelet coefficients are obtained. The optimized complex wavelet coefficients corresponding to each of the multiple channels are obtained by using the method of obtaining any one of the optimized complex wavelet coefficients.

6. The method according to claim 5, characterized in that, The step of obtaining the frequency domain attention weights of any initial complex wavelet coefficients includes: Based on any initial complex wavelet coefficient, the initial frequency domain energy of any initial complex wavelet coefficient is obtained by using the fast Fourier transform method. The initial frequency domain energy is bandpass filtered to obtain the target frequency domain energy; The target frequency domain energy is normalized to obtain the frequency domain attention weight.

7. The method according to claim 5, characterized in that, The process of obtaining scale attention weights at any scale includes: Determine the phase deviation corresponding to each of the multiple initial complex wavelet coefficients in any given scale; Under the condition of meeting the preset conditions, the plurality of initial complex wavelet coefficients are filtered to obtain the filtered complex wavelet coefficients. The preset conditions indicate that the phase deviation of any initial complex wavelet coefficient is less than a preset phase deviation threshold. The filtered complex wavelet coefficients represent the complex wavelet coefficients that are filtered from the plurality of initial complex wavelet coefficients and are associated with the partial discharge characteristics of the power equipment. Based on the filtered complex wavelet coefficients, the amplitude energy at any scale is determined; The amplitude energy corresponding to each of the multiple scales is obtained by using the method of obtaining the amplitude energy at any of the scales. The scale attention weight for any one of the multiple scales is determined based on the amplitude energy corresponding to each scale.

8. The method according to claim 1, characterized in that, In a scenario where any channel includes multiple frame signals, and any frame signal is decomposed into multiple scales, with each scale corresponding to a specific target complex wavelet coefficient, the step of denoising the optimized complex wavelet coefficients corresponding to each of the multiple channels to obtain the target complex wavelet coefficients corresponding to each of the multiple channels includes: The target complex wavelet coefficients are obtained by denoising any optimized complex wavelet coefficients in the following manner: Determine the phase deviation of any of the optimized complex wavelet coefficients; Based on the preset phase deviation and the phase deviation of any optimized complex wavelet coefficient, the optimized complex wavelet coefficient is masked to obtain any masked complex wavelet coefficient. Based on any of the mask complex wavelet coefficients, an adaptive threshold denoising method is used to obtain any of the target complex wavelet coefficients; The target complex wavelet coefficients for each of the multiple channels are obtained by using the method of obtaining any one of the target complex wavelet coefficients.

9. The method according to any one of claims 1 to 8, characterized in that, In a scenario where any channel includes multiple frame signals, and any frame signal is decomposed into multiple scales, and the multiple scales correspond one-to-one with multiple target complex wavelet coefficients, obtaining the target partial discharge sound signal of the power equipment based on the target complex wavelet coefficients corresponding to each of the multiple channels includes: The dual-tree complex wavelet inverse transform method is used to reconstruct the complex wavelet coefficients of multiple targets in any frame of signal to obtain the reconstructed partial discharge sound signal of the any frame of signal. The reconstructed partial discharge sound signals corresponding to each of the multiple frame signals are overlapped and spliced ​​to obtain the reconstructed partial discharge sound signal of any channel. The recombined partial discharge sound signals corresponding to each of the multiple channels are integrated to obtain the target partial discharge sound signal.

10. An electronic device, characterized in that, The device includes one or more processors and a memory, the memory being used to store one or more programs for execution, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the partial discharge sound signal denoising processing method for power equipment as described in any one of claims 1 to 9.

Images