Partial discharge signal denoising method, device, electronic device and storage medium

Abstract

The present invention discloses a method, device, electronic device and storage medium for denoising a local discharge signal. The method comprises: determining a time-frequency domain spectrum of an acquired local discharge signal, determining an edge image of the time-frequency domain spectrum, determining a discharge feature region in the edge image according to the waveform characteristics of the local discharge signal, and removing noise in the time-frequency domain spectrum according to the discharge feature region. The embodiment of the present invention utilizes the waveform feature difference between the local discharge signal and the noise signal in the frequency domain spectrum, and then determines the discharge feature region in the time-frequency domain spectrum of the local discharge signal according to the waveform characteristics of the local discharge signal. Based on the discharge feature region, it is convenient to remove the noise signal in the local discharge signal, and can assist the staff to quickly identify the working status of the power equipment.

Description

Technical Field

[0001] The present invention relates to the field of signal processing technology, and in particular to a local discharge signal denoising method, device, electronic equipment and storage medium. Background Art

[0002] The insulation degradation of medium and high voltage power equipment will shorten the service life of the equipment. In severe cases, the defects will develop into permanent faults such as short circuit and open circuit. Partial discharge is the main manifestation of insulation degradation of power equipment. The status of power equipment can be effectively identified by analyzing the partial discharge signal of power equipment. However, due to the influence of the measurement site environment and measuring instruments, the collected partial discharge waveform will inevitably be interfered by noise, resulting in a large deviation when extracting the characteristics of the partial discharge signal, which has a negative impact on the fault diagnosis of power equipment.

[0003] Since the partial discharge signal is a typical non-stationary signal, there will be large errors if only the time domain tool or frequency domain tool is used for analysis. At the same time, there is a large electromagnetic interference phenomenon in the partial discharge test, and the time domain diagram and Fourier transform frequency domain diagram cannot effectively separate the noise signal from the partial discharge signal. Summary of the invention

[0004] The present invention provides a local discharge signal denoising method, device, electronic device and storage medium, which obtains the time-frequency domain spectrum of the local discharge signal, determines the discharge feature area in the edge image of the time-frequency domain spectrum according to the waveform characteristics of the local discharge signal, and removes the noise in the time-frequency domain spectrum according to the discharge feature area. By utilizing the waveform feature difference between the local discharge signal and the noise signal in the frequency domain spectrum, the noise signal in the local discharge signal is easily removed based on the discharge feature area, which can assist the staff to quickly identify the working status of the power equipment.

[0005] According to one aspect of the present invention, a method for denoising a partial discharge signal is provided, the method comprising:

[0006] Determine the time-frequency domain spectrum of the acquired partial discharge signal;

[0007] Determine the edge image of the time-frequency domain spectrum;

[0008] The discharge feature area in the edge image is determined according to the waveform characteristics of the local discharge signal, and the noise in the time-frequency domain spectrum is eliminated according to the discharge feature area.

[0009] According to another aspect of the present invention, a partial discharge signal denoising device is provided, the device comprising:

[0010] A spectrum determination module, used to determine the time-frequency domain spectrum of the acquired partial discharge signal;

[0011] An edge determination module, used for determining an edge image of a time-frequency domain spectrum;

[0012] The noise removal module is used to determine the discharge feature area in the edge image according to the waveform characteristics of the local discharge signal, and remove the noise in the time-frequency domain spectrum according to the discharge feature area.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the partial discharge signal denoising method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the partial discharge signal denoising method described in any embodiment of the present invention when executed.

[0018] The technical solution of the embodiment of the present invention determines the time-frequency domain spectrum of the acquired partial discharge signal, determines the edge image of the time-frequency domain spectrum, determines the discharge feature area in the edge image according to the waveform characteristics of the partial discharge signal, and removes the noise in the time-frequency domain spectrum according to the discharge feature area. The embodiment of the present invention utilizes the waveform feature difference between the partial discharge signal and the noise signal in the frequency domain spectrum, and then determines the discharge feature area in the time-frequency domain spectrum of the partial discharge signal according to the waveform characteristics of the partial discharge signal. Based on the discharge feature area, it is easy to remove the noise signal in the partial discharge signal, which can assist the staff to quickly identify the working status of the power equipment.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of a partial discharge signal denoising method provided according to Embodiment 1 of the present invention;

[0022] Figure 2 is a flow chart of a partial discharge signal denoising method provided according to Embodiment 2 of the present invention;

[0023] Figure 3 is a flow chart of a partial discharge signal denoising method provided according to Embodiment 3 of the present invention;

[0024] Figure 4 is a structural example diagram of a test circuit provided according to Embodiment 3 of the present invention;

[0025] Figure 5 is a time domain waveform diagram of a partial discharge signal provided according to Embodiment 3 of the present invention;

[0026] Figure 6 is a time-frequency domain spectrum of a partial discharge signal provided according to the third embodiment of the present invention;

[0027] Figure 7 This is an example diagram of Hough transform detection provided according to Embodiment 3 of the present invention;

[0028] Figure 8 is a time-frequency domain spectrum of a denoised partial discharge signal provided by Embodiment 3 of the present invention;

[0029] Fig. 9 is a structural schematic diagram of a partial discharge signal denoising device provided according to a fourth embodiment of the present invention;

[0030] Fig.10 It is a structural schematic diagram of an electronic device for implementing the partial discharge signal denoising method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] Embodiment 1

[0034] Figure 1 A flowchart of a method for denoising a partial discharge signal is provided for the first embodiment of the present invention. This embodiment is applicable to the case of removing noise signals from a partial discharge signal. The method can be performed by a partial discharge signal denoising device, which can be implemented in the form of hardware and / or software. Figure 1 As shown, a partial discharge signal denoising method provided in this embodiment 1 specifically includes the following steps:

[0035] S110, determining a time-frequency domain spectrum of the acquired partial discharge signal.

[0036] The partial discharge signal may refer to a target signal to be denoised, and the partial discharge signal may be a signal actually collected at the power equipment, or may be a signal simulated by a signal generator, which is not limited in the embodiment of the present invention. The time-frequency domain spectrum may be understood as the result obtained after the partial discharge signal is analyzed and processed in the time-frequency domain, and the horizontal axis of the time-frequency domain spectrum is time, and the vertical axis is frequency.

[0037] In an embodiment of the present invention, a local discharge signal containing noise can be first obtained by means including but not limited to actual collection at the power equipment or simulation by a signal generator, and then the local discharge signal can be processed using a pre-configured time-frequency domain analysis tool or other time-frequency domain analysis algorithm to obtain a corresponding time-frequency domain spectrum, wherein the time-frequency domain analysis tool or other time-frequency domain analysis algorithm may include but is not limited to: short-time Fourier transform, wavelet transform, Cohen-type time-frequency distribution and Affine-type time-frequency distribution, etc., and the embodiment of the present invention does not limit this.

[0038] S120, determining the edge image of the time-frequency domain spectrum.

[0039] The edge image may refer to an image obtained after edge detection is performed on a time-frequency domain spectrum.

[0040] In an embodiment of the present invention, a pre-configured image edge detection rule can be called to detect edge features in the time-frequency domain spectrum to obtain a corresponding edge image, wherein the image edge detection rules used may include but are not limited to the following: Sobel edge detection rule, Roberts edge detection rule and Canny edge detection rule, etc., and the embodiment of the present invention does not limit this.

[0041] S130 , determining a discharge feature region in the edge image according to the waveform features of the partial discharge signal, and removing noise in the time-frequency domain spectrum according to the discharge feature region.

[0042] The waveform feature may refer to the waveform shape feature of the partial discharge signal in the time-frequency domain spectrum. Exemplarily, the waveform feature at least includes that the waveform shape in the time-frequency domain spectrum is an ellipse or other specific shapes. The discharge characteristic region may be understood as a characteristic region that needs to be retained after denoising in the time-frequency domain spectrum corresponding to the partial discharge signal. The discharge characteristic region may be displayed as an ellipse or other specific shapes.

[0043] In the embodiment of the present invention, since the waveform shape of the local discharge signal in its time-frequency domain spectrum is a certain specific shape, for example, when the local discharge signal is a decaying oscillation waveform, it is presented as an elliptical aperture in the time-frequency domain spectrum after the Choi-Williams transform, and the noise signal is mostly scattered or linearly distributed in the time-frequency domain diagram. Based on this principle, a pre-configured waveform feature detection rule can be called to detect the discharge feature area with waveform characteristics in the edge image, and only the pixel value of the discharge feature area is retained in the time-frequency domain spectrum of the local discharge signal, that is, all the pixel values ​​outside the discharge feature area are filled with 0, thereby removing the noise of the local discharge signal in its time-frequency domain spectrum, wherein the waveform feature detection rule may include but is not limited to the following: Hough transform detection rule, watershed detection rule and GrabCut detection rule, etc., which are not limited in the embodiment of the present invention.

[0044] The technical solution of the embodiment of the present invention determines the time-frequency domain spectrum of the acquired partial discharge signal, determines the edge image of the time-frequency domain spectrum, determines the discharge feature area in the edge image according to the waveform characteristics of the partial discharge signal, and removes the noise in the time-frequency domain spectrum according to the discharge feature area. The embodiment of the present invention utilizes the waveform feature difference between the partial discharge signal and the noise signal in the frequency domain spectrum, and then determines the discharge feature area in the time-frequency domain spectrum of the partial discharge signal according to the waveform characteristics of the partial discharge signal. Based on the discharge feature area, it is easy to remove the noise signal in the partial discharge signal, which can assist the staff to quickly identify the working status of the power equipment.

[0045] Embodiment 2

[0046] Figure 2 This is a flow chart of a partial discharge signal denoising method provided in the second embodiment of the present invention, which is further optimized and expanded based on the above embodiment and can be combined with various optional technical solutions in the above embodiment. Figure 2 As shown, a partial discharge signal denoising method provided in the second embodiment specifically includes the following steps:

[0047] S210, calling a preset time-frequency domain analysis tool to process the partial discharge signal to obtain a time-frequency domain spectrum.

[0048] Among them, the preset time-frequency domain analysis tool may refer to a pre-configured signal analysis tool for processing partial discharge signals to obtain time-frequency domain spectra. The preset time-frequency domain analysis tools may include: Gabor transform, S transform, Wigner distribution and Choi-Williams transform, etc.

[0049] In an embodiment of the present invention, a pre-configured preset time-frequency domain analysis tool can be called to perform time-frequency domain analysis processing on the acquired partial discharge signal to obtain a corresponding time-frequency domain spectrum, wherein the preset time-frequency domain analysis tool may include but is not limited to the following: Gabor transform, S transform, Wigner distribution and Choi-Williams transform, etc. Since the partial discharge signal contains noise, the time-frequency domain spectrum obtained after the time-frequency domain analysis processing will contain a large number of linear and point-shaped signal components, which need to be removed by some means. It should be understood that the acquisition method of the partial discharge signal can be either a signal actually collected at the power equipment or a signal simulated by a signal generator, and the embodiment of the present invention does not limit this.

[0050] Further, based on the above-mentioned embodiment of the invention, the preset time-frequency domain analysis tool may at least include Choi-Williams transformation, and the formula of Choi-Williams transformation is expressed as follows:

[0051]

[0052] Wherein, x(t) represents the local discharge signal, S(t,w) represents the Choi-Williams transform result of the local discharge signal, W(t,τ,u) represents the smoothing window function, t represents time, w represents frequency, * represents the complex conjugate operation, e represents the exponential operation with a natural constant as the base, and j represents the imaginary unit symbol.

[0053] S220, grayscale the time-frequency domain spectrum to obtain a grayscale image of the spectrum.

[0054] The atlas grayscale image may refer to a grayscale image obtained by graying the time-frequency domain atlas.

[0055] In an embodiment of the present invention, after obtaining the time-frequency domain spectrum corresponding to the partial discharge signal, it can be grayed to obtain the corresponding spectrum grayscale image. The grayscale value of the pixel in the spectrum grayscale image ranges from 0 to 255, where a grayscale value of 255 represents the brightest brightness and a grayscale value of 0 represents the darkest brightness.

[0056] S230, using pixel blocks with gray value of 0 to perform edge expansion on the atlas grayscale image, then using a window of a preset window size to traverse and select all image blocks in the atlas grayscale image with edge expansion, and composing the image blocks into an element matrix.

[0057] In an embodiment of the present invention, if the resolution of the atlas grayscale image is m×n, the edge of the image can be filled with pixel blocks with a grayscale value of 0. The resolution of the image matrix obtained after the edge filling is (m+2)×(n+2). Then, a preset window size, such as a 3×3 window, is used to traverse and select all image blocks in the atlas grayscale image with edge expansion. The selected image blocks are used as single elements to form a new matrix, namely, an element matrix, whose resolution is m×n, and each element in the element matrix is ​​a 3×3 matrix.

[0058] S240 , calling a horizontal Sobel operator and a vertical Sobel operator to respectively multiply the elements in the element matrix to obtain a horizontal gradient matrix and a vertical gradient matrix.

[0059] In an embodiment of the present invention, a preconfigured horizontal Sobel operator and a vertical Sobel operator may be called to multiply the elements in the element matrix respectively, thereby obtaining a corresponding horizontal gradient matrix and a vertical gradient matrix, wherein the resolution of the horizontal gradient matrix and the vertical gradient matrix are both m×n, and each element in the matrix is ​​a 3×3 matrix.

[0060] Further, based on the above-mentioned embodiment of the invention, the horizontal Sobel operator and the vertical Sobel operator at least include:

[0061]

[0062] Wherein, Dx represents the horizontal Sobel operator, and Dy represents the vertical Sobel operator.

[0063] S250: Taking the sum of the square of the horizontal gradient matrix and the square of the vertical gradient matrix as the gradient matrix.

[0064] In an embodiment of the present invention, after obtaining the horizontal gradient matrix and the vertical gradient matrix, the corresponding gradient value can be further calculated, that is, the sum of the square of the horizontal gradient matrix and the square of the vertical gradient matrix is ​​used as the gradient matrix. Exemplarily, if the horizontal gradient matrix is ​​Gx and the vertical gradient matrix is ​​Gy, then the gradient matrix G = (Gx) 2 +(Gy) 2 .

[0065] S260 , performing a sum operation on all elements in the gradient matrix to obtain an edge image.

[0066] In an embodiment of the present invention, each element in the gradient matrix is ​​a 3×3 matrix, and each element in the gradient matrix can be added to obtain a new grayscale value. After performing a summation operation on all elements in the gradient matrix, an edge image can be obtained.

[0067] S270 , calling a preset Hough transform detection rule to detect a discharge feature region with waveform characteristics in the edge image after binarization processing.

[0068] The preset Hough transform detection rule may refer to a pre-configured Hough transform detection rule for detecting a discharge feature area. The Hough transform is a method for extracting image features and can be used to detect shapes such as a straight line, a circle, and an ellipse.

[0069] In the embodiment of the present invention, since the waveform shape of the local discharge signal in the time-frequency domain spectrum is a certain specific shape, for example, when the local discharge signal is a decaying oscillation waveform, it appears as an elliptical aperture in the time-frequency domain spectrum after the Choi-Williams transform, while the noise signal mostly presents a scattered distribution or a linear distribution in the time-frequency domain diagram, based on this principle, the pre-configured preset Hough transform detection rule can be called to extract the waveform features in the edge image after the binarization processing, so as to extract the required discharge feature area.

[0070] S280 , assigning a value of 0 to a coordinate point corresponding to a non-discharge characteristic region in the time-frequency domain spectrum according to the discharge characteristic region, so as to eliminate noise in the time-frequency domain spectrum.

[0071] The non-discharge feature region may refer to a feature region that needs to be removed in the time-frequency domain spectrum, that is, a region represented by noise or other interference.

[0072] In an embodiment of the present invention, after the discharge characteristic region is detected, the coordinate points of the non-discharge characteristic region, i.e., the region outside the discharge characteristic region, can be assigned a value of 0 in the time-frequency domain spectrum, and only the coordinate point information in the discharge characteristic region is retained, thereby eliminating the noise of the local discharge signal in the time-frequency domain spectrum. Further, after eliminating the noise of the local discharge signal in the time-frequency domain spectrum, if the discharge characteristic region is an ellipse, the center point of each ellipse in the time-frequency domain spectrum can be extracted, and the horizontal coordinate of the center point is used as the generation time of the local discharge signal, and the vertical coordinate of the center point is used as the oscillation frequency of the local discharge signal. The oscillation frequency characteristics of the local discharge signal can help analyze and identify the working state of the power equipment.

[0073] The technical solution of the embodiment of the present invention is to process the local discharge signal by calling a preset time-frequency domain analysis tool to obtain a time-frequency domain spectrum, grayscale the time-frequency domain spectrum to obtain a spectrum grayscale image, use a pixel block with a grayscale value of 0 to expand the edge of the spectrum grayscale image, and then use a window of a preset window size to traverse and select all image blocks in the edge-expanded spectrum grayscale image, and form an element matrix with the image blocks, call the horizontal Sobel operator and the vertical Sobel operator to multiply the elements in the element matrix respectively to obtain a horizontal gradient matrix and a vertical gradient matrix, take the sum of the square of the horizontal gradient matrix and the square of the vertical gradient matrix as the gradient matrix, perform a sum operation on all elements in the gradient matrix to obtain an edge image, call a preset Hough transform detection rule to detect the discharge feature area with waveform characteristics in the edge image after binarization, and assign the coordinate point corresponding to the non-discharge feature area in the time-frequency domain spectrum to 0 according to the discharge feature area to achieve the removal of noise in the time-frequency domain spectrum. The embodiment of the present invention calculates the time-frequency domain spectrum of the partial discharge signal by calling a preset time-frequency domain analysis tool and performs grayscale processing, then uses the Sobel operator to detect the edge features of the partial discharge signal in the time-frequency domain distribution, and finally uses the waveform feature difference between the partial discharge signal and the noise signal in the frequency domain spectrum to call the preset Hough transform detection rule to extract the component containing the waveform feature in the edge image. Based on the waveform feature component, the noise signal in the partial discharge signal is effectively and quickly removed, which can assist the staff to quickly identify the working status of the power equipment.

[0074] Embodiment 3

[0075] Figure 3 This is a flow chart of a method for denoising a partial discharge signal provided in the third embodiment of the present invention. Based on the above embodiments, this embodiment provides an implementation of a method for denoising a partial discharge signal, which can remove noise signals from partial discharge signals. Figure 3 As shown, a method for denoising a partial discharge signal provided by Embodiment 3 of the present invention specifically includes the following steps:

[0076] S310. Build different types of partial discharge models.

[0077] In the embodiment of the present invention, four types of partial discharge models (i.e., partial discharge models) are constructed, namely, corona discharge, surface discharge, internal discharge, and suspension discharge. The test circuit structure is as follows: Figure 4 As shown in the figure, a high frequency current transformer (HFCT) with a working frequency band of 1 to 30 MHz and a transmission impedance greater than 8Ω is used as a monitoring terminal for partial discharge detection. It is placed in a ground loop and the partial discharge signal generated by the partial discharge model is observed through the trigger mode of the RIGOL MSO5104 oscilloscope. The oscilloscope has 4 analog channels and a bandwidth of 100 MHz, a real-time sampling rate of 8 Gsa / s, a storage depth of 100 Mpts, and a waveform capture rate of more than 500,000 waveforms per second, which can fully meet the needs of experimental testing.

[0078] S320, collecting time domain data of the original partial discharge signal, and obtaining its corresponding time-frequency domain spectrum based on Choi-Williams transformation.

[0079] In the embodiment of the present invention, the time domain spectrum of the original partial discharge signal can be collected by using a signal acquisition device such as an oscilloscope, and its corresponding time-frequency domain spectrum can be obtained based on the Choi-Williams transformation. Specifically, assuming that the original partial discharge signal is x(t), the Cohen-type time-frequency distribution is used to obtain the converted function S(t,w). The commonly used time-frequency domain distribution method is the Wigner-Vill distribution. The definition of the Wigner-Vill distribution is as follows:

[0080]

[0081] In order to reduce the influence of cross terms on the judgment of partial discharge signals, the embodiment of the present invention adopts Choi-Williams transform to improve the time-frequency domain spectrum. The essence of Choi-Williams transform is to introduce a smoothing window function based on the Wigner-Vill distribution. The smoothing window function adopted in the embodiment of the present invention is shown in the following formula:

[0082]

[0083] Wherein, ξ represents the attenuation coefficient of the window function. The principle of selecting the attenuation coefficient is to ensure as high a resolution of the time-frequency spectrum as possible without cross-term interference. In the embodiment of the present invention, ξ=1 is selected. After substituting into the window function formula, the Choi-Williams transformation formula for processing partial discharge signals can be obtained:

[0084]

[0085] S330, calling the Sobel detection rule to obtain an edge grayscale image of the time-frequency domain spectrum.

[0086] In an embodiment of the present invention, it is assumed that the resolution of the time-frequency domain image is m×n, and the time-frequency domain image is gray-scale processed to obtain an m×n partial discharge signal gray-scale matrix G. Before edge detection, the periphery of the matrix G is filled with pixel blocks with a gray-scale value of 0 to form a new matrix G' of size (m+2)×(n+2). Then, a 3×3 square pixel frame is used to traverse and select different pixel blocks in the matrix G'. The selected pixel blocks are used as single elements to form a matrix E, and the size of the matrix E is m×n. The Sobel operator is multiplied with each element in the matrix E to obtain the x-direction gradient matrix Gx and the y-direction gradient matrix Gy formulas as follows:

[0087]

[0088] The sizes of Gx and Gy are both m×n, and each element of the matrix is ​​also a 3×3 matrix. The matrices Gx and Gy are further operated to obtain the gradient matrix G. The operation process is as follows:

[0089] G(m,n)=[G x (m,n)] 2 +[G y (m,n)] 2

[0090] Among them, the size of the gradient matrix G is m×n, and each element of the matrix G is also a 3×3 matrix. Each element in the gradient matrix G(m,n) is added to obtain a new grayscale value. The sum of all elements in the matrix G can be calculated to obtain an edge image grayscale matrix A of size m×n. The element grayscale value range of this matrix is ​​0 to 255. The matrix is ​​visualized to obtain an edge grayscale image of the time-frequency domain spectrum.

[0091] S340, calling Hough transform detection rules to extract partial discharge signals that do not contain noise signals and remove noise signals.

[0092] In the embodiment of the present invention, since the local discharge signal generally adopts the local discharge signal of the exponential decay oscillation waveform, it is presented as an aperture similar to an ellipse in the time-frequency domain spectrum after the Choi-Williams transformation, and the noise signal is mostly scattered or linearly distributed in the time-frequency domain spectrum, it is possible to select a similar elliptical waveform for the noisy time-frequency domain waveform to extract the local discharge signal that does not contain the noise signal. Specifically, the edge grayscale image can be binarized first, and then the processed edge grayscale image can be subjected to the Hough transform. The principle of the Hough transform is to use the following elliptical coordinate equation:

[0093]

[0094] Where a and b represent the semi-major axis and semi-minor axis of the ellipse, respectively, (p0, q0) represents the center point of the ellipse, (p, q) represents the coordinate point on the ellipse, and θ represents the rotation angle. Substitute all the pixel points of the edge grayscale image after binary processing into the above ellipse coordinate equation. After continuous iterative fitting, several elliptical apertures will be extracted. In the time-frequency domain spectrum, the coordinate points outside the elliptical aperture are assigned to 0, and the denoised time-frequency domain image of the partial discharge signal can be obtained. In addition, the generation time and oscillation frequency of the partial discharge signal can be determined by the coordinates of the center points of each elliptical aperture (p0, q0), where the generation time is p0 and the oscillation frequency is q0. The oscillation frequency characteristics of the partial discharge signal can help analyze and identify the working status of the power equipment.

[0095] Figure 5 This is a time domain waveform diagram of a partial discharge signal provided by the third embodiment of the present invention. Figure 5 As shown in Figure 1, the local discharge signal has a large amount of electromagnetic noise. The corresponding time-frequency domain spectrum is obtained by performing Choi-Williams transformation on the local discharge signal containing noise. Figure 6 As shown by Figure 6 It can be seen that the brightness of the local discharge signal in the time-frequency domain spectrum is weakened after the superposition of noise. At the same time, in addition to the obvious elliptical aperture, there are also a large number of linear and point-shaped signal components, which proves the existence of noise interference in the time-frequency domain spectrum. Figure 7 This is an example diagram of a Hough transform detection provided by the third embodiment of the present invention. Figure 7 As shown in Figure 2, several elliptical apertures can be extracted using the Hough transform. Figure 6 Assign the coordinate points outside the elliptical aperture to 0, and you can get the following Figure 8 The denoised time-frequency domain image is shown in Figure 2. Figure 8 It can be seen that the occurrence time and discharge frequency of the partial discharge signal are relatively clear, and the denoising of the partial discharge signal is achieved.

[0096] The local discharge signal denoising method provided in the third embodiment of the present invention can be used to collect and analyze different types of typical local discharge signals in the laboratory, obtain the oscillation frequency characteristics of different types of local discharge signals, and use this parameter as a reference for field data, which has great reference value for actual engineering.

[0097] The technical solution of the embodiment of the present invention is to build different types of partial discharge models, collect the time domain data of the original partial discharge signal, obtain the corresponding time-frequency domain spectrum based on the Choi-Williams transform, call the Sobel detection rule to obtain the edge grayscale image of the time-frequency domain spectrum, call the Hough transform detection rule to extract the partial discharge signal that does not contain the noise signal and remove the noise signal. The embodiment of the present invention builds different types of partial discharge models, obtains the time-frequency domain spectrum of the partial discharge signal based on the Choi-Williams transform, and then calls the Sobel detection rule to detect the edge characteristics of the partial discharge signal in the time-frequency domain distribution. Finally, the waveform feature difference between the partial discharge signal and the noise signal in the frequency domain spectrum is used to call the Hough transform detection rule to extract the component containing the elliptical feature in the edge image. Based on the elliptical feature component, the noise signal in the partial discharge signal is effectively and quickly removed. The filtered waveform can effectively identify the characteristic frequency and time domain features of the partial discharge signal, which can assist the staff to quickly identify the working status of the power equipment.

[0098] Embodiment 4

[0099] Fig. 9 This is a schematic diagram of the structure of a partial discharge signal denoising device provided in Embodiment 4 of the present invention. Fig. 9 As shown, the device comprises:

[0100] The spectrum determination module 41 is used to determine the time-frequency domain spectrum of the acquired partial discharge signal.

[0101] The edge determination module 42 is used to determine the edge image of the time-frequency domain spectrum.

[0102] The noise removal module 43 is used to determine the discharge feature area in the edge image according to the waveform characteristics of the partial discharge signal, and remove the noise in the time-frequency domain spectrum according to the discharge feature area.

[0103] The technical solution of the embodiment of the present invention is to determine the time-frequency domain spectrum of the acquired partial discharge signal through the spectrum determination module, determine the edge image of the time-frequency domain spectrum by the edge determination module, and determine the discharge feature area in the edge image according to the waveform characteristics of the partial discharge signal by the noise removal module, and remove the noise in the time-frequency domain spectrum according to the discharge feature area. The embodiment of the present invention utilizes the waveform feature difference between the partial discharge signal and the noise signal in the frequency domain spectrum, and then determines the discharge feature area in the time-frequency domain spectrum of the partial discharge signal according to the waveform characteristics of the partial discharge signal. Based on the discharge feature area, it is convenient to remove the noise signal in the partial discharge signal, and can assist the staff to quickly identify the working status of the power equipment.

[0104] Further, based on the above-mentioned embodiment of the invention, the spectrum determination module 41 includes:

[0105] The spectrum acquisition unit is used to call a preset time-frequency domain analysis tool to process the partial discharge signal to obtain a time-frequency domain spectrum.

[0106] Further, based on the above-mentioned embodiment of the invention, the preset time-frequency domain analysis tool at least includes Choi-Williams transformation, and the formula of Choi-Williams transformation is expressed as follows:

[0107]

[0108] Wherein, x(t) represents the local discharge signal, S(t,w) represents the Choi-Williams transform result of the local discharge signal, W(t,τ,u) represents the smoothing window function, t represents time, w represents frequency, * represents the complex conjugate operation, e represents the exponential operation with a natural constant as the base, and j represents the imaginary unit symbol.

[0109] Further, based on the above-mentioned embodiment of the invention, the edge determination module 42 includes:

[0110] The grayscale processing unit is used to perform grayscale processing on the time-frequency domain spectrum to obtain a grayscale image of the spectrum.

[0111] The element matrix acquisition unit is used to use pixel blocks with gray value 0 to expand the edge of the atlas grayscale image, and then use a window of a preset window size to traverse and select all image blocks in the atlas grayscale image with edge expansion, and form the image blocks into an element matrix.

[0112] The operator processing unit is used to call the horizontal Sobel operator and the vertical Sobel operator to multiply the elements in the element matrix respectively to obtain the horizontal gradient matrix and the vertical gradient matrix.

[0113] The gradient matrix acquisition unit is used to take the sum of the square of the horizontal gradient matrix and the square of the vertical gradient matrix as the gradient matrix.

[0114] The edge matrix acquisition unit is used to perform a sum operation on all elements in the gradient matrix to obtain an edge image.

[0115] Further, based on the above-mentioned embodiment of the invention, the horizontal Sobel operator and the vertical Sobel operator at least include:

[0116]

[0117] Further, based on the above-mentioned embodiment of the invention, the noise removal module 43 includes:

[0118] The feature region detection unit is used to call a preset Hough transform detection rule to detect a discharge feature region with waveform characteristics in the edge image after binarization processing.

[0119] The noise removal unit is used to assign the coordinate points corresponding to the non-discharge characteristic areas in the time-frequency domain spectrum to 0 according to the discharge characteristic areas, so as to remove the noise in the time-frequency domain spectrum.

[0120] Further, based on the above-mentioned embodiment of the invention, the waveform feature of the partial discharge signal at least includes that the waveform shape in the time-frequency domain spectrum is an ellipse.

[0121] The partial discharge signal denoising device provided in the embodiment of the present invention can execute the partial discharge signal denoising method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0122] Embodiment 5

[0123] Fig.10 A schematic diagram of an electronic device 50 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0124] like Fig.10As shown, the electronic device 50 includes at least one processor 51, and a memory connected to the at least one processor 51 in communication, such as a read-only memory (ROM) 52, a random access memory (RAM) 53, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 51 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 52 or the computer program loaded from the storage unit 58 to the random access memory (RAM) 53. In the RAM 53, various programs and data required for the operation of the electronic device 50 can also be stored. The processor 51, the ROM 52, and the RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0125] A number of components in the electronic device 50 are connected to the I / O interface 55, including: an input unit 56, such as a keyboard, a mouse, etc.; an output unit 57, such as various types of displays, speakers, etc.; a storage unit 58, such as a disk, an optical disk, etc.; and a communication unit 59, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 59 allows the electronic device 50 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0126] The processor 51 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 51 performs the various methods and processes described above, such as a partial discharge signal denoising method.

[0127] In some embodiments, the partial discharge signal denoising method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 58. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 50 via the ROM 52 and / or the communication unit 59. When the computer program is loaded into the RAM 53 and executed by the processor 51, one or more steps of the partial discharge signal denoising method described above may be performed. Alternatively, in other embodiments, the processor 51 may be configured to perform the partial discharge signal denoising method in any other appropriate manner (e.g., by means of firmware).

[0128] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0129] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0130] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0131] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0132] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0133] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0134] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0135] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for denoising a partial discharge signal, characterized in that: The method comprises: Determine the time-frequency domain spectrum of the acquired partial discharge signal; Determining an edge image of the time-frequency domain spectrum; Determine a discharge feature region in the edge image according to the waveform characteristics of the local discharge signal, and remove noise in the time-frequency domain spectrum according to the discharge feature region; Wherein, determining the discharge feature area in the edge image according to the waveform characteristics of the local discharge signal, and removing the noise in the time-frequency domain spectrum according to the discharge feature area, includes: Calling a preset Hough transform detection rule to detect the discharge feature region having the waveform feature in the edge image after binarization processing; According to the discharge characteristic area, the coordinate points corresponding to the non-discharge characteristic area in the time-frequency domain spectrum are assigned 0 to eliminate the noise in the time-frequency domain spectrum.

2. The method according to claim 1, characterized in that: The step of determining the acquired time-frequency domain spectrum of the partial discharge signal comprises: A preset time-frequency domain analysis tool is called to process the partial discharge signal to obtain the time-frequency domain spectrum.

3. The method according to claim 2, characterized in that The preset time-frequency domain analysis tool at least includes Choi-Williams transformation, and the formula of Choi-Williams transformation is as follows: Wherein, x(t) represents the local discharge signal, S(t,w) represents the Choi-Williams transform result of the local discharge signal, W(t,τ,u) represents the smoothing window function, t represents time, w represents frequency, * represents the complex conjugate operation, e represents the exponential operation with a natural constant as the base, and j represents the imaginary unit symbol.

4. The method according to claim 1, characterized in that: The step of determining the edge image of the time-frequency domain spectrum comprises: Performing grayscale processing on the time-frequency domain spectrum to obtain a grayscale image of the spectrum; Using pixel blocks with gray value of 0 to perform edge expansion on the atlas grayscale image, then using a window of a preset window size to traverse and select all image blocks in the atlas grayscale image with the edge expanded, and composing the image blocks into an element matrix; Calling a horizontal Sobel operator and a vertical Sobel operator to multiply the elements in the element matrix respectively to obtain a horizontal gradient matrix and a vertical gradient matrix; Taking the sum of the square of the horizontal gradient matrix and the square of the vertical gradient matrix as a gradient matrix; A sum operation is performed on all elements in the gradient matrix to obtain the edge image.

5. The method according to claim 4, characterized in that The horizontal Sobel operator and the vertical Sobel operator at least include: ， 。 6 . The method according to claim 1 , wherein the waveform feature of the partial discharge signal at least includes that the waveform shape in the time-frequency domain spectrum is an ellipse.

7. A local discharge signal denoising device, characterized in that: The device comprises: A spectrum determination module, used to determine the time-frequency domain spectrum of the acquired partial discharge signal; An edge determination module, used to determine the edge image of the time-frequency domain spectrum; A noise removal module, used to determine a discharge feature area in the edge image according to the waveform characteristics of the local discharge signal, and remove noise in the time-frequency domain spectrum according to the discharge feature area; Wherein, the noise removal module includes: A feature region detection unit, used for calling a preset Hough transform detection rule to detect the discharge feature region having the waveform feature in the edge image after binarization processing; The noise removal unit is used to assign the coordinate points corresponding to the non-discharge characteristic areas in the time-frequency domain spectrum to 0 according to the discharge characteristic areas, so as to remove the noise in the time-frequency domain spectrum.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the partial discharge signal denoising method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the partial discharge signal denoising method according to any one of claims 1 to 6 when executed.

Images