A morphological feature-based interference spectrum identification method and device, electronic equipment and storage medium

Abstract

The application discloses a morphological feature-based interference spectrum identification method and device, an electronic device and a storage medium. The method and device are applied to the electronic device, specifically, original PRPD spectrum to be identified is subjected to binary processing to obtain binary PRPD spectrum; the amplitude threshold of the maximum background noise signal in the binary PRPD spectrum is calculated; the distribution curve of partial discharge pixel points greater than the maximum background noise signal is calculated based on the amplitude threshold; the phase correlation matrix and the partial discharge phase center matrix are obtained by calculation based on the distribution curve; the original PRPD spectrum is judged based on the constructed phase distribution judgment threshold and the phase correlation matrix and the partial discharge phase center matrix, and an identification result is obtained. The scheme can realize preliminary screening of the interference spectrum in the collected PRPD spectrum, thereby greatly reducing the workload of subsequent experts in analyzing the partial discharge condition based on the PRPD spectrum.

Description

Technical Field

[0001] This application relates to the field of ultra-high frequency partial discharge monitoring technology, and more specifically, to a method, apparatus, electronic device, and storage medium for interference spectrum identification based on morphological features. Background Technology

[0002] Ultra-high frequency (UHF) partial discharge monitoring technology is currently the main means of identifying the types of partial discharge defects inside GIS (Gas Insulated Switcher) equipment. In the UHF detection process, the accumulated PRPD (Phase-Resolved Partial Discharge) spectrum can obtain statistical information on the phase, discharge amplitude, and discharge frequency of the partial discharge signal, and is widely used in the process of identifying partial discharge types in GIS.

[0003] The PRPD spectrum is a two-dimensional spectrum, where the horizontal axis represents the phase information of the alternating current and the vertical axis represents the amplitude information of the received ultra-high frequency signal. The color rendering under fixed horizontal and vertical axes (phase and amplitude) represents the frequency information of the ultra-high frequency pulse signal within the statistical time of the PRPD spectrum. The morphological characteristics of the ultra-high frequency pulse signal excited by partial discharge of different physical factors are different under the statistical time. Based on the morphological characteristics of the PRPD spectrum, four partial discharge modes can be identified, namely metal particle discharge, burr tip discharge, floating electrode discharge, and insulating air gap discharge.

[0004] However, the inventors of this application discovered that in actual partial discharge detection, the signals acquired by the ultra-high frequency sensor include not only partial discharge signals but also environmental noise and interference signals within its response frequency band. Therefore, in practical applications, based on the PRPD spectrum formation mechanism, the final PRPD spectrum morphology features include not only the four typical partial discharge types but also interference noise spectra. If the interference noise spectra cannot be identified, measures cannot be taken to remove the interference, which increases the workload for experts in subsequent analysis of partial discharge conditions based on the PRPD spectrum. Summary of the Invention

[0005] In view of this, this application provides a method, apparatus, electronic device and storage medium for identifying interference patterns based on morphological features, used to identify whether the acquired PRPD pattern is an interference pattern, so as to avoid partial discharge identification failure.

[0006] To achieve the above objectives, the following solution is proposed:

[0007] An interference pattern recognition method based on morphological features, applied to electronic devices, includes the following steps:

[0008] The original PRPD map to be identified is binarized to obtain a binarized PRPD map.

[0009] Calculate the amplitude threshold of the maximum noise floor signal in the binarized PRPD map;

[0010] The distribution curve of partial discharge pixels with a noise floor greater than the maximum noise floor signal is calculated based on the amplitude threshold.

[0011] Based on the distribution curve, the phase correlation matrix and the partial discharge phase center matrix are calculated to obtain the phase correlation matrix and the partial discharge phase center matrix.

[0012] Based on the constructed phase distribution decision threshold, the phase correlation matrix, and the partial discharge phase center matrix, the original PRPD spectrum is judged to obtain the identification result.

[0013] Optionally, the step of binarizing the original PRPD map to be identified to obtain a binarized PRPD map includes the following steps:

[0014] The original PRPD map is converted into a two-dimensional grayscale image;

[0015] The grayscale threshold is obtained by calculating the two-dimensional grayscale image based on a preset grayscale threshold coefficient.

[0016] Each pixel of the two-dimensional gray value is calibrated based on the gray threshold to obtain the binarized PRPD map.

[0017] Optionally, calculating the amplitude threshold of the maximum background noise signal in the binarized PRPD spectrum includes the following steps:

[0018] Based on the binarized PRPD map, the pixels corresponding to the minimum amplitude and minimum phase in the original PRPD map are defined as basic pixels;

[0019] Based on the basic pixel points, the phase distribution characteristics of each pixel point in the original PRPD map under a fixed amplitude are statistically analyzed;

[0020] The amplitude threshold is determined by judging whether each pixel has a phase correlation with the phase distribution feature.

[0021] Optionally, the step of calculating the distribution curve of partial discharge pixels with a noise floor greater than the maximum noise floor signal based on the amplitude threshold includes the following steps:

[0022] Based on the amplitude threshold, the signal amplitude of each pixel in the original PRPD map is determined, and the set of all pixels whose signal amplitude is less than the amplitude threshold is defined as the noise set;

[0023] The pixel points in the binarized PRPD map that correspond to the noise set are assigned a value of 0 to obtain the denoised pixel set.

[0024] The 0 and 1 values ​​in the denoised pixel set are accumulated according to the 0-360° phase to obtain the distribution curve.

[0025] Optionally, the step of calculating the phase correlation matrix and the partial discharge phase center matrix based on the distribution curve includes the following steps:

[0026] Using 0 as a threshold, the distribution curve is sequentially searched for multiple partial discharge pixel distribution values ​​in the 0-360° phase coordinate system. The partial discharge pixel points whose previous partial discharge pixel distribution value and their next partial discharge pixel distribution value are both 0 or not both greater than 0 are determined as the set definition points.

[0027] The phase correlation matrix and the partial discharge behavior center matrix are obtained by calculating based on all the defined points of the set.

[0028] Optionally, based on the constructed phase distribution decision threshold, the phase correlation matrix, and the partial discharge phase center matrix, the original PRPD map is judged to obtain the identification result, including the following steps:

[0029] Based on the distribution criterion threshold, the phase correlation matrix and the partial discharge phase center matrix are used to obtain the modified phase correlation matrix and the modified partial discharge phase center matrix.

[0030] The original PRPD map is identified based on the modified phase correlation matrix and the modified partial discharge phase center matrix to obtain the identification result.

[0031] Optionally, the step of identifying the original PRPD map based on the corrected phase correlation matrix and the corrected partial discharge phase center matrix to obtain the identification result includes the following steps:

[0032] The corresponding data in the modified phase correlation matrix and the modified partial discharge phase center matrix are calculated to obtain the data to be identified.

[0033] The data to be identified is judged based on the threshold to be identified, and the original PRPD map of the data to be identified that meets the threshold to be identified is identified as an interference map.

[0034] An interference pattern recognition device based on morphological features, applied to electronic devices, wherein the interference pattern recognition method includes:

[0035] The binarization processing module is configured to binarize the original PRPD map to be identified to obtain a binarized PRPD map.

[0036] The threshold calculation module is configured to calculate the amplitude threshold of the maximum background noise signal in the binarized PRPD spectrum;

[0037] The curve calculation module is configured to calculate the distribution curve of partial discharge pixels that are greater than the maximum noise floor signal based on the amplitude threshold.

[0038] The matrix calculation module is configured to calculate based on the distribution curve to obtain the phase correlation matrix and the partial discharge phase center matrix;

[0039] The recognition execution module is configured to judge the original PRPD spectrum based on the constructed phase distribution decision threshold, the phase correlation matrix, and the partial discharge phase center matrix, and obtain the recognition result.

[0040] An electronic device includes at least one processor and a memory connected to the processor, wherein:

[0041] The memory is used to store computer programs or instructions;

[0042] The processor is used to execute the computer program or instructions to enable the electronic device to implement the interference pattern recognition method as described above.

[0043] A computer-readable storage medium is applied to an electronic device, the storage medium carrying one or more computer programs that can be executed by the electronic device, thereby enabling the electronic device to implement the interference pattern identification method as described above.

[0044] As can be seen from the above technical solution, this application discloses a method, apparatus, electronic device, and storage medium for interference spectrum identification based on morphological features. The method and apparatus are applied to electronic devices, specifically involving binarizing the original PRPD spectrum to be identified to obtain a binarized PRPD spectrum; calculating the amplitude threshold of the maximum background noise signal in the binarized PRPD spectrum; calculating the distribution curve of partial discharge pixels greater than the maximum background noise signal based on the amplitude threshold; calculating the phase correlation matrix and the partial discharge phase center matrix based on the distribution curve; and judging the original PRPD spectrum based on the constructed phase distribution decision threshold, phase correlation matrix, and partial discharge phase center matrix to obtain the identification result. This solution can achieve preliminary screening of interference spectra within the acquired PRPD spectrum, thereby greatly reducing the workload of subsequent expert analysis of partial discharge conditions based on PRPD spectra.

[0045] In addition, this application identifies interference maps based on morphological features, without using machine learning and deep learning methods, but only using expert knowledge-driven methods, which greatly saves the cost of interference map data collection and labeling in the dataset production process; finally, the algorithm utilizes the statistical information of a simple PRPD two-dimensional image matrix, with low algorithm complexity and no need to occupy the hardware and software resources of the partial discharge equipment. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of an interference pattern recognition method according to an embodiment of this application;

[0048] Figure 2a A schematic diagram of a PRPD pattern for a burr tip;

[0049] Figure 2b This is a schematic diagram of the PRPD pattern of a metal particle.

[0050] Figure 2c This is a schematic diagram of the PRPD pattern of an insulation defect.

[0051] Figure 2d This is a schematic diagram of the PRPD pattern of a suspended electrode.

[0052] Figure 2e This is a schematic diagram of an interference spectrum;

[0053] Figure 3a A schematic diagram of the partial discharge pixel distribution curve after noise reduction of a PRPD map of a burr tip;

[0054] Figure 3b A schematic diagram of the partial discharge pixel distribution curve after noise reduction of the PRPD spectrum of a metal particle.

[0055] Figure 3c A schematic diagram of the partial discharge pixel distribution curve after noise reduction of the PRPD spectrum of an insulation defect;

[0056] Figure 3d This is a schematic diagram of the partial discharge pixel distribution curve after noise reduction of the PRPD spectrum of a floating electrode.

[0057] Figure 3eThis is a schematic diagram of the partial discharge pixel distribution curve after noise reduction of an interference spectrum.

[0058] Figure 4 This is a block diagram of an interference pattern recognition device according to an embodiment of this application;

[0059] Figure 5 This is a block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0060] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0061] The inventors of this application discovered through research that existing UHF partial discharge (PD) devices typically employ two methods to address environmental noise interference during detection. One method involves using signal processing to physically eliminate noise interference signals, ensuring that the resulting PRPD spectrum does not contain interference patterns. The other method involves constructing a rich dataset of interference patterns and training a feature extractor using machine learning to directly identify PRPD interference patterns. However, completely eliminating noise signals using signal processing methods requires manual intervention, and machine learning methods struggle to construct the diverse datasets of interference patterns needed for spectrum recognition. To address these difficulties, this invention, based on the image features of PRPD spectra, identifies four typical PD defects with more specific morphological distribution patterns than interference patterns. By summarizing the differences between interference patterns and the four typical PD defects, and drawing on expert knowledge decision-making methods, a data-free, morphological feature-based PRPD interference pattern recognition method is designed. Based on these ideas, this application proposes the following specific implementation methods.

[0062] Figure 1 This is a flowchart of an interference pattern identification method according to an embodiment of this application.

[0063] like Figure 1 As shown, the interference image recognition method provided in this embodiment is applied to an electronic device to identify PRPD maps collected by a UHF sensor from a GIS device, thereby identifying interference maps. This electronic device can be understood as a computer, server, cloud platform, or embedded device with data computing and information processing capabilities. The interference map recognition method specifically includes the following steps:

[0064] S1. Binarize the original PRPD map to obtain a binarized PRPD map.

[0065] The original PRPD image is binarized to obtain a binarized PRPD image. The original PRPD image can be represented as a three-dimensional matrix h×w×3, where h is the total number of pixels corresponding to the amplitude of the image, w is the total number of pixels corresponding to the phase of the image, and 3 represents the three RGB channels. The original PRPD image is then converted from RGB three-channel to a two-dimensional grayscale image G_PRPD, represented as a two-dimensional matrix h×w. The grayscale threshold G of the original PRPD image is calculated by inputting the grayscale threshold coefficient α. th And by assigning values ​​less than G to the two-dimensional grayscale image th All pixels are defined as 0, otherwise as 1, thus achieving the binarization of the PRPD map. The binarized PRPD map can be represented as B_PRPD.

[0066] This invention converts the original PRPD image into a grayscale image. Since the background grayscale of the image is uniform, while the grayscale values ​​of the signal pixels are significantly different from the background grayscale due to the different discharge frequencies of the pulse signal at different phase amplitude positions, a grayscale threshold coefficient is set and a grayscale threshold is calculated. The grayscale value information is used to divide the pixels in the original PRPD image into two parts: background and signal pixels. This facilitates the subsequent calculation and statistics of the morphological characteristics of the signal pixels.

[0067] In one specific embodiment of this application, binarization is achieved using the following steps:

[0068] First, obtain the three-dimensional matrix I_PRPD ∈ h×w×3 of the original PRPD map, and then use the averaging method to obtain the gray-level matrix G_PRPD of the map. The specific calculation method is shown in the following formula:

[0069]

[0070] Where x ranges from [1, w], y ranges from [1, h], and I_PRPD_R, ​​I_PRPD_G and I_PRPD_B represent the grayscale values ​​of the three RGB channels, respectively.

[0071] Then, based on the input grayscale threshold coefficient α, if the grayscale value of the background pixel in the original PRPD map is G... b Then the grayscale threshold of the image can be expressed as: G th =G b *α. In this application, α is defined as having a specific value of 0.1, G b The specific value is 200. The grayscale value range of the pixels in the entire original PRPD map is [0, 255].

[0072] Finally, based on the calculated spectral threshold G th Define the binary interval of the graph as [G]. b-G th G b +G th ], the pixel grayscale range within the original PRPD map is set to [G b -G th G b +G th The pixel positions within the specified range are defined as background and marked as 0, while the remaining pixels are marked as 1, representing partial discharge signals, background noise, or interference pixels, thus obtaining the binarized PRPD map B_PRPD. This map is a two-dimensional matrix with r rows and c columns, and the matrix size is represented as r*c. The values ​​within the matrix are either 0 or 1.

[0073] S2. Calculate the amplitude threshold of the maximum noise floor signal in the binarized PRPD spectrum.

[0074] The amplitude threshold n of the maximum noise floor signal is calculated based on the binarized PRPD map B_PRPD. Specifically, based on this B_PRPD, the pixel coordinates corresponding to the minimum amplitude and minimum phase in the original PRPD map are defined as (0, 0). The phase distribution characteristics of the PRPD partial discharge pixels in the original PRPD map under a fixed amplitude are statistically analyzed. The amplitude threshold n of the maximum noise floor signal is determined by judging whether the partial discharge pixels under a fixed amplitude have phase correlation.

[0075] Within four typical PRPD partial discharge and PRPD interference maps, the phase distribution of the noise floor signal is uncorrelated, meaning the signal's appearance is independent of phase. Utilizing this expert prior, a noise floor pixel discrimination method was designed. Based on the physical prior of a uniform phase distribution of noise floor pixels within the map (0-360°), the amplitude of the signal pixel distribution with uniform phase characteristics after fixing the signal amplitude is defined as the noise floor amplitude. The largest noise floor amplitude is identified as the amplitude threshold n of this application.

[0076] In one specific embodiment of this application, the above-mentioned amplitude threshold n is calculated using the following steps:

[0077] First, the value under the (i,j) coordinates of the binarized PRPD map is directly located as g. When (i,j) is a partial discharge pixel, g = 1; otherwise, g = 0.

[0078] Then, the pixel magnitude coordinates j that satisfy the following conditions are denoted as the set N = [n1, n2, n3, ..., n N ]. g(i,j)=1,j=1,2,3,...,w.

[0079] Finally, the maximum value in set N is defined as the amplitude threshold n.

[0080] S3. Calculate the distribution curve of partial discharge pixels that are greater than the maximum noise floor signal based on the amplitude threshold.

[0081] Based on the amplitude threshold n of the maximum noise floor signal, calculate the partial discharge pixel distribution curve I for values ​​greater than this amplitude threshold. fb Specifically, the set of locations of partial discharge (PD) pixels in the original PRPD image where the signal amplitude is less than the amplitude threshold n is defined as N_noise. Then, all position information values ​​in the N_noise set corresponding to the binarized PRPD image are assigned to 0, resulting in the B_PRPD_qz matrix. At this point, the positions with a value of 1 in the B_PRPD_qz matrix represent the PD signal after noise removal. The 0 and 1 values ​​in B_PRPD_qz are accumulated according to the 0–360° phase to obtain the denoised PD pixel distribution curve I. fb Its array dimension can be represented as w×1.

[0082] Since both the PRPD partial discharge spectrum and the interference spectrum contain background noise information, the background noise pixels within the spectrum are not considered valid morphological features in the identification process. Therefore, this patent, based on the amplitude threshold n obtained in step S2, sets all background noise pixels without phase correlation within the spectrum as background pixels of the PRPD spectrum. The remaining pixels are then accumulated according to their phase coordinates, and under specific phase coordinate values, the distribution curve I... fb The larger the value, the richer the amplitude of the pulse signal gathered under that phase.

[0083] The above distribution curve I fb Specifically, it is calculated using the following formula:

[0084]

[0085] Where j represents the j-th phase coordinate, i represents the i-th signal amplitude coordinate, and g(i, j) is the specific value of the position coordinate (i, j) in the B_PRPD_qz matrix.

[0086] S4. Based on the distribution curve, calculate the phase correlation matrix and the partial discharge phase center matrix.

[0087] That is, the distribution curve I based on partial discharge pixels fb Calculate the phase correlation coefficient matrix A and the partial discharge phase center matrix C of the original PRPD image. I is the partial discharge pixel distribution curve after removing the noise floor from the original PRPD image. fb If the value is greater than or equal to 0, use 0 as the threshold to sequentially search for I in the 0-360° phase coordinate range. fb The curve contains w partial discharge pixel distribution values. For the i-th phase coordinate, when the previous I... fb Pixel value and the next I fbIf the pixel values ​​are not all 0 at the same time, or not all of them are greater than 0 at the same time, record the i-th phase coordinate. Define P as the set of all statistically analyzed phase coordinates. Calculate the phase correlation matrix A and the partial discharge phase center matrix C based on P.

[0088] This application uses a phase correlation matrix A to statistically analyze the phase distribution information of signal pixels within the original PRPD spectrum and calculate the information of continuous phase distribution in the spectrum. Specifically, this application divides the phase within the spectrum into multiple continuous phase segments, storing the phase segment information in the phase correlation matrix A, and storing the phase center value corresponding to each phase segment in the partial discharge phase center matrix C. Signal pulses are detected at each phase coordinate in the selected phase segment. The phases of typical four types of partial discharge PRPD spectra exhibit unipolar or bipolar correlation. If bipolar, the centers of two continuous phase segments differ by 180°. Therefore, by calculating matrices A and C, it can be determined whether the spectrum is a PRPD partial discharge spectrum. If it does not conform to the characteristics of partial discharge spectra, the PRPD spectrum can be identified as an interference spectrum.

[0089] In one specific embodiment of this application, the phase correlation coefficient matrix A and the partial discharge phase center matrix C are calculated using the following steps:

[0090] First, based on the pixel distribution curve I after removing the background noise within the original PRPD map. fb Based on the phase coordinates [1,2,...,w], determine the corresponding I values ​​sequentially. fb The value is defined as the set matrix of phase coordinates that satisfies the following conditions.

[0091] 1) Input I fb The value I at the first phase coordinate fb (1), determine I fb (1) Whether it is greater than 0. If it is greater than 0, then the phase coordinate is the first element in matrix P; if I fb (1) If it equals 0, then continue to judge I. fb (2), until the first I greater than 0 is found. fb If the value of (m) is given, then the phase coordinate m is the first element in P.

[0092] 2) Sequentially determine I under phase coordinates [m+1, m+2, ... w] fb Does the value satisfy the following condition, i.e., the previous I? fb Pixel value and the next I fb The pixel values ​​are not all 0 at the same time, or not all greater than 0 at the same time. The phase coordinates that meet this condition are stored sequentially in P to obtain the phase coordinate matrix P.

[0093] if I fb (i-1)=0and I fb(i+1)>0i∈[m+1,....,w-1)

[0094] if I fb (i-1)>0and I fb (i+1) = 0, i ∈ [m+1,....,w-1]

[0095] Then, based on the calculated phase coordinate matrix P and I fb The calculation methods for the phase correlation coefficient matrix A and the partial discharge phase center matrix C are as follows.

[0096] 1) Define a phase coordinate matrix P with Np phase coordinate samples. Determine the phase coordinate value I between the i-th phase coordinate P(i) and P(i+1). fb Does the value satisfy the condition of being greater than 0? As shown in the following formula, where i∈[1,Np-1].

[0097] if I fb (j)>0j∈[P(i),P(i+1)]

[0098] 2) Combine the phase coordinates that satisfy the above. i = {P(i), P(i+1)} are recorded. If there are m combinations that satisfy the conditions, then the combination matrix is ​​represented as [a1, a2, ..., a... m ].

[0099] 3) Based on the computation of the combination matrix [a1, a2, ..., a] m The correlation coefficient matrix A and the signal phase center matrix C are calculated based on the elements in the combined matrix. The i-th element A in A is... i The calculation method is based on the i-th coordinate combination a in the combination matrix. i = {P(i), P(i+1)}. The specific calculation method is as follows:

[0100] A i =(P(i+1)-P(i)) / w

[0101] The i-th element C in the partial discharge phase center matrix C i The calculation method is based on the i-th coordinate combination a in the combination matrix. i = {P(i), P(i+1)}. The specific calculation method is as follows:

[0102] C i = (P(i+1)-P(i)) / 2

[0103] S5. The original PRPD spectrum is judged based on the phase correlation matrix and the partial discharge phase center matrix to obtain the identification result.

[0104] A threshold for phase distribution in a PRPD interference pattern is constructed. Based on the phase correlation matrix A and the partial discharge phase center matrix C, it is determined whether the pattern belongs to the interference pattern. Scattered phases in the phase correlation matrix A and the partial discharge phase center matrix C are removed to obtain a modified phase correlation matrix A' and a modified phase center matrix C'. The total number of correlated phase distributions in A' is calculated as N_A', and the difference matrix between phase centers in C' is calculated as D_C'. When N_A' and D_C' satisfy the set threshold conditions, the pattern is considered an interference pattern.

[0105] Since the phase correlation matrix A above contains all continuous phase segments of signal pixels within the PRPD spectrum, when the signal within the spectrum exhibits both aggregated and discrete characteristics, phase correlation matrix A will inevitably contain phase segments with extremely small phase differences. To avoid the influence of discrete features on phase statistics, this application uses a threshold to filter out extremely small phase differences in phase correlation matrix A, resulting in a corrected phase correlation matrix A' and a corrected phase center matrix C'. By determining the number of phase segments in matrix A' and the phase difference between the centers of each phase segment in matrix C', PRPD interference spectrum identification can be achieved.

[0106] In one specific embodiment of this application, the calculation process for the two matrices is implemented using the following steps:

[0107] First, given that the phase correlation coefficient matrix A and the partial discharge phase center matrix C both contain m elements, each element in matrix A represents the degree of clustering of phase correlations. The larger the value of the i-th element in matrix A, the stronger the corresponding coordinate combination a. i The larger the phase span, the more this application sets a threshold A. th =0.1, to set the matrix A to the nearest integer A. th Similarly, the corresponding elements in the corresponding signal phase center matrix C are also removed.

[0108] After phase correction, the elements at the same search position in matrices A' and C' correspond one-to-one, physically representing the length of the clustered phase segment of the searched signal pixels and the position of the center pixel, respectively.

[0109] The process of identifying interference patterns is as follows:

[0110] First, obtain matrices A' and C'. Since both matrices have the same number of elements, denote them as N_A'. Because the phase of typical four types of partial discharge patterns clusters into unipolar and bipolar distributions, the value of N_A' is either 1 or 2.

[0111] Then, considering phase cyclicity, the physical meaning of 0° phase and 360° phase is the same. Therefore, when the phase range of signal pixel aggregation is near 0°, the signal phase aggregation cluster may be divided into two clusters, one half of which is distributed in the first half of the spectrum and the other half of which is distributed in the second half of the spectrum. In this special case, the value of N_A' in the unipolar partial discharge spectrum is equal to 2, and the value of N_A' in the bipolar spectrum is equal to 3.

[0112] For the PRPD map to be identified, its N_A' is calculated. When N_A' is greater than 3, the map is determined to be an interference map. When N_A' is less than or equal to 3, the interference map is further subdivided as follows:

[0113] 1) When N_A' equals 1, the PRPD pattern is a partial discharge pattern; otherwise, it is an interference pattern.

[0114] 2) When N_A' equals 2, calculate the difference D_C' between the two elements in the phase center matrix C'. There is only one difference number. When the phase difference number is approximately equal to w / 2 or w, the PRPD spectrum is a partial discharge spectrum; otherwise, it is an interference spectrum.

[0115] 3) When N_A' equals 3, calculate the difference D_C' between each pair of elements in the phase center matrix C'. When the difference between C1 and C2 is approximately equal to w / 2, and the difference between C1 and C3 is approximately equal to w, the PRPD spectrum is a partial discharge spectrum; otherwise, it is an interference spectrum.

[0116] As can be seen from the above technical solution, this embodiment provides a method for identifying interference patterns based on morphological features. This method is applied to electronic devices. Specifically, it involves binarizing the original PRPD pattern to be identified to obtain a binarized PRPD pattern; calculating the amplitude threshold of the maximum background noise signal in the binarized PRPD pattern; calculating the distribution curve of partial discharge pixels greater than the maximum background noise signal based on the amplitude threshold; calculating the phase correlation matrix and the partial discharge phase center matrix based on the distribution curve; and judging the original PRPD pattern based on the constructed phase distribution decision threshold, phase correlation matrix, and partial discharge phase center matrix to obtain the identification result. This solution can achieve preliminary screening of interference patterns within the acquired PRPD pattern, thereby greatly reducing the workload of subsequent expert analysis of partial discharge conditions based on the PRPD pattern.

[0117] In addition, this application identifies interference maps based on morphological features, without using machine learning and deep learning methods, but only using expert knowledge-driven methods, which greatly saves the cost of interference map data collection and labeling in the dataset production process; finally, the algorithm utilizes the statistical information of a simple PRPD two-dimensional image matrix, with low algorithm complexity and no need to occupy the hardware and software resources of the partial discharge equipment.

[0118] The inventors of this application have implemented simulation calculations based on the above-mentioned technical solution.

[0119] Figures 2a-2e These are grayscale images of the five original PRPD maps to be identified in this application. Figure 2a PRPD pattern of the burr tip. Figure 2b This is the PRPD spectrum of the metal particles. Figure 2c From the PRPD spectrum of insulation defects. Figure 2d The PRPD spectrum of the suspended electrode. Figure 2e The image shows the interference spectrum. As can be seen from the figure, the signal phase distribution of the partial discharge PRPD spectrum exhibits unipolar or bipolar convergence characteristics, while the interference spectrum does not.

[0120] Figures 3a-3e To be respectively with Figures 2a-2e The Ifb curve, which corresponds to the original PRPD spectrum after noise reduction, is shown in the figure. After noise reduction, the corresponding value of the Ifb curve in the signal pixel distribution is greater than 0.

[0121] The table below shows the phase correlation matrix A' and the corrected phase center matrix C' calculated from the five original PRPD maps after correction.

[0122] (a) [0.16]

[36] (b) [0.968]

[60] (c) [0.272,0.312] [19,83] (d) [0.344,0.504] [36,94] (e) null null

[0123] In (a) and (b), if N_A' equals 1, the spectrum to be identified is directly determined to be a PRPD partial discharge spectrum. In (c) and (d), if the corresponding spectrum to be identified has N_A' equal to 2, in (c), the difference D_C' between the two elements in the phase center matrix C' is 64, which is approximately equal to 125 / 2, so the spectrum to be identified is a partial discharge spectrum. Similarly, in (d), the difference D_C' between the two elements in the phase center matrix C' is 58, which is approximately equal to 125 / 2, so the spectrum to be identified is a partial discharge spectrum. After correction, all phases of the spectrum to be identified (e) do not meet the threshold condition, so the correlation matrix after correction is empty, and therefore it is directly determined to be an interference spectrum.

[0124] Although the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous.

[0125] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0126] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer.

[0127] Figure 4 This is a flowchart of an interference pattern identification method according to an embodiment of this application.

[0128] like Figure 4 As shown, the interference image recognition device provided in this embodiment is applied to an electronic device to identify PRPD maps collected by a UHF sensor from a GIS device, thereby identifying interference maps. This electronic device can be understood as a computer, server, cloud platform, or embedded device with data computing and information processing capabilities. Specifically, the interference map recognition device includes a binarization processing module 10, a threshold calculation module 20, a curve calculation module 30, a matrix calculation module 40, and a recognition execution module 50.

[0129] The binarization module is used to binarize the original PRPD map to obtain a binarized PRPD map.

[0130] The original PRPD image is binarized to obtain a binarized PRPD image. The original PRPD image can be represented as a three-dimensional matrix h×w×3, where h is the total number of pixels corresponding to the amplitude of the image, w is the total number of pixels corresponding to the phase of the image, and 3 represents the three RGB channels. The original PRPD image is then converted from RGB three-channel to a two-dimensional grayscale image G_PRPD, represented as a two-dimensional matrix h×w. The grayscale threshold G of the original PRPD image is calculated by inputting the grayscale threshold coefficient α. th And by assigning values ​​less than G to the two-dimensional grayscale image th All pixels are defined as 0, otherwise as 1, thus achieving the binarization of the PRPD map. The binarized PRPD map can be represented as B_PRPD.

[0131] This invention converts the original PRPD image into a grayscale image. Since the background grayscale of the image is uniform, while the grayscale values ​​of the signal pixels are significantly different from the background grayscale due to the different discharge frequencies of the pulse signal at different phase amplitude positions, a grayscale threshold coefficient is set and a grayscale threshold is calculated. The grayscale value information is used to divide the pixels in the original PRPD image into two parts: background and signal pixels. This facilitates the subsequent calculation and statistics of the morphological characteristics of the signal pixels.

[0132] The threshold calculation module is used to calculate the amplitude threshold of the maximum noise floor signal in the binarized PRPD spectrum.

[0133] The amplitude threshold n of the maximum noise floor signal is calculated based on the binarized PRPD map B_PRPD. Specifically, based on this B_PRPD, the pixel coordinates corresponding to the minimum amplitude and minimum phase in the original PRPD map are defined as (0, 0). The phase distribution characteristics of the PRPD partial discharge pixels in the original PRPD map under a fixed amplitude are statistically analyzed. The amplitude threshold n of the maximum noise floor signal is determined by judging whether the partial discharge pixels under a fixed amplitude have phase correlation.

[0134] Within four typical PRPD partial discharge and PRPD interference maps, the phase distribution of the noise floor signal is uncorrelated, meaning the signal's appearance is independent of phase. Utilizing this expert prior, a noise floor pixel discrimination method was designed. Based on the physical prior of a uniform phase distribution of noise floor pixels within the map (0-360°), the amplitude of the signal pixel distribution with uniform phase characteristics after fixing the signal amplitude is defined as the noise floor amplitude. The largest noise floor amplitude is identified as the amplitude threshold n of this application.

[0135] The curve calculation module is used to calculate the distribution curve of partial discharge pixels that are greater than the maximum noise floor signal based on the amplitude threshold.

[0136] Based on the amplitude threshold n of the maximum noise floor signal, calculate the partial discharge pixel distribution curve I for values ​​greater than this amplitude threshold. fb Specifically, the set of locations of partial discharge (PD) pixels in the original PRPD image where the signal amplitude is less than the amplitude threshold n is defined as N_noise. Then, all position information values ​​in the N_noise set corresponding to the binarized PRPD image are assigned to 0, resulting in the B_PRPD_qz matrix. At this point, the positions with a value of 1 in the B_PRPD_qz matrix represent the PD signal after noise removal. The 0 and 1 values ​​in B_PRPD_qz are accumulated according to the 0–360° phase to obtain the denoised PD pixel distribution curve I. fb Its array dimension can be represented as w×1.

[0137] Since both the PRPD partial discharge spectrum and the interference spectrum contain background noise information, the background noise pixels within the spectrum are not considered valid morphological features in the identification process. Therefore, this patent, based on the amplitude threshold n obtained in step S2, sets all background noise pixels without phase correlation within the spectrum as background pixels of the PRPD spectrum. The remaining pixels are then accumulated according to their phase coordinates, and under specific phase coordinate values, the distribution curve I... fb The larger the value, the richer the amplitude of the pulse signal gathered under that phase.

[0138] The matrix calculation module is used to perform calculations based on the distribution curve to obtain the phase correlation matrix and the partial discharge phase center matrix.

[0139] That is, the distribution curve I based on partial discharge pixels fb Calculate the phase correlation coefficient matrix A and the partial discharge phase center matrix C of the original PRPD image. I is the partial discharge pixel distribution curve after removing the noise floor from the original PRPD image. fb If the value is greater than or equal to 0, use 0 as the threshold to sequentially search for I in the 0-360° phase coordinate range. fb The curve contains w partial discharge pixel distribution values. For the i-th phase coordinate, when the previous I... fb Pixel value and the next I fb If the pixel values ​​are not all 0 at the same time, or not all of them are greater than 0 at the same time, record the i-th phase coordinate. Define P as the set of all statistically analyzed phase coordinates. Calculate the phase correlation matrix A and the partial discharge phase center matrix C based on P.

[0140] This application uses a phase correlation matrix A to statistically analyze the phase distribution information of signal pixels within the original PRPD spectrum and calculate the information of continuous phase distribution in the spectrum. Specifically, this application divides the phase within the spectrum into multiple continuous phase segments, storing the phase segment information in the phase correlation matrix A, and storing the phase center value corresponding to each phase segment in the partial discharge phase center matrix C. Signal pulses are detected at each phase coordinate in the selected phase segment. The phases of typical four types of partial discharge PRPD spectra exhibit unipolar or bipolar correlation. If bipolar, the centers of two continuous phase segments differ by 180°. Therefore, by calculating matrices A and C, it can be determined whether the spectrum is a PRPD partial discharge spectrum. If it does not conform to the characteristics of partial discharge spectra, the PRPD spectrum can be identified as an interference spectrum.

[0141] The recognition execution module is used to judge the original PRPD spectrum based on the phase correlation matrix and the partial discharge phase center matrix to obtain the recognition result.

[0142] A threshold for phase distribution in a PRPD interference pattern is constructed. Based on the phase correlation matrix A and the partial discharge phase center matrix C, it is determined whether the pattern belongs to the interference pattern. Scattered phases in the phase correlation matrix A and the partial discharge phase center matrix C are removed to obtain a modified phase correlation matrix A' and a modified phase center matrix C'. The total number of correlated phase distributions in A' is calculated as N_A', and the difference matrix between phase centers in C' is calculated as D_C'. When N_A' and D_C' satisfy the set threshold conditions, the pattern is considered an interference pattern.

[0143] Since the phase correlation matrix A above contains all continuous phase segments of signal pixels within the PRPD spectrum, when the signal within the spectrum exhibits both aggregated and discrete characteristics, phase correlation matrix A will inevitably contain phase segments with extremely small phase differences. To avoid the influence of discrete features on phase statistics, this application uses a threshold to filter out extremely small phase differences in phase correlation matrix A, resulting in a corrected phase correlation matrix A' and a corrected phase center matrix C'. By determining the number of phase segments in matrix A' and the phase difference between the centers of each phase segment in matrix C', PRPD interference spectrum identification can be achieved.

[0144] As can be seen from the above technical solution, this embodiment provides an interference spectrum identification device based on morphological features. This device is applied to electronic devices. Specifically, it binarizes the original PRPD spectrum to be identified to obtain a binarized PRPD spectrum; calculates the amplitude threshold of the maximum background noise signal in the binarized PRPD spectrum; calculates the distribution curve of partial discharge pixels greater than the maximum background noise signal based on the amplitude threshold; calculates the phase correlation matrix and the partial discharge phase center matrix based on the distribution curve; and judges the original PRPD spectrum based on the constructed phase distribution decision threshold, phase correlation matrix, and partial discharge phase center matrix to obtain the identification result. This solution can achieve preliminary screening of interference spectra within the acquired PRPD spectrum, thereby greatly reducing the workload of subsequent expert analysis of partial discharge status based on PRPD spectrum.

[0145] The units described in the embodiments of this disclosure can be implemented in software or in hardware. The name of a unit does not necessarily limit the unit itself; for example, the first acquisition unit can also be described as "a unit that acquires at least two Internet Protocol addresses".

[0146] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0147] Figure 5 This is a block diagram of an electronic device according to an embodiment of this application.

[0148] The following is for reference. Figure 5 This document illustrates a structural diagram suitable for implementing the electronic device in the embodiments of this disclosure. The terminal device in the embodiments of this disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. This electronic device is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this disclosure.

[0149] The electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from an input device 506 into a random access memory (RAM) 503. The RAM also stores various programs and data required for the operation of the electronic device. The processing unit, ROM, and RAM are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0150] Typically, the following devices can be connected to the I / O interface: input devices including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows the electronic device to communicate wirelessly or wiredly with other devices to exchange data. Although electronic devices with various devices are shown in the figures, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0151] This application also provides an embodiment of a computer-readable storage medium.

[0152] The aforementioned computer-readable storage medium is applied to an electronic device and carries one or more programs. When these programs are executed by the electronic device, the device binarizes the original PRPD image to be identified, obtaining a binarized PRPD image; calculates the amplitude threshold of the maximum noise floor signal in the binarized PRPD image; calculates the distribution curve of partial discharge pixels greater than the maximum noise floor signal based on the amplitude threshold; calculates the phase correlation matrix and the partial discharge phase center matrix based on the distribution curves; and judges the original PRPD image based on the constructed phase distribution decision threshold, phase correlation matrix, and partial discharge phase center matrix to obtain the identification result. This scheme can achieve preliminary screening of interference patterns within the acquired PRPD image, thereby greatly reducing the workload of subsequent expert analysis of partial discharge conditions based on the PRPD image.

[0153] It should be noted that the computer-readable medium described above in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0154] In this disclosure, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0155] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0156] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0157] The technical solution provided by the present invention has been described in detail above. Specific examples have been used to illustrate the principle and implementation of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation and application scope based on the idea of ​​the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying interference maps based on morphological features, applied to electronic devices, characterized in that, The interference pattern identification method includes the following steps: The original PRPD map to be identified is binarized to obtain a binarized PRPD map. The amplitude threshold of the maximum noise floor signal in the binarized PRPD map is calculated by defining the pixel corresponding to the minimum amplitude and minimum phase in the original PRPD map as the base pixel based on the binarized PRPD map, statistically analyzing the phase distribution characteristics of each pixel in the original PRPD map under a fixed amplitude based on the base pixel, and determining the amplitude threshold by judging whether each pixel has a phase correlation with the phase distribution characteristics. The distribution curve of partial discharge pixels with a noise floor greater than the maximum noise floor signal is calculated based on the amplitude threshold. Based on the distribution curve, the phase correlation matrix and the partial discharge phase center matrix are calculated. Specifically, multiple partial discharge pixel distribution values ​​under the 0~360° phase coordinates are sequentially retrieved from the distribution curve with 0 as the threshold. The partial discharge pixel points whose previous partial discharge pixel distribution value and their next partial discharge pixel distribution value are both 0 or not both greater than 0 are determined as set definition points. Based on all the set definition points, the phase correlation matrix and the partial discharge phase center matrix are calculated. Based on the constructed phase distribution decision threshold, the phase correlation matrix, and the partial discharge phase center matrix, the original PRPD spectrum is judged to obtain the identification result. Specifically, based on the phase distribution decision threshold, the phase correlation matrix and the partial discharge phase center matrix are used to obtain the modified phase correlation matrix and the modified partial discharge phase center matrix. The corresponding data in the modified phase correlation matrix and the modified partial discharge phase center matrix are calculated to obtain the data to be identified. Based on the threshold to be identified, the data to be identified is judged, and the original PRPD spectrum that meets the threshold requirement is identified as an interference spectrum.

2. The interference pattern identification method as described in claim 1, characterized in that, The step of binarizing the original PRPD map to be identified to obtain a binarized PRPD map includes the following steps: The original PRPD map is converted into a two-dimensional grayscale image; The grayscale threshold is obtained by calculating the grayscale image based on a preset grayscale threshold coefficient; Each pixel of the two-dimensional grayscale image is calibrated based on the grayscale threshold to obtain the binarized PRPD map.

3. The interference pattern identification method as described in claim 1, characterized in that, The step of calculating the distribution curve of partial discharge pixels that are greater than the maximum noise floor signal based on the amplitude threshold includes the following steps: Based on the amplitude threshold, the signal amplitude of each pixel in the original PRPD map is determined, and the set of all pixels whose signal amplitude is less than the amplitude threshold is defined as the noise set; The pixel points in the binarized PRPD map that correspond to the noise set are assigned a value of 0 to obtain the denoised pixel set. The 0 and 1 values ​​in the denoised pixel set are accumulated according to the 0-360° phase to obtain the distribution curve.

4. A morphological feature-based interference pattern recognition device, applied to electronic devices, characterized in that, The interference pattern identification device includes: The binarization processing module is configured to binarize the original PRPD map to be identified to obtain a binarized PRPD map. The threshold calculation module is configured to calculate the amplitude threshold of the maximum noise floor signal in the binarized PRPD map. Specifically, based on the binarized PRPD map, the pixels corresponding to the minimum amplitude and minimum phase in the original PRPD map are defined as base pixels. Based on the base pixels, the phase distribution characteristics of each pixel in the original PRPD map under a fixed amplitude are statistically analyzed. The amplitude threshold is determined by judging whether each pixel has a phase correlation with the phase distribution characteristics. The curve calculation module is configured to calculate the distribution curve of partial discharge pixels that are greater than the maximum noise floor signal based on the amplitude threshold. The matrix calculation module is configured to perform calculations based on the distribution curve to obtain the phase correlation matrix and the partial discharge phase center matrix. Specifically, it sequentially searches the distribution curve for multiple partial discharge pixel distribution values ​​in the 0~360° phase coordinates with 0 as the threshold, and determines the partial discharge pixel points whose previous partial discharge pixel distribution value and its next partial discharge pixel distribution value are both 0 or not both greater than 0 as set definition points. Based on all the set definition points, it calculates to obtain the phase correlation matrix and the partial discharge phase center matrix. The identification execution module is configured to judge the original PRPD spectrum based on the constructed phase distribution decision threshold, the phase correlation matrix, and the partial discharge phase center matrix to obtain the identification result. Specifically, based on the phase distribution decision threshold, the module obtains a modified phase correlation matrix and a modified partial discharge phase center matrix from the phase correlation matrix and the partial discharge phase center matrix. The module calculates the corresponding data in the modified phase correlation matrix and the modified partial discharge phase center matrix to obtain the data to be identified. Based on the threshold to be identified, the module judges the data to be identified and identifies the original PRPD spectrum that meets the threshold requirement as an interference spectrum.

5. An electronic device, characterized in that, The electronic device includes at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs or instructions; The processor is used to execute the computer program or instructions to enable the electronic device to implement the interference pattern recognition method as described in any one of claims 1 to 3.

6. A computer-readable storage medium for use in electronic devices, characterized in that, The storage medium carries one or more computer programs that can be executed by the electronic device, thereby enabling the electronic device to implement the interference pattern recognition method as described in any one of claims 1 to 3.

Images