Analysis method and system for partial discharge energy distribution in oil

Abstract

The invention discloses an analysis method and system for partial discharge energy distribution in oil, and belongs to the technical field of maintenance and diagnosis of high-voltage power transmission and transformation equipment. The oil paper insulation partial discharge experiment platform collects a partial discharge PRPD spectrogram, a pulse waveform and a streamer image; compressing the streamer image, and performing multi-scale correction on the compressed image to obtain a primary image; performing gray level layering and segmented stretching processing on the primary image to obtain a secondary image; carrying out image binarization eight-neighborhood contour extraction on the secondary image to obtain a final contour of the partial discharge stream, and carrying out region division according to a gray value interval of the discharge stream; analyzing the relationship between the discharge stream form and the discharge pulse in combination with the pulse waveform; and analyzing the high energy ratio and the energy distribution complexity of partial discharge, and calculating the partial discharge energy. The oil paper insulation degradation mechanism is disclosed, and the insulation fault risk of the converter transformer is reduced.

Description

Technical Field

[0001] This application belongs to the field of high-voltage power transmission and transformation equipment maintenance and diagnosis technology, specifically relating to an analysis method and system for partial discharge energy distribution in oil. Background Technology

[0002] Converter transformers and converter valves together form the core of AC-DC transmission system connections, and the safe operation of converter transformers is crucial to ensuring the stability and reliability of DC transmission systems. However, insulation accidents in converter transformers occur frequently. The energy released by partial discharge is the main cause of oil-paper insulation faults, and its distribution is closely related to the morphology of the discharge current.

[0003] As a type of extremely lightweight magnetohydrodynamic fluid, the evolution of partial discharge beams exhibits rapidly changing dynamic characteristics. Current image processing algorithms are only suitable for discharge arcs with relatively regular shapes and slow development speeds, and cannot accurately extract the edge features of partial discharges. Furthermore, the correlation between partial discharge beams and pulse waveforms and PRPD spectra is unclear, making it impossible to elucidate the energy distribution patterns of partial discharges.

[0004] Therefore, there is an urgent need for a technical solution based on the extraction of the morphology of local discharge current beams in oil and the analysis of their energy distribution. By extracting the morphology of local discharge current beams and combining the partial discharge pulse waveform and PRPD spectrum information, the energy distribution law of partial discharge can be clarified. Summary of the Invention

[0005] The purpose of this invention is to address the problems in the prior art of extracting partial discharge current beams from oil and the unclear changes in the energy distribution of partial discharge, by proposing an analytical method and system for the energy distribution of partial discharge in oil.

[0006] In a first aspect, embodiments of the present invention provide a method for analyzing the energy distribution of partial discharge in oil, the method comprising the following steps: S1. Construct a visual experimental oil tank for partial discharge to simulate the real oil channel electric field environment of a converter transformer. Monitor the temperature inside the oil tank using a temperature sensor and monitor the pressure in the oil using a pressure sensor. S2. Construct an oil-paper insulation partial discharge experimental platform to collect PRPD spectrum, pulse waveform, and streamer image of partial discharge in oil; S3. The streamer image is compressed using the R, G, and B channels respectively. Weights are assigned to the optical signal image of each channel, and color restoration coefficients are added for multi-scale correction to obtain a primary image. S4. Perform grayscale layering and piecewise stretching on the primary image, connect adjacent pixels with equal grayscale values ​​into closed grayscale contour lines, and use a piecewise linear transformation function to perform contrast stretching to obtain a secondary image. S5. The eight-neighbor contour is extracted by image binarization of the secondary image to obtain the final contour of the local discharge current beam, and the discharge area is divided into Class I, Class II and Class III discharge areas with different brightness according to the gray value range of the discharge current beam. S6. Record the gray area, number of branches, and horizontal displacement of all partial discharges before the insulation paper breaks down, and analyze the relationship between the discharge current beam morphology and the discharge pulse in combination with the pulse waveform. S7. Define the high-energy proportion coefficient k by the ratio of the area of ​​Class I and Class III discharge regions, and use it to analyze the high-energy proportion of partial discharge; characterize the complexity of energy distribution by the fractal dimension and horizontal displacement of the discharge; calculate the partial discharge energy by multiplying the single discharge quantity Q and the applied voltage U in the PRPD spectrum.

[0007] Furthermore, the visualization experimental tank in step S1 is made of a fully transparent, high-mechanical-strength acrylic container, which has good thermal stability, light transmittance and electrical insulation. Based on the T-shaped cross oil passage at the end of the high-voltage winding of the converter transformer, the L-shaped oil passage near the electrostatic ring, and the U-shaped oil passage extending into the riser, an oil passage model of the experimental oil tank was constructed in an acrylic container using insulating cardboard. The high-voltage line and grounding line are connected to the copper foil attached to the outside of the insulating paper to simulate the real oil duct electric field environment of the converter transformer. A forced oil-guided circulation system consisting of an oil pump, oil flow sleeve, heating tube, and cooling fins is incorporated, and temperature and pressure sensors are installed.

[0008] Furthermore, in step S2, the experimental platform includes: a waveform generator, a power amplifier, a current-limiting resistor, a visualization experimental tank, a high-speed camera, a coupling capacitor, a high-voltage probe, an oscilloscope, a partial discharge tester, and a detection impedance. The waveform generator is connected to the power amplifier, the power amplifier is connected to the current-limiting resistor, the current-limiting resistor is connected to the visualization experiment tank and the coupling capacitor, a high-speed camera is installed on the visualization experiment tank, the coupling capacitor is connected to the detection impedance and the high-voltage probe, the detection impedance is connected to the visualization experiment tank, the oscilloscope and the partial discharge instrument, and the high-voltage probe is connected to the oscilloscope.

[0009] Furthermore, the specific steps for acquiring the PRPD spectrum, pulse waveform, and streamer image of partial discharge in oil in step S2 are as follows: S21. Set the waveform generator to generate an electrical signal, which is then amplified by a power amplifier to obtain an experimental voltage of 0-30kV; S22. The experimental voltage is applied to the high-voltage side of the experimental oil tank; S23. Acquire the PRPD spectrum of partial discharge using a partial discharge instrument, acquire the pulse waveform of partial discharge using a high-voltage probe and oscilloscope, and acquire the streamer of partial discharge using a high-speed camera; S24. Taking into account the exposure time of the high-speed camera and the transmission delay of the coaxial cable signal, the time relationship between the optical signal and the electrical signal is corrected.

[0010] Furthermore, the specific steps of step S3 are as follows: S31. The streamer image is dynamically compressed using the R, G, and B channels respectively to obtain the optical signal image of each channel; S32. Assign weights to the optical signal images of each channel, add color restoration coefficients for multi-scale correction, and then re-overlay the corrected images to form a new image. The formula for multi-scale correction is:

[0011]

[0012]

[0013] In the formula, i It can be any one of the three optical channels: R, G, and B. R i ( x , y (reflection image) R In the i In each optical channel ( x , y Pixel value at point ) I i ( x , y ) is the original image in the 1st i One optical channel ( x , y The pixel value at point ); * indicates convolution operation. G j ( x , y ) is the first j Gaussian wrap function at various scales; N For the total number of scales, j For a certain count, ω j For the first j The weights corresponding to each scale; C i ( x , y ) is the first i The color correction factor for each channel is used to adjust the color ratio of the three optical channels;α and β These represent the controlled nonlinear intensity and the gain constant, respectively.

[0014] Further, in step S4, the piecewise linear transformation function is:

[0015] Where (a, c) and (b, d) are the coordinates of two gray-level inflection points, f(x, y) is the gray-level function of the first-order image, representing the gray-level value of a point in the first-order image, and g(x, y) is the new gray-level function after stretching, representing the gray-level value of a point in the stretched image. M f , M g These are the maximum grayscale values ​​of the original image and the stretched image, respectively.

[0016] Furthermore, in step S5: Eight-neighbor contour extraction is performed on the secondary image based on image binarization. The eight neighbor pixels in the middle position are arranged on both sides of the gray-level threshold line. All pixels with gray levels greater than or equal to the threshold are judged as discharge beams, with a gray level of 260, and appear as white. Pixels with gray levels less than the threshold are excluded from the object area, representing the background or surrounding object area, with a gray level of 0, and appear as black. All pixels in the secondary image are stored in matrix format and assigned values ​​in row and column coordinates. If a pixel's grayscale value is 260, it means that the pixel belongs to the discharge area. The pixel is selected and an eight-neighbor counterclockwise search is performed. If the pixel's grayscale value is not 260, the pixel is determined to be a background area, it is discarded, and the process returns to review other pixels. The pixel that enters the counterclockwise search of the eight neighborhood will be searched again to determine whether the gray value of its surrounding pixels is 0. If it is 0, it means that the pixels in the neighborhood of the pixel are in the background except for itself, so the pixel is selected as part of the edge and stored in the database. If it is 1, it means that there are still other pixels in the discharge area in the neighborhood of the pixel except for itself, so it cannot be determined whether the pixel is the discharge edge, and it is necessary to return to the initial gray value search process. After multiple iterations of identification, the closed shape of the pixels that meet the criteria, presented with a grayscale value of 260, is regarded as the edge contour of the discharge.

[0017] Further, in step S5, the grayscale value range of the discharge beam is [140, 260]. The beam region is divided into three discharge regions of different brightness: I, II, and III, based on the grayscale value. Type I discharge region: grayscale value range 140-180; Class II discharge region: grayscale value range 181-220; Class III discharge region: grayscale value range 221-260.

[0018] Furthermore, the specific steps of step S7 are as follows: S71. Define the high-energy proportion coefficient k as the ratio of the area of ​​Class I to Class III discharge regions, and study the high-energy proportion of partial discharge:

[0019] in, S Ⅰ The area of ​​the Class I discharge region. S Ⅲ To be consistent with the area of ​​the Class III discharge region; S72. The complexity of partial discharge energy distribution is studied using the fractal dimension of discharge and horizontal displacement; S73. The magnitude of partial discharge energy is quantified using the corresponding discharge information in the PRPD spectrum. The energy magnitude formula is:

[0020] in, Q This represents the single discharge quantity in the PRPD spectrum. U This refers to the magnitude of the applied voltage during this discharge. Secondly, embodiments of this application also provide an analysis system for the energy distribution of partial discharge in oil, the system comprising: The experimental environment simulation module is used to construct a visualized experimental oil tank for partial discharge, simulating the real oil channel electric field environment of a converter transformer. The temperature inside the oil tank is monitored by a temperature sensor, and the pressure in the oil is monitored by a pressure sensor. The data acquisition module is used to build an experimental platform for partial discharge of oil-paper insulation, and to acquire PRPD spectrum, pulse waveform and streamer image of partial discharge in oil; The primary image preprocessing module is used to compress the streamer image through the R, G, and B channels respectively, assign weights to the optical signal image of each channel, add color restoration coefficients for multi-scale correction, and obtain the primary image. The grayscale processing and enhancement module is used to perform grayscale layering and segmented stretching on the primary image. It connects adjacent pixels with equal grayscale values ​​into closed grayscale contour lines and uses a piecewise linear transformation function to perform contrast stretching to obtain a secondary image. The contour processing and region segmentation module is used to extract the eight-neighbor contours of the secondary image through image binarization, obtain the final contour of the local discharge current beam, and divide the discharge region into Class I, Class II and Class III discharge regions with different brightness according to the gray value range of the discharge current beam. The feature parameter recording and morphology analysis module is used to record the gray area, number of branches, and horizontal displacement of all partial discharges before the insulation paper breaks down, and to analyze the relationship between the discharge current beam morphology and the discharge pulse in combination with the pulse waveform. The energy distribution analysis and quantification module is used to define the high-energy proportion coefficient k by the ratio of the area of ​​Class I and Class III discharge regions, and to analyze the high-energy proportion of partial discharge; the fractal dimension and horizontal displacement of the discharge characterize the complexity of the energy distribution; and the partial discharge energy is calculated by multiplying the single discharge quantity Q and the applied voltage U in the PRPD spectrum.

[0021] As can be seen from the above technical solutions, the present invention has the following advantages: The method and system for analyzing the energy distribution of partial discharge in oil in this application utilizes multi-scale channel-enhanced streamer image preprocessing, gray-level layering and segmented stretching image secondary processing, and eight-neighborhood contour extraction based on image binarization. This method is suitable for extracting partial discharge contours in low-brightness environments with background interference, and can solve the tomographic problem of edge point sampling. It can be used to extract most low-brightness discharge images with background interference. The method establishes the correlation between partial discharge streamers, pulse waveforms, and PRPD spectra. Based on the streamer morphology of different gray levels of partial discharge, combined with the pulse waveform of partial discharge, the energy distribution of partial discharge can be analyzed. Simultaneously, based on PRPD spectrum information, the energy magnitude of a single partial discharge can be quantified. This application reveals the mechanism of oil-paper insulation degradation under the influence of partial discharge and energy release, which is of great significance for reducing the risk of insulation faults in converter transformers. Attached Figure Description

[0022] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the analytical method for partial discharge energy distribution in oil according to this application.

[0024] Figure 2 This is a schematic diagram of an experimental oil tank used for visualizing the evolution of partial discharge in the oil, which is the method for analyzing the energy distribution of partial discharge in oil according to this application.

[0025] Figure 3 This is a schematic diagram of the oil-paper insulation partial discharge experimental platform for the analysis method of partial discharge energy distribution in oil according to this application.

[0026] Figure 4 This is a schematic diagram of the multi-scale channel enhanced streamer image preprocessing algorithm, which is the method for analyzing the partial discharge energy distribution in oil according to this application.

[0027] Figure 5 This is a grayscale histogram of the original partial discharge image used in the analysis method of partial discharge energy distribution in oil according to this application.

[0028] Figure 6 This is a grayscale histogram of a primary partial discharge image, representing the method for analyzing the energy distribution of partial discharge in oil according to this application.

[0029] Figure 7 This is a schematic diagram of the three-dimensional curved surface grayscale distribution of the original partial discharge image used in the analysis method of partial discharge energy distribution in oil according to this application. Figure 7 (a) in the image is the grayscale distribution map of the three-dimensional surface of the original image. Figure 7 (b) in the figure shows the relationship between the grayscale value and the pixel of the original image.

[0030] Figure 8 This is a schematic diagram of the three-dimensional curved surface grayscale distribution of a primary partial discharge image, used in the analysis method of partial discharge energy distribution in oil according to this application. Figure 8 (a) in the image is a three-dimensional surface grayscale distribution map of a linear image. Figure 8 (b) in the figure shows the relationship between grayscale values ​​and pixels in a single image.

[0031] Figure 9 This is a schematic diagram of the grayscale projection contour lines of the original partial discharge image used in the analysis method of partial discharge energy distribution in oil according to this application.

[0032] Figure 10 This is a schematic diagram of the grayscale projection contour lines of a primary partial discharge image, used in the analysis method for partial discharge energy distribution in oil according to this application.

[0033] Figure 11 This is a schematic diagram of the eight-neighbor contour extraction process of the binarized image used in the analysis method of partial discharge energy distribution in oil according to this application.

[0034] Figure 12 This is a schematic diagram illustrating the principle of eight-neighbor binarization for the analysis method of partial discharge energy distribution in oil in this application.

[0035] Figure 13 This is a schematic diagram illustrating the partial discharge profile extraction method applicable to low-brightness environments with background interference, which is the method for analyzing the partial discharge energy distribution in oil according to this application.

[0036] Figure 14 This is a schematic diagram of the partial discharge profile extracted by the Roberts, log, Prewitt, Sobel, and Canny algorithms used in the analysis of partial discharge energy distribution in oil according to this application. Figure 14 (a) in the image is a schematic diagram of the partial discharge profile extracted by the Roberts algorithm. Figure 14(b) in the figure is a schematic diagram of the partial discharge profile extracted by the log algorithm. Figure 14 (c) in the diagram is a schematic diagram of the partial discharge profile extracted by the Prewitt algorithm. Figure 14 (d) in the diagram is a schematic diagram of the partial discharge profile extracted by the Sobel algorithm. Figure 14 (e) in the figure is a schematic diagram of the partial discharge profile extracted by the Canny algorithm.

[0037] Figure 15 This is a schematic diagram of the partial discharge profile extracted using the analysis method for partial discharge energy distribution in oil in this application.

[0038] Figure 16 This is a schematic diagram of the grayscale region division of partial discharge in the analysis method of partial discharge energy distribution in oil according to this application.

[0039] Figure 17 This is a schematic diagram of pulse waveforms of a certain group of partial discharges, which is part of the method for analyzing the energy distribution of partial discharges in oil according to this application.

[0040] Figure 18 This is a schematic diagram of three types of gray areas corresponding to the discharge sequence in the analysis method of partial discharge energy distribution in oil according to this application. Figure 18 In the diagram, (a) represents the gray area of ​​type I. Figure 18 In the diagram, (b) represents the gray area of ​​type II. Figure 18 (c) in the figure represents the gray area of ​​type III.

[0041] Figure 19 This is a schematic diagram showing the number of discharge branches and horizontal displacement corresponding to the discharge sequence in the analysis method of partial discharge energy distribution in oil according to this application.

[0042] Figure 20 This is a schematic diagram illustrating the high-energy proportion of partial discharge in the analysis method of partial discharge energy distribution in oil presented in this application.

[0043] Figure 21 This is a schematic diagram illustrating the complexity of the partial discharge energy distribution in the analysis method for partial discharge energy distribution in oil according to this application.

[0044] Figure 22 This is a schematic diagram of the energy change of partial discharge in the analysis method of partial discharge energy distribution in oil according to this application.

[0045] Figure 23 This is a schematic diagram of the analysis system for the partial discharge energy distribution in the oil described in this application. Detailed Implementation

[0046] Various embodiments of the invention will be described more fully in the detailed steps of the analytical method based on improved partial discharge energy distribution in oil, which will be described in detail below. The invention may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the invention to the specific embodiments disclosed herein, but rather the invention should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of the invention.

[0047] It should be understood that, when used in this specification, the term "comprising" indicates the presence of the described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0048] The terms "one embodiment" or "some embodiments" used in this application mean that one or more embodiments of this application include the specific features, structures, or characteristics described in that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this application do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0049] To make the objectives, features, and advantages of this invention more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solutions protected by this invention. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0050] Please see Figure 1 The diagram shows a flowchart of a method for analyzing the energy distribution of partial discharge in oil. The method includes the following steps: S1. Construct a visual experimental oil tank for partial discharge to simulate the real oil channel electric field environment of a converter transformer. Monitor the temperature inside the oil tank using a temperature sensor and monitor the pressure in the oil using a pressure sensor. S2. Construct an oil-paper insulation partial discharge experimental platform to collect PRPD spectrum, pulse waveform, and streamer image of partial discharge in oil; S3. The streamer image is compressed using the R, G, and B channels respectively. Weights are assigned to the optical signal image of each channel, and color restoration coefficients are added for multi-scale correction to obtain a primary image. S4. Perform grayscale layering and piecewise stretching on the primary image, connect adjacent pixels with equal grayscale values ​​into closed grayscale contour lines, and use a piecewise linear transformation function to perform contrast stretching to obtain a secondary image. S5. The eight-neighbor contour is extracted by image binarization of the secondary image to obtain the final contour of the local discharge current beam, and the discharge area is divided into Class I, Class II and Class III discharge areas with different brightness according to the gray value range of the discharge current beam. S6. Record the gray area, number of branches, and horizontal displacement of all partial discharges before the insulation paper breaks down, and analyze the relationship between the discharge current beam morphology and the discharge pulse in combination with the pulse waveform. S7. Define the high-energy proportion coefficient k by the ratio of the area of ​​Class I and Class III discharge regions, and use it to analyze the high-energy proportion of partial discharge; characterize the complexity of energy distribution by the fractal dimension and horizontal displacement of the discharge; calculate the partial discharge energy by multiplying the single discharge quantity Q and the applied voltage U in the PRPD spectrum.

[0051] It should be noted that by simulating the real oil channel environment through a visualized oil tank, temperature and pressure monitoring ensures the reliability of the experiment; multi-channel image enhancement and layered processing improve the accuracy of the streamer images, and combined with partitioning and multi-parameter quantification of energy distribution, the discharge mode and energy can be accurately correlated, supporting the diagnosis and condition assessment of insulation defects in converter transformers.

[0052] As a refinement and extension of the specific implementation methods described above, and to fully illustrate the specific implementation process in this embodiment, another method for analyzing the energy distribution of partial discharge in oil is provided, including: S1. Construct a visual experimental oil tank for partial discharge to simulate the real oil channel electric field environment of a converter transformer. Monitor the temperature inside the oil tank using a temperature sensor and the pressure in the oil using a pressure sensor; such as Figure 2 As shown, the visualization experimental tank in step S1 is a fully transparent, high-mechanical-strength acrylic container with good thermal stability, light transmittance and electrical insulation. Based on the T-shaped cross oil passage at the end of the high-voltage winding of the converter transformer, the L-shaped oil passage near the electrostatic ring, and the U-shaped oil passage extending into the riser, an oil passage model of the experimental oil tank was constructed in an acrylic container using insulating cardboard. The high-voltage line and grounding line are connected to the copper foil attached to the outside of the insulating paper to simulate the real oil duct electric field environment of the converter transformer. A forced oil-guided circulation system consisting of an oil pump, oil flow sleeve, heating tube, and cooling fins is incorporated, and temperature and pressure sensors are installed.

[0053] S2. Construct an oil-paper insulation partial discharge experimental platform to collect PRPD spectrum, pulse waveform, and streamer image of partial discharge in oil; like Figure 3 As shown, in step S2, the experimental platform includes: a waveform generator, a power amplifier, a current-limiting resistor, a visualization experimental tank, a high-speed camera, a coupling capacitor, a high-voltage probe, an oscilloscope, a partial discharge instrument, and a detection impedance. The waveform generator is connected to the power amplifier, the power amplifier is connected to the current-limiting resistor, the current-limiting resistor is connected to the visualization experiment tank and the coupling capacitor, a high-speed camera is installed on the visualization experiment tank, the coupling capacitor is connected to the detection impedance and the high-voltage probe, the detection impedance is connected to the visualization experiment tank, the oscilloscope and the partial discharge instrument, and the high-voltage probe is connected to the oscilloscope.

[0054] The specific steps for acquiring the PRPD spectrum, pulse waveform, and streamer image of partial discharge in oil in step S2 are as follows: S21. Set the waveform generator to generate an electrical signal, which is then amplified by a power amplifier to obtain an experimental voltage of 0-30kV; S22. The experimental voltage is applied to the high-voltage side of the experimental oil tank; S23. Acquire the PRPD spectrum of partial discharge using a partial discharge instrument, acquire the pulse waveform of partial discharge using a high-voltage probe and oscilloscope, and acquire the streamer of partial discharge using a high-speed camera; S24. Taking into account the exposure time of the high-speed camera and the transmission delay of the coaxial cable signal, the time relationship between the optical signal and the electrical signal is corrected.

[0055] It should be noted that an electrical signal is generated by a waveform generator, and after being amplified by a power amplifier, an experimental voltage of 0-30kV is obtained. This experimental voltage is applied to the high-voltage side of the experimental tank. The partial discharge PRPD spectrum is acquired using a partial discharge instrument, the pulse waveform of the partial discharge is acquired using a high-voltage probe and an oscilloscope, and the streamers of the partial discharge are acquired using a high-speed camera. Taking into account the exposure time of the high-speed camera and the transmission delay of the coaxial cable signal, the time relationship between the optical and electrical signals is corrected to achieve synchronous acquisition of partial discharge spectrum characteristics, waveform characteristics, and streamer characteristics.

[0056] The process of steps S3-S5 is as follows: Figure 13 As shown: S3. The streamer image is compressed using the R, G, and B channels respectively. Weights are assigned to the optical signal image of each channel, and color restoration coefficients are added for multi-scale correction to obtain a primary image; such as... Figure 4 As shown, the specific steps of step S3 are as follows: S31. The streamer image is dynamically compressed using the R, G, and B channels respectively to obtain the optical signal image of each channel; S32. Assign weights to the optical signal images of each channel, add color restoration coefficients for multi-scale correction, and then re-overlay the corrected images to form a new image. The formula for multi-scale correction is:

[0057]

[0058]

[0059] In the formula, i It can be any one of the three optical channels: R, G, and B. R i ( x , y (reflection image) R In the i In each optical channel ( x , y Pixel value at point ) I i ( x , y ) is the original image in the 1st i One optical channel ( x , y The pixel value at point ); * indicates convolution operation. G j ( x , y ) is the first j Gaussian wrap function at various scales; N For the total number of scales, j For a certain count, ω j For the first j The weights corresponding to each scale; C i ( x , y ) is the first i The color correction factor for each channel is used to adjust the color ratio of the three optical channels; α and β These represent the controlled nonlinear intensity and the gain constant, respectively.

[0060] It should be noted that the original streamer image is separated into three channels: R, G, and B. The optical signal image of each channel is subjected to noise filtering and Gaussian convolution enhancement, and then weighted. After color correction, the signals are superimposed to form the preprocessed image.

[0061] S4. Perform grayscale layering and piecewise stretching on the primary image, connecting adjacent pixels with equal grayscale values ​​into closed grayscale contour lines, and then perform contrast stretching using a piecewise linear transformation function to obtain a secondary image; in step S4, the piecewise linear transformation function is:

[0062] Where (a, c) and (b, d) are the coordinates of two gray-level inflection points, f(x, y) is the gray-level function of the first-order image, representing the gray-level value of a point in the first-order image, and g(x, y) is the new gray-level function after stretching, representing the gray-level value of a point in the stretched image. M f , M g These are the maximum grayscale values ​​of the original image and the stretched image, respectively.

[0063] It should be noted that the primary image of partial discharge is processed a second time based on grayscale layering and segmented stretching to unify the measurement standard of all pixel values ​​within the layer and highlight the brightness area that reflects the discharge.

[0064] The grayscale values ​​and distribution of each pixel in the preprocessed image are statistically analyzed. The grayscale histograms of the original partial discharge image and the primary image are shown below. Figure 5 and Figure 6 As shown.

[0065] A z-coordinate axis is established based on the grayscale values ​​of each pixel in the partial discharge image to achieve a one-to-one correspondence between the grayscale distribution and the pixel coordinates of a given point. The three-dimensional surface grayscale distributions of the original partial discharge image and the primary image are shown below. Figure 7 and Figure 8 As shown, after multi-scale enhanced contrast processing, it was found that in one image, except for the discharge area where the gray value was greater than 200, the gray values ​​of most surrounding objects were limited to less than 150, clearly separating the gray levels.

[0066] Connecting adjacent points with equal gray values ​​to form a closed curve and projecting it vertically onto a horizontal plane yields a gray-level contour distribution that corresponds one-to-one with the discharge image. The gray-level projection contour distributions of the original partial discharge image and the primary image are shown below. Figure 9 and Figure 10 As shown, in a single image, the brightness of the incident light source is reduced by weakening the R, G, and B channels, thus lowering the grayscale of the background area. Consequently, the discharge separates from the surrounding area, achieving the initial formation of the discharge profile.

[0067] The discharge concentration area in the grayscale image is segmented and stretched. A suitable threshold is selected using grayscale contour lines to obtain a secondary image that reflects both the overall and local features of the image, allowing for the segmentation of discharge contours and the extraction of morphological indicators. Contrast stretching uses a piecewise linear transformation function to improve the dynamic range of grayscale levels during image processing, selectively stretching a certain grayscale range to improve the output image.

[0068] S5. The secondary image is binarized to extract the eight-neighbor contour, obtaining the final contour of the local discharge beam. The discharge region is then divided into Class I, Class II, and Class III discharge regions with different brightness levels according to the grayscale value range of the discharge beam. In step S5: Eight-neighbor contour extraction is performed on the secondary image based on image binarization. The eight neighbor pixels in the middle position are arranged on both sides of the gray-level threshold line. All pixels with gray levels greater than or equal to the threshold are judged as discharge beams, with a gray level of 260, and appear as white. Pixels with gray levels less than the threshold are excluded from the object area, representing the background or surrounding object area, with a gray level of 0, and appear as black. All pixels in the secondary image are stored in matrix format and assigned values ​​in row and column coordinates. If a pixel's grayscale value is 260, it means that the pixel belongs to the discharge area. The pixel is selected and an eight-neighbor counterclockwise search is performed. If the pixel's grayscale value is not 260, the pixel is determined to be a background area, it is discarded, and the process returns to review other pixels. The pixel that enters the counterclockwise search of the eight neighborhood will be searched again to determine whether the gray value of its surrounding pixels is 0. If it is 0, it means that the pixels in the neighborhood of the pixel are in the background except for itself, so the pixel is selected as part of the edge and stored in the database. If it is 1, it means that there are still other pixels in the discharge area in the neighborhood of the pixel except for itself, so it cannot be determined whether the pixel is the discharge edge, and it is necessary to return to the initial gray value search process. After multiple iterations of identification, the closed shape of the pixels that meet the criteria, presented with a grayscale value of 260, is regarded as the edge contour of the discharge.

[0069] It should be noted that the process for extracting the eight-neighbor contour from the secondary image based on image binarization is as follows: Figure 11 As shown. The principle of eight-neighbor binarization is as follows. Figure 12 As shown, the squares in the figure are the pixel units of the image, and each pixel is represented by P(row, column): the center pixel is P(i, j); the eight neighboring pixels around it (top, bottom, left, right, and diagonal directions) are the eight neighborhoods of P(i, j); the yellow curve is the ideal edge, and the green curve is the direction indicator of the local edge segment, used to mark the extension direction of the edge in the region.

[0070] In step S5, such as Figure 16As shown, the grayscale range of the discharge beam is [140, 260]. Based on the grayscale value, the beam region is divided into three discharge regions of different brightness: I, II, and III. Type I discharge region: grayscale value range 140-180; Class II discharge region: grayscale value range 181-220; Class III discharge region: grayscale value range 221-260.

[0071] It should be noted that five other algorithms with good performance in existing research were used to extract the contours of the discharge images, and their extraction effects were compared with those of the algorithm proposed in this invention. Figure 14 In the image, from left to right, are the partial discharge profiles extracted by the Roberts, Log, Prewitt, Sobel, and Canny algorithms, respectively. Figure 15 This is the partial discharge contour extracted in the embodiments of this application. It can be seen that the processing of the five algorithms all showed discontinuities at the irregular edges of the discharge. The main reason for the point-like discontinuities at the edges is that the brightness of the background and the discharge halo area were not suppressed or weakened. In the embodiments of this application, a partial discharge current beam contour with continuous edges and clear details was obtained.

[0072] S6. Record the gray area, number of branches, and horizontal displacement of all partial discharges before the insulation paper breaks down, and analyze the relationship between the discharge current beam morphology and the discharge pulse in combination with the pulse waveform. It should be noted that the pulse waveform of a certain group of partial discharges is as follows: Figure 17 As shown in the figure, the numbers represent the discharge order of this group of discharges. The three types of gray areas corresponding to the discharge order are as follows: Figure 18 As shown. The number of discharge branches and horizontal displacement corresponding to the discharge sequence are as follows. Figure 19 As shown in the figure, the numbers represent the number of stream branches, while those not marked represent single stream discharges.

[0073] S7. Define a high-energy proportion coefficient k by the ratio of the areas of Class I and Class III discharge regions, used to analyze the high-energy proportion of partial discharge; characterize the complexity of energy distribution by the fractal dimension and horizontal displacement of the discharge; calculate the partial discharge energy by multiplying the single discharge quantity Q and the applied voltage U in the PRPD spectrum. The specific steps of step S7 are as follows: S71. Define the high-energy proportion coefficient k as the ratio of the area of ​​Class I to Class III discharge regions, and study the high-energy proportion of partial discharge:

[0074] in, S Ⅰ The area of ​​the Class I discharge region. S Ⅲ To be consistent with the area of ​​the Class III discharge region; S72. The complexity of partial discharge energy distribution is studied using the fractal dimension of discharge and horizontal displacement; S73. The magnitude of partial discharge energy is quantified using the corresponding discharge information in the PRPD spectrum. The energy magnitude formula is:

[0075] in, Q This represents the single discharge quantity in the PRPD spectrum. U This refers to the magnitude of the applied voltage during this discharge.

[0076] It should be noted that, taking the discharge group in step S6 as an example, the high-energy proportion of partial discharge is as follows: Figure 20 As shown, the complexity of the partial discharge energy distribution is as follows: Figure 21 As shown, the change in partial discharge energy is as follows: Figure 22 As shown.

[0077] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0078] Please see Figure 23 The diagram shows a schematic of an analysis system for partial discharge energy distribution in oil. The system includes: The experimental environment simulation module is used to construct a visualized experimental oil tank for partial discharge, simulating the real oil channel electric field environment of a converter transformer. The temperature inside the oil tank is monitored by a temperature sensor, and the pressure in the oil is monitored by a pressure sensor. The data acquisition module is used to build an experimental platform for partial discharge of oil-paper insulation, and to acquire PRPD spectrum, pulse waveform and streamer image of partial discharge in oil; The primary image preprocessing module is used to compress the streamer image through the R, G, and B channels respectively, assign weights to the optical signal image of each channel, add color restoration coefficients for multi-scale correction, and obtain the primary image. The grayscale processing and enhancement module is used to perform grayscale layering and segmented stretching on the primary image. It connects adjacent pixels with equal grayscale values ​​into closed grayscale contour lines and uses a piecewise linear transformation function to perform contrast stretching to obtain a secondary image. The contour processing and region segmentation module is used to extract the eight-neighbor contours of the secondary image through image binarization, obtain the final contour of the local discharge current beam, and divide the discharge region into Class I, Class II and Class III discharge regions with different brightness according to the gray value range of the discharge current beam. The feature parameter recording and morphology analysis module is used to record the gray area, number of branches, and horizontal displacement of all partial discharges before the insulation paper breaks down, and to analyze the relationship between the discharge current beam morphology and the discharge pulse in combination with the pulse waveform. The energy distribution analysis and quantification module is used to define the high-energy proportion coefficient k by the ratio of the area of ​​Class I and Class III discharge regions, and to analyze the high-energy proportion of partial discharge; the fractal dimension and horizontal displacement of the discharge characterize the complexity of the energy distribution; and the partial discharge energy is calculated by multiplying the single discharge quantity Q and the applied voltage U in the PRPD spectrum.

[0079] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of analyzing the energy distribution of partial discharges in oil, characterized in that The method comprises the following steps: S1. Construct a visual experimental oil tank of partial discharge, simulate the real oil channel electric field environment of the converter transformer, monitor the temperature in the oil tank through a temperature sensor, and monitor the pressure in the oil through a pressure sensor; S2. Build an oil-paper insulation partial discharge experimental platform, collect the PRPD spectrum, pulse waveform and streamer image of the partial discharge in the oil; S3. Compress the streamer image through R, G and B channels respectively, assign weights to the optical signal images of each channel, add a color repair coefficient for multi-scale correction, and obtain a primary image; S4. Perform gray scale layering and segmented stretching on the primary image, connect the adjacent pixel points with equal gray values into closed gray contour lines, perform contrast stretching through a segmented linear transformation function, and obtain a secondary image; S5. Perform eight-neighborhood contour extraction of image binarization on the secondary image, obtain the final contour of the partial discharge streamer, and divide the discharge region into discharge regions of different brightness according to the gray value interval of the discharge streamer, i.e., type I, type II and type III discharge regions; S6. Record the gray area, branch number and horizontal displacement of all partial discharges before the breakdown of the insulating paper, analyze the relationship between the discharge streamer shape and the discharge pulse in combination with the pulse waveform; S7. Define a high-energy proportion coefficient k as the ratio of the areas of type I and type III discharge regions, which is used to analyze the high-energy proportion of the partial discharge; the fractal dimension and horizontal displacement of the discharge are used to represent the complexity of the energy distribution; and the product of the single discharge quantity Q and the applied voltage U in the PRPD spectrum is used to calculate the partial discharge energy.

2. The method of analyzing partial discharge energy distribution in oil according to claim 1, characterized in that, The visual experimental oil tank in step S1 adopts a transparent and high-mechanical-strength acrylic container, which has good thermal stability, light transmittance and electrical insulation; According to the T-shaped cross oil channel at the end of the high-voltage winding of the converter transformer, the L-shaped oil channel close to the electrostatic ring and the U-shaped oil channel extending into the elevated seat structure, an oil channel model of the experimental oil tank is built in the acrylic container using insulating paper boards; The high-voltage line and the grounding line are connected to the copper foil attached to the outside of the insulating paper to simulate the real oil channel electric field environment of the converter transformer; A forced oil guiding and circulating system composed of an oil pump, an oil flow sleeve, a heating pipe and a refrigeration sheet is added, and a temperature sensor and a pressure sensor are arranged.

3. The method of analyzing partial discharge energy distribution in oil according to claim 1, characterized in that, In step S2, the experimental platform comprises a waveform generator, a power amplifier, a current limiting resistor, a visual experimental oil tank, a high-speed camera, a coupling capacitor, a high-voltage probe, an oscilloscope, a partial discharge instrument and a detection impedance; The waveform generator is connected with the power amplifier, the power amplifier is connected with the current limiting resistor, the current limiting resistor is connected with the visual experimental mailbox and the coupling capacitor, the high-speed camera is installed on the visual experimental oil tank, the coupling capacitor is connected with the detection impedance and the high-voltage probe, the detection impedance is connected with the visual experimental oil tank, the oscilloscope and the partial discharge instrument, and the high-voltage probe is connected with the oscilloscope.

4. The method of analyzing partial discharge energy distribution in oil according to claim 3, characterized in that, The specific steps for collecting the PRPD spectrum, pulse waveform and streamer image of the partial discharge in the oil in step S2 are as follows: S21. Set the waveform generator to generate an electrical signal, and obtain an experimental voltage of 0-30 kV after amplification by the power amplifier; S22. Apply the experimental voltage to the high-voltage side of the experimental oil tank; S23. The partial discharge overvoltage instrument collects the PRPD spectrum of the partial discharge, the high-voltage probe and the oscilloscope collect the pulse waveform of the partial discharge, and the high-speed camera collects the streamer of the partial discharge; S24. The time relationship between the optical signal and the electrical signal is corrected by comprehensively considering the exposure time of the high-speed camera and the transmission delay factor of the coaxial cable signal.

5. The method of analyzing partial discharge energy distribution in oil according to claim 1, characterized in that, The specific steps of step S3 are: S31. The streamer image is processed by dynamic compression through R, G, and B channels respectively to obtain the optical signal image of each channel; S32. The optical signal image of each channel is weighted and color repair coefficients are added for multi-scale correction. After correction, the first image is recombined. The formula for multi-scale correction is: wherein, i R, G, B are any one of the three optical channels; R i ( x , y ) are pixel values of the reflected image R at the point of the i th optical channel x , y ; I i ( x , y ) are pixel values of the original image at the point of the i th optical channel x , y ; G j ( x , y ) is a Gaussian wrap function at the j th scale; N N is the total number of scales, j n is the scale in the current count, ω j w is the weight corresponding to the j th scale; C i ( x , y ) is the color correction factor of the i th channel, used to adjust the color ratio of the three optical channels; α and β are the controlled nonlinear strength and gain constant, respectively.

6. The method of analyzing partial discharge energy distribution in oil according to claim 1, wherein, In step S4, the piecewise linear transformation function is: Where (a, c) and (b, d) are two gray transition point coordinates, f(x, y) is a first image gray function, representing the gray value of a certain point in the first image, g(x, y) is a new gray function after stretching, representing the gray value of a certain point in the image after stretching M f , M g are the maximum gray values of the first image and the stretched image, respectively.

7. The method of analyzing partial discharge energy distribution in oil according to claim 1, wherein, In step S5: The eight-neighborhood contour extraction based on image binarization is performed on the secondary image. The eight-neighborhood pixels in the middle position are arranged on both sides of the gray threshold line. All pixels with a gray value greater than or equal to the threshold value are determined as discharge streamers, with a gray value of 260, appearing white. The pixel points with a gray value less than the threshold value are excluded from the object area, representing the background or peripheral object area, with a gray value of 0, appearing black. All pixels included in the secondary image are stored in matrix format and assigned row and column coordinates. If the gray value of a pixel is 260, it means that the pixel belongs to the discharge area, and the pixel is selected for eight-neighborhood counterclockwise retrieval. If the pixel gray value is not 260, it is determined that the pixel is in the background area and is discarded and returned to review other pixel points. The pixel points that enter the eight-neighborhood counterclockwise retrieval are re-searched to determine whether the gray values of their surrounding pixels are 0. If they are, it means that all other pixels in the neighborhood except themselves are in the background, and the pixel is selected as part of the edge and stored in the library. If they are not, it means that there are still other pixels in the discharge area in the neighborhood except themselves, and it is not possible to determine whether the pixel is a discharge edge, so the process of initially searching for the gray value needs to be restarted. After multiple cycles of identification, the closed shape of the pixel points that meet the conditions and have a gray value of 260 is considered as the edge profile of the discharge.

8. The method of analyzing partial discharge energy distribution in oil according to claim 1, characterized in that, In step S5, the gray value range of the discharge streamer is [140, 260], and the streamer area is divided into discharge areas of different brightness types I, II, and III according to the gray value, where: Type I discharge area: gray value range 140-180; Type II discharge area: gray value range 181-220; Type III discharge area: gray value range 221-260.

9. The method of analyzing partial discharge energy distribution in oil according to claim 1, wherein, The specific steps of step S7 are: S71. The ratio of the areas of type I and type III discharge areas is defined as the high-energy proportion coefficient k to study the high-energy proportion of the partial discharge: wherein S Ⅰ is the area of the class I discharge region, S Ⅲ is the area of the class III discharge region; S72. The complexity of the energy distribution of the partial discharge is studied by the fractal dimension and horizontal displacement of the discharge; S73. The size of the partial discharge energy is quantified by the corresponding discharge information in the PRPD spectrum, and the energy size formula is: wherein, Q is the single discharge quantity in the PRPD spectrum, U is the external voltage size at this discharge.

10. A system for analyzing the energy distribution of partial discharges in oil, characterized in that The system comprises: An experimental environment simulation module is used to build a visual oil tank for partial discharge, simulate the real oil channel electric field environment of a converter transformer, monitor the temperature in the oil tank through a temperature sensor, and monitor the pressure in the oil through a pressure sensor; A data acquisition module is used to build an oil-paper insulation partial discharge experimental platform, acquire the PRPD spectrum, pulse waveform and streamer image of partial discharge in oil, and acquire the streamer image of partial discharge in oil; An image primary preprocessing module is used to compress the streamer image through R, G and B channels, assign weights to the optical signal image of each channel, add a color repair coefficient for multi-scale correction, and obtain a primary image; A gray processing and enhancement module is used to perform gray layering and segmented stretching on the primary image, connect adjacent pixel points with equal gray values into closed gray contour lines, perform contrast stretching through a segmented linear transformation function, and obtain a secondary image; A contour processing and region division module is used to perform eight-neighborhood contour extraction of image binarization on the secondary image, obtain the final contour of partial discharge streamer, and divide the discharge region into discharge regions of different brightness, i.e., type I, type II and type III discharge regions, according to the gray value interval of the discharge streamer; A characteristic parameter recording and morphology analysis module is used to record the gray area, branch number and horizontal displacement of all partial discharges before the breakdown of the insulating paper, analyze the relationship between the discharge streamer morphology and the discharge pulse in combination with the pulse waveform, and record the gray area, branch number and horizontal displacement of all partial discharges before the breakdown of the insulating paper. An energy distribution analysis and quantification module is used to define a high-energy proportion coefficient k by the area ratio of type I and type III discharge regions, analyze the high-energy proportion of partial discharge, represent the energy distribution complexity by the fractal dimension and horizontal displacement of discharge, and calculate the partial discharge energy by the product of single discharge quantity Q and applied voltage U in the PRPD spectrum.

Images