Insulation defect identification method and device, computer equipment and storage medium

Abstract

The invention relates to an insulation defect identification method and device, computer equipment and a storage medium, and relates to the technical field of insulation detection. The method comprises the following steps: acquiring a PRPD map corresponding to power equipment to be detected, and dividing the PRPD map into a first sub-map and a second sub-map; obtaining a first signal group in the first sub-atlas, and determining a first time domain waveform data set corresponding to the first signal group and a first similarity in the first time domain waveform data set; if the first similarity is greater than a first threshold value, acquiring a second signal group in the second sub-atlas, and determining a second time domain waveform data set corresponding to the second signal group and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal group is determined based on the phase angle of the first signal group; and if the second similarity is greater than a second threshold, determining that an insulation defect signal exists in the to-be-detected power equipment. By adopting the method, the accuracy of insulation defect identification can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of insulation detection, and in particular to an insulation defect identification method, device, computer equipment, storage medium and computer program product. Background Art

[0002] The atlas technology used to judge partial discharge has developed rapidly, but the automatic judgment technology has been stagnant. The judgment of insulation defects is usually performed by testers or experts. With the promotion of urban high voltage technology and the advancement of urbanization, the workload of insulation defect identification has doubled. Therefore, in related technologies, automatic identification can be performed through artificial intelligence models to improve the efficiency of insulation defect identification.

[0003] In the related technology, threshold judgment of physical quantities is usually performed based on multiple logic gates, and the relationship between the amplitude and phase of the PRPD (Phase Resolved Partial Discharge) spectrum is used to identify insulation defects based on neural networks, and the correlation between multiple logic gates and neural networks is integrated to obtain the identification result. However, the identification of insulation defects through the features provided by the PRPD spectrum and the logical relationship of the feature quantities is essentially the identification of the phase distribution features of the PRPD spectrum, which makes the data dimension of the features used for insulation defect identification low, and the correlation between different features is insufficient, resulting in insufficient accuracy in insulation defect identification. Summary of the invention

[0004] Based on this, it is necessary to provide an insulation defect identification method, device, computer equipment, computer-readable storage medium and computer program product that can improve identification accuracy in response to the above technical problems.

[0005] In a first aspect, the present application provides a method for identifying insulation defects. The method comprises:

[0006] Acquire a PRPD spectrum corresponding to the power equipment to be detected, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum;

[0007] Acquire a first signal cluster in the first sub-spectrum, determine a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set; the first similarity is determined based on the labeled data and the comparison data in the first time-domain waveform data set; the labeled data is data corresponding to a pre-set number of samples configured in the first time-domain waveform data set, the comparison data is data in the first time-domain waveform data set other than the labeled data, and the number of the labeled data and the comparison data is the same;

[0008] If the first similarity is greater than a first threshold, obtaining a second signal cluster in the second sub-spectrum, and determining a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal cluster is determined based on a phase angle of the first signal cluster;

[0009] If the second similarity is greater than a second threshold, it is determined that an insulation defect signal exists in the power equipment to be detected.

[0010] In one embodiment, obtaining the first signal cluster in the first sub-spectrum, determining a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set, includes:

[0011] Retrieving the signal clusters contained in the first sub-spectrum, and if the area of ​​any signal cluster retrieved is larger than a preset area, determining the signal cluster as the first signal cluster, and the phase angle corresponding to the first signal cluster;

[0012] Based on the phase angle of the first signal cluster, a first time domain waveform data set corresponding to the first signal cluster is determined, and based on the marked data and the comparison data in the first time domain waveform data set, a first similarity is determined.

[0013] In one embodiment, determining the first similarity based on the marked data and the comparison data in the first time domain waveform data set includes:

[0014] The calculation formula for determining the first similarity is as follows:

[0015]

[0016] Among them, the first time domain waveform data set is , the label data is , the comparison data is , n represents the total number of elements in the first time domain waveform data set, and C represents the first similarity.

[0017] In one embodiment, the obtaining of the second signal cluster in the second sub-spectrum and determining the second time-domain waveform data set corresponding to the second signal cluster and the second similarity between the first time-domain waveform data set and the second time-domain waveform data set comprises:

[0018] Determining a search range in the second sub-spectrum based on a phase angle corresponding to the first signal cluster;

[0019] In the search range, a second signal group and a second time domain waveform data set corresponding to the second signal group are determined; the second time domain waveform data set is determined by taking the inverse of the data of the time domain waveform corresponding to the second signal group;

[0020] A second similarity is determined based on the first time-domain waveform data set and the second time-domain waveform data set.

[0021] In one embodiment, determining the second similarity based on the first time-domain waveform data set and the second time-domain waveform data set includes:

[0022] The calculation formula for determining the second similarity is as follows:

[0023]

[0024] Among them, the first time domain waveform data set is , the second time domain waveform data set is , n represents the total number of elements in the first time domain waveform data set, and E represents the second similarity.

[0025] In one embodiment, dividing the PRPD spectrum into a first sub-spectrum and a second sub-spectrum includes:

[0026] A dividing line is determined, and the PRPD spectrum is divided into a first sub-spectrum and a second sub-spectrum based on the dividing line; the dividing line is a vertical line, and the dividing point of the dividing line is located at a position corresponding to 180 degrees.

[0027] In one embodiment, the step of obtaining a PRPD spectrum corresponding to the power equipment to be detected includes:

[0028] Acquire a partial discharge signal data stream corresponding to the power equipment to be detected; the partial discharge signal data stream includes moments in a working cycle and discharge signal amplitudes corresponding to each moment;

[0029] For each working cycle, the PRPD spectrum corresponding to the power equipment to be detected is determined based on the number of discharges in each phase segment included in the working cycle and the discharge signal amplitude corresponding to each moment in the partial discharge signal data stream.

[0030] In a second aspect, the present application also provides an insulation defect identification device. The device comprises:

[0031] A spectrum division module, used for obtaining a PRPD spectrum corresponding to the power equipment to be detected, and dividing the PRPD spectrum into a first sub-spectrum and a second sub-spectrum;

[0032] A first similarity determination module is used to obtain a first signal cluster in the first sub-spectrum, determine a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set; the first similarity is determined based on the marked data and the comparison data in the first time-domain waveform data set; the marked data is data corresponding to a pre-set number of samples configured in the first time-domain waveform data set, the comparison data is data in the first time-domain waveform data set other than the marked data, and the number of the marked data and the comparison data is the same;

[0033] A second similarity determination module, configured to obtain a second signal cluster in the second sub-spectrum if the first similarity is greater than a first threshold, and determine a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal cluster is determined based on a phase angle of the first signal cluster;

[0034] The insulation defect recognition module is used to determine that an insulation defect signal exists in the power equipment to be detected if the second similarity is greater than a second threshold.

[0035] In a third aspect, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method described in the first aspect are implemented.

[0036] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0037] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0038] The above insulation defect identification method, device, computer equipment, storage medium and computer program product obtain the PRPD spectrum corresponding to the power equipment to be detected, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum. The first signal group in the first sub-spectrum is obtained, the first time domain waveform data set corresponding to the first signal group is determined, and the data corresponding to the pre-set number of samples configured in the first time domain waveform data set and the remaining data are equal, and the first similarity is determined. When the first similarity is greater than the first threshold, the second signal group in the second sub-spectrum is obtained, and the second time domain waveform data set corresponding to the second signal group is determined, and the second similarity is determined based on the first time domain waveform data set and the second time domain waveform data set. When the second similarity is greater than the second threshold, it is determined that there is an insulation defect signal in the power equipment to be detected, that is, it is identified that there is an insulation defect in the power equipment to be detected. Due to the periodicity of the PRPD spectrum and the symmetry of the PRPD spectrum of insulation defects, the PRPD spectrum of noise interference can be screened out by dividing the PRPD spectrum to obtain the first sub-spectrum and the second sub-spectrum. And by calculating the first similarity between the marked data and the comparison data in the first time domain waveform data set corresponding to the first sub-spectrum, it can be determined whether the first signal group is a defective pulse waveform from the same source. Afterwards, after determining that the first signal group is a defective pulse waveform from the same source, the corresponding second signal group can be found in the second sub-spectrum, and it can be determined whether the second time domain waveform data set corresponding to the second signal group is similar to the first time domain waveform data set. If similar, it can be determined that the PRPD spectrum has symmetry, and the second signal group is a defective pulse waveform from the same source, and it is determined that the power equipment to be detected contains an insulation defect signal, so that the PRPD spectrum and the time domain waveform data are combined through the above process to determine whether the power equipment has insulation defects, increase the data dimension used for insulation defect identification, and improve the accuracy of insulation defect identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A diagram showing an application environment of an insulation defect identification method in one embodiment;

[0041] Figure 2 A schematic diagram of a flow chart of an insulation defect identification method in one embodiment;

[0042] Figure 3is a schematic diagram of a typical insulation defect detection spectrum in one embodiment;

[0043] Figure 4 is a schematic diagram of a spectrum of typical defects in one embodiment;

[0044] Figure 5 is a schematic diagram of a spectrum of typical noise interference in one embodiment;

[0045] Figure 6 is a flow chart of an insulation defect identification method in another embodiment;

[0046] Figure 7 is a schematic diagram of a gridded PRPD spectrum in one embodiment;

[0047] Figure 8 A schematic diagram of a process for identifying insulation defects in another embodiment;

[0048] Fig. 9 A schematic diagram of a process for identifying insulation defects in another embodiment;

[0049] Fig.10 is a structural block diagram of an insulation defect identification device in one embodiment;

[0050] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0052] The insulation defect identification method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 can detect the power equipment to be detected, obtain the PRPD (Phase Resolved Partial Discharge) spectrum corresponding to the power equipment to be detected, and the terminal 102 can record the voltage amplitude at each moment through the acquisition rate of the ADC (Analog-to-Digital Converter) chip, so as to obtain the time domain waveform data corresponding to the power equipment to be detected. The terminal 102 can send the PRPD spectrum and time domain waveform data corresponding to the power equipment to be detected to the server 104. The server 104 can divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum. The server 104 obtains the first signal group in the first sub-spectrum and determines the first time domain waveform data set corresponding to the first signal group. The server 104 calculates the similarity of the marked data and the comparison data inside the first time domain waveform data set, the marked data is the data corresponding to the pre-set number of samples configured in the first time domain waveform set, and the comparison data is the data in the first time domain waveform data set except the marked data, and the number of the marked data and the comparison data is the same, and the first similarity in the first time domain waveform data set is obtained. When the first similarity is greater than the first threshold, the server obtains the second signal group in the second sub-map and determines the second time domain waveform data set corresponding to the second signal group. The server 104 calculates the similarity between the first time domain waveform data set and the second time domain waveform data set to obtain a second similarity. When it is determined that the second similarity is greater than the second threshold, the server 104 can determine that there is an insulation defect signal in the power equipment to be detected, and the first signal group and the second signal group in the PRPD spectrum, and the time domain waveforms corresponding to the first signal group and the second signal group are the parts corresponding to the insulation defect signal. The terminal 102 communicates with the server 104 via the network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers.

[0053] The terminal 102 may be, but is not limited to, various computers and partial discharge detection equipment, and the partial discharge detection equipment may be a high-frequency current sensor, an ultrasonic sensor, an ultra-high frequency sensor, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers.

[0054] In an exemplary embodiment, Figure 2 As shown, a method for identifying insulation defects is provided. Figure 1 The server 104 in the example is used as an example to illustrate the method, which includes the following steps S202 to S208. Among them:

[0055] Step S202: obtain a PRPD spectrum corresponding to the power equipment to be detected, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum.

[0056] Among them, the power equipment to be tested may include transformers, switchgear, gas-insulated combination electrical appliances, high-voltage cables, ring main units, reactors and other equipment. The PRPD spectrum is a graph that corresponds the amplitude of the partial discharge signal to the phase angle of the power frequency voltage, and displays the discharge activity in a complete voltage cycle from 0 to 360 degrees. The horizontal axis of the PRPD spectrum represents the phase angle of the power frequency voltage, which can range from 0 to 360 degrees, and the vertical axis represents the amplitude of the partial discharge signal. The color or brightness represents the frequency of occurrence of the discharge time.

[0057] Specifically, the server can obtain the PRPD spectrum of the power equipment to be detected through the terminal, and split the PRPD spectrum at a preset phase angle within a complete voltage cycle to obtain two sub-spectra, namely, a first sub-spectra and a second sub-spectra. In an example, the server can determine that the sub-spectra with a relatively small phase angle is the first sub-spectra, and determine that the sub-spectra with a relatively large phase angle is the second sub-spectra.

[0058] Step S204, obtaining a first signal cluster in the first sub-spectrum, determining a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set.

[0059] The first similarity is determined based on the marked data and the comparison data in the first time-domain waveform data set. The marked data is data corresponding to a preset number of samples configured in the first time-domain waveform data set, and the comparison data can be unmarked data, that is, data that is not marked in the first time-domain waveform data set, and the number of the marked data and the comparison data is the same.

[0060] Specifically, the server can search in the first sub-spectrum to determine whether there is a first signal cluster. When the first signal cluster is determined, the server can determine the time domain waveform data corresponding to each phase angle based on the phase angle, voltage amplitude and color of the first signal cluster in the PRPD spectrum, thereby obtaining a first time domain waveform data set. For example, the server can determine a scatter point with a phase angle of 90 degrees and all colors and all voltage amplitudes, and determine the time domain waveform data corresponding to the scatter points to obtain the corresponding time domain waveform data at the phase angle. After determining all the scattered points within the phase angle range corresponding to the first signal cluster, the first time domain waveform data set corresponding to the first signal cluster can be obtained.

[0061] The server may mark the first preset number of samples in the first time-domain waveform data set based on a preset marking strategy to obtain marked data, and determine other samples in the first time-domain waveform data set as comparison data. The server may compare the marked data and the comparison data at corresponding positions based on the positions of the marked data and the comparison data in the first time-domain waveform data set, thereby obtaining a first similarity between each marked data and each comparison data.

[0062] Step S206, if the first similarity is greater than the first threshold, obtain the second signal cluster in the second sub-spectrum, and determine the second time domain waveform data set corresponding to the second signal cluster, and the second similarity between the first time domain waveform data set and the second time domain waveform data set.

[0063] The second signal group is determined based on the phase angle of the first signal group. Since the PRPD spectrum corresponding to the power equipment with insulation defects has a certain symmetry, the server can determine the phase angle at the symmetrical position based on the phase angle (i.e., phase angle) of the first signal group, and determine the second signal group at the phase angle at the symmetrical position.

[0064] Specifically, the server can determine whether the first similarity is greater than the first threshold value. If the first similarity is not greater than the first threshold value, it is determined that the first signal group does not have an insulation defect. The server can obtain a new PRPD spectrum, and return to the step to divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum, and obtain the first signal group in the first sub-spectrum, determine the first time domain waveform data set corresponding to the first signal group, and the first similarity within the first time domain waveform data set. If the first similarity is greater than the first threshold value, it is determined that the first signal group may have an insulation defect, and the phase angle of the symmetrical position can be further determined based on the phase angle of the first signal group, and the second signal group can be retrieved in the second sub-spectrum based on the phase angle of the symmetrical position.

[0065] The server can determine the time domain waveform data corresponding to each phase angle according to the phase angle, voltage amplitude and color of the second signal group in the PRPD spectrum, thereby obtaining a second time domain waveform data set. Afterwards, the server can compare data in different data sets and at the same position based on the position of each sample in the first time domain waveform data set and the second time domain waveform data set, thereby obtaining a second similarity between the first time domain waveform data set and the second time domain waveform data set.

[0066] Step S208: If the second similarity is greater than the second threshold, it is determined that an insulation defect signal exists in the power equipment to be detected.

[0067] Specifically, if the second similarity is less than or equal to the second threshold, it can be determined that the second similarity between the first time domain waveform data set and the second time domain waveform data set is insufficient, that is, the distribution symmetry of the scattered points in the PRPD spectrum is insufficient, the PRPD spectrum is not the spectrum of the insulation defect, and the time domain waveform corresponding to the PRPD spectrum is not the time domain waveform corresponding to the insulation defect. If the second similarity is greater than the second threshold, it can be determined that the distribution of the scattered points in the PRPD spectrum is symmetrical, that is, it meets the typical characteristics of the insulation defect. Based on this, if the second similarity is greater than the second threshold, the server can determine that there is an insulation defect signal in the power equipment to be detected.

[0068] In the above insulation defect identification method, the PRPD spectrum corresponding to the power equipment to be detected is obtained, and the PRPD spectrum is divided into a first sub-spectrum and a second sub-spectrum. The first signal group in the first sub-spectrum is obtained, and the first time domain waveform data set corresponding to the first signal group is determined, and the data corresponding to the pre-set number of samples configured in the first time domain waveform data set and the remaining data are equal, and the first similarity is determined. When the first similarity is greater than the first threshold, the second signal group in the second sub-spectrum is obtained, and the second time domain waveform data set corresponding to the second signal group is determined, and the second similarity is determined based on the first time domain waveform data set and the second time domain waveform data set. When the second similarity is greater than the second threshold, it is determined that there is an insulation defect signal in the power equipment to be detected, that is, it is identified that the power equipment to be detected has an insulation defect. Due to the periodicity of the PRPD spectrum and the symmetry of the PRPD spectrum of insulation defects, the PRPD spectrum with noise interference can be screened out by dividing the PRPD spectrum to obtain the first sub-spectrum and the second sub-spectrum. And by calculating the first similarity between the marked data and the comparison data in the first time domain waveform data set corresponding to the first sub-spectrum, it can be determined whether the first signal group is a defective pulse waveform from the same source. Afterwards, after determining that the first signal group is a defective pulse waveform from the same source, the corresponding second signal group can be found in the second sub-spectrum, and it can be determined whether the second time domain waveform data set corresponding to the second signal group is similar to the first time domain waveform data set. If similar, it can be determined that the PRPD spectrum has symmetry, and the second signal group is a defective pulse waveform from the same source, and it is determined that the power equipment to be detected contains an insulation defect signal, so that the PRPD spectrum and the time domain waveform data are combined through the above process to determine whether the power equipment has insulation defects, increase the data dimension used for insulation defect identification, and improve the accuracy of insulation defect identification.

[0069] In an exemplary embodiment, the specific implementation process of the step of "obtaining the first signal cluster in the first sub-spectrum, determining the first time domain waveform data set corresponding to the first signal cluster, and the first similarity within the first time domain waveform data set" includes:

[0070] Retrieve the signal clusters contained in the first sub-spectrum, and if the area of ​​any signal cluster retrieved is larger than the preset area, determine the signal cluster as the first signal cluster, and the phase angle corresponding to the first signal cluster; based on the phase angle of the first signal cluster, determine the first time domain waveform data set corresponding to the first signal cluster, and determine the first similarity based on the marked data and comparison data in the first time domain waveform data set.

[0071] Specifically, the server can configure a sliding window of a preset area, and perform a sliding search in the first sub-spectrum through the sliding window to determine whether the first sub-spectrum contains a signal cluster with an area greater than the preset area. If there is a signal cluster with an area greater than the preset area, the signal cluster is determined to be the first signal cluster. The server can determine the first time domain waveform data set corresponding to the first signal cluster based on the phase angle of the first signal cluster. The server can determine the voltage amplitude and color of the phase angle in the PRPD spectrum, and determine the time domain waveform data corresponding to each phase angle based on the phase angle of the first signal cluster to obtain the first time domain waveform data set.

[0072] In one example, the phase angle of the first signal group may be a range, and the range may include multiple phase angles. The first time domain waveform data set is determined by merging the time domain waveform data corresponding to the multiple phase angles. When merging, the server may sequentially connect according to the order of the phase angles to obtain the first time domain waveform data set.

[0073] Afterwards, the server may mark the first preset number of samples in the first time domain waveform data set according to a preset marking strategy to obtain marked data, and determine that other samples in the first time domain waveform data set are comparison data. For example, the first time domain waveform data set may be .

[0074] Among them, the first p samples are determined as labeled data, and the samples after p+1 are determined as remaining data. for The dividing data point between the labeled data and the remaining data is The server can calculate the similarity between the samples, and merge the similarity corresponding to each sample to obtain the first similarity corresponding to the first time domain waveform data set.

[0075] In this embodiment, by obtaining a signal cluster with an area larger than a preset area, the first signal cluster in the PRPD spectrum can be obtained, and based on the first signal cluster, the first time domain waveform data set can be obtained, and the first similarity corresponding to the first time domain waveform data set can be obtained, which can improve the accuracy of the first time domain waveform data set.

[0076] In an exemplary embodiment, the specific implementation process of the step of “determining the first similarity based on the marked data and the comparison data in the first time domain waveform data set” includes:

[0077] The calculation formula for determining the first similarity is as follows:

[0078]

[0079] Among them, the first time domain waveform data set is , the labeled data is , the comparison data is , n represents the total number of elements in the first time domain waveform data set, and C represents the first similarity.

[0080] Specifically, the server may calculate the difference between the marked data of the corresponding position and the comparison data of the corresponding position, respectively, to obtain the difference value of each position. The server may divide the position with the difference value by all the positions to be compared, to obtain the ratio of the difference, and the server may subtract the ratio from 1 to obtain the ratio of no difference, which is used as the first similarity.

[0081] In one example, n can be 2 times the value of p. In this case, the calculation formula can determine that the data from 1 to p is the marked data, and the data from p+1 to 2p is the comparison data. The server can sequentially convert (w 11 , w 1p+1 ), (w 12 , w 1p+2 )......(w 1p , w 12p ) to find out whether there is a difference in each position. The server can confirm that W 1c and W 1p Subtract the difference value, and define a group of data with a difference value greater than a preset difference as having a difference, and define a group of data with a difference value less than or equal to the preset difference as having no difference. The value of the difference can be 1, and the value of the no difference can be 0. Based on this, the server can obtain the ratio of the data with differences in all the data, for example, the ratio is 0.2, thereby obtaining the first similarity 1-0.2=0.8.

[0082] In this embodiment, the similarity between the marked data and the compared data can be determined through the marked data of each corresponding position and the compared data of the corresponding position, thereby obtaining the first similarity, which can improve the efficiency and accuracy of determining the first similarity.

[0083] In an exemplary embodiment, the specific implementation process of the step of "obtaining a second signal cluster in the second sub-spectrum, and determining a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set" includes:

[0084] Based on the phase angle corresponding to the first signal cluster, determine the search range in the second sub-spectrum; in the search range, determine the second signal cluster and the second time domain waveform data set corresponding to the second signal cluster; the second time domain waveform data set is determined by taking the inverse of the data of the time domain waveform corresponding to the second signal cluster; based on the first time domain waveform data set and the second time domain waveform data set, determine the second similarity.

[0085] Among them, the search range can be obtained by adding the first angle and the second angle on the basis of the phase angle, thereby obtaining a search range consisting of the sum of the phase angle and the first angle, and the sum of the phase angle and the second angle. For example, in the PRPD spectrum of a typical defect discharge signal, the signal group of the positive half cycle (1st and 2nd quadrants) corresponds to the negative half cycle (3rd and 4th quadrants), and the phase difference between the center points of the first quadrant and the third quadrant is 180°. For this, the range is increased, that is, 160°~200°, the phase angle is f1, the first angle is 160 degrees, and the second angle is 200 degrees. Then the search range is [f1+160°, f1+200°]. In actual testing, it is almost impossible for f1 to be equal to 10°, because the phase angle 10° corresponds to a voltage close to the zero axis, that is, no power generation will be generated when the voltage is zero, so the actual phase of the discharge can only be at a higher voltage position, which is about 45° to 135° according to experience, so the range of the phase angle f1 is specified as [45°~135°].

[0086] Specifically, the server can construct a search range for searching the second sub-spectrum based on the phase angle of the first signal group. The server can search in the search range through a sliding window corresponding to a preset area to see whether there is a signal group with an area greater than the preset area. If there is a signal group with an area greater than the preset area, the signal group is determined to be the second signal group. The server can determine the time domain waveform data corresponding to the second signal group based on the phase angle of the second signal group. Since the positive half-cycle (1st and 2nd quadrants) signal group corresponds to the negative half-cycle (3rd and 4th quadrants) in the PRPD spectrum of a typical defect discharge signal, the server can take the inverse of the time domain waveform data corresponding to the second signal group, and determine the time domain waveform data after taking the inverse as the second time domain waveform data set.

[0087] Afterwards, the server can compare the first time domain waveform data set corresponding to the first signal group and the second time domain waveform data set corresponding to the second signal group. For each sample in the first time domain waveform data set and the second time domain waveform data set, a group of samples at the same position are compared to obtain the similarity corresponding to each group of samples, and finally the similarities corresponding to all groups of samples are merged to obtain the second similarity.

[0088] In this embodiment, the retrieval range is determined by the phase angle of the first signal group, and the second time domain waveform data set is determined within the retrieval range. The second similarity is determined by the first time domain waveform data set and the second time domain waveform data set. The symmetry of the PRPD spectrum with insulation defects can be utilized to accurately obtain the second signal group and the second time domain waveform data of the first-grade steel, thereby improving the accuracy of determining the second similarity.

[0089] In an exemplary embodiment, the specific implementation process of the step of “determining the second similarity based on the first time domain waveform data set and the second time domain waveform data set” includes:

[0090] The calculation formula for determining the second similarity is as follows:

[0091]

[0092] Among them, the first time domain waveform data set is , the second time domain waveform data set is , n represents the total number of elements in the first time domain waveform data set, and E represents the second similarity.

[0093] Specifically, the server may respectively calculate the difference values ​​between the samples of the first time domain waveform data set and the second time domain waveform data set at the same position to obtain the difference value of each group of samples. The server may divide the number of samples with difference values ​​by the total number of samples to be compared to obtain a ratio with differences, and the server may subtract the ratio from 1 to obtain a ratio with no differences, thereby serving as the second similarity.

[0094] In one example, the number of samples in the first time-domain waveform data set and the number of samples in the second time-domain waveform data set may be the same or different. The embodiment of the present application describes the same number of samples. The server may take samples at the same position from the first time-domain waveform data set and the second time-domain waveform data set, respectively, to obtain multiple sample groups (w 11 , w 21 ), (w 12 , w 22 )......(w 1n , w 2n), based on multiple sample groups, the formula W2-W1 can be executed to obtain the difference value between each sample group. If the difference value is greater than the preset difference, it is determined that the sample group has a difference. If the difference value is less than or equal to the preset difference, it is determined that the sample group has no difference, that is, the sample groups are similar. At the same time, the value of the sample group with a difference can be defined as 1, and the value of the similar sample group can be defined as 0, so as to obtain the sum of W2-W1, and determine the proportion of the sum in n, so as to obtain the difference degree, and use 1 to subtract the difference degree to obtain the second similarity.

[0095] In this embodiment, the second similarity is obtained by calculating the difference between samples of the first time-domain waveform data set and the second time-domain waveform data set at the same position, which can improve the efficiency and accuracy of determining the second similarity.

[0096] In an exemplary embodiment, the specific implementation process of the step of "dividing the PRPD spectrum into a first sub-spectrum and a second sub-spectrum" includes:

[0097] A dividing line is determined, and the PRPD spectrum is divided into a first sub-spectrum and a second sub-spectrum based on the dividing line; the dividing line is a vertical line, and the dividing point of the dividing line is located at a position corresponding to 180 degrees.

[0098] Specifically, the server can determine a demarcation point in the PRPD spectrum where the phase angle is 180 degrees, and take a vertical line perpendicular to the horizontal axis of the PRPD spectrum at the demarcation point to obtain a demarcation line. The server can divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum based on the demarcation line. Usually, the sub-spectrum on the left can be defined as the first sub-spectrum, and the sub-spectrum on the right can be defined as the second sub-spectrum.

[0099] In this embodiment, the PRPD spectrum is made of vertical lines at the dividing point of 180 degrees, which can fully utilize the symmetry of the waveform in the PRPD spectrum and improve the accuracy and rationality between the first sub-spectrum and the second sub-spectrum obtained by division.

[0100] In an exemplary embodiment, the specific implementation process of the step of "obtaining the PRPD spectrum corresponding to the power equipment to be detected" includes:

[0101] Obtain a partial discharge signal data stream corresponding to the power equipment to be detected; the partial discharge signal data stream includes moments in a working cycle and the discharge signal amplitude corresponding to each moment; for each working cycle, based on the number of discharges in each phase segment included in the working cycle and the discharge signal amplitude corresponding to each moment in the partial discharge signal data stream, determine the PRPD spectrum corresponding to the power equipment to be detected.

[0102] Among them, partial discharge occurs in electrical equipment when there are defects such as air gaps, impurities, moisture, etc. in the insulating medium. Under the action of a strong electric field, partial discharge occurs inside the insulation. This causes the charge in the insulating medium to be quickly transferred and redistributed, forming physical phenomena such as local pulse currents and electromagnetic waves. The initial partial discharge signal is an analog signal, which needs to be converted into a digital signal through analog-to-digital conversion (ADC) technology for subsequent processing, transmission, and analysis. The ADC chip samples and quantizes the analog partial discharge signal at a certain sampling frequency and quantization accuracy, and converts it into a series of discrete digital data, which constitute the partial discharge signal data stream. The working cycle can be an industrial frequency cycle.

[0103] Specifically, the server can obtain the discharge signal amplitude corresponding to multiple power frequency cycles. For each power frequency cycle, the discharge signal amplitude corresponding to each phase angle in the power frequency cycle can be represented as a scatter point in the PRPD spectrum. After completing the traversal of multiple power frequency cycles, scatter points may be repeatedly generated at positions corresponding to similar (same) phase angles and similar (same) discharge signal amplitudes. Then, the scatter point at this position can change color according to the discharge frequency, thereby indicating the frequency of discharge of the scatter point defect.

[0104] In one example, the server records the partial discharge signal data stream, requiring the recorded data stream to be a two-dimensional data set. :

[0105]

[0106] Among them, T k is the kth moment in the working cycle, Q k is the kth amplitude of the discharge signal, N k is the kth discharge frequency of the signal, and k is the number of data measurements. Two-dimensional data set Includes T and Q, while N can be represented by color depth.

[0107] After that, the server can divide the time interval of a power frequency cycle (working cycle) into a number of (M) phase segments in equal proportion, and divide the number of discharges in each phase segment into It can be calculated by the following formula:

[0108]

[0109] As shown in the above formula, M is the number of phase segments divided by the detected power frequency cycle, h d is the number of discharges in the dth phase segment. The formatted two-dimensional data set , filled into the grid, where T is the horizontal coordinate and Q is the vertical coordinate, and the frequency of defect discharge is expressed by the color depth of the scattered points.

[0110] In this embodiment, the accuracy of generating the PRPD spectrum can be improved by determining the PRPD spectrum corresponding to the power equipment to be detected based on the number of discharges in each phase segment included in the working cycle and the discharge signal amplitude corresponding to each moment in the local discharge signal data stream.

[0111] The specific implementation process of the above insulation defect identification method is described in detail below in conjunction with a specific embodiment.

[0112] First, a typical insulation defect detection map, such as Figure 3 As shown, it includes PRPD spectrum, PRPS (Phase-Resolved Pulse Sequenc) spectrum and spectrum diagram, and the lower part also includes qt diagram, nt diagram and time domain waveform, among which:

[0113] PRPD spectrum, the horizontal axis is the phase angle (0-360°), corresponding to the four quadrants of the sine wave; the vertical axis is the discharge amount (pC), reflecting the relationship between the discharge signal and the phase angle; Figure 3 The PRPD spectrum is the result that has been inverted, with the aim of:

[0114] 1) Use less screen space to display graph features;

[0115] 2) Prepare for the next step of comparing the first signal cluster and the second signal cluster.

[0116] The PRPS spectrum is based on the three-dimensional coordinates of the PRPD spectrum. The added coordinate axis represents time. As time increases, the slices of the spectrum increase (migrate) forward, reflecting the relationship between the discharge signal and the phase angle and time accumulation.

[0117] The spectrum diagram, the horizontal axis is the frequency (1-100MHz), the vertical axis is the discharge amplitude (dB or V), which reflects the relationship between the discharge signal amplitude and frequency distribution;

[0118] The qt graph, with the horizontal axis representing time and the vertical axis representing discharge amount, reflects the changing trend of discharge amount over time;

[0119] nt spectrum, the horizontal axis is time, and the vertical axis is discharge density (frequency, how many discharges per second), reflecting the changing trend of discharge density over time;

[0120] The time domain waveform, with the horizontal axis being time and the vertical axis being amplitude, reflects the relationship between the discharge signal amplitude and time. Its relationship with the PRPD spectrum is that each point in the PRPD diagram can be interpreted as a time domain waveform.

[0121] It can be seen that in the PRPD spectrum, the red PD can indicate the location where partial discharge exists. In the PRPS spectrum, the red PD can indicate the location where partial discharge exists. In the spectrum, the red PD can indicate the location where partial discharge exists, and the green lines in the spectrum can indicate noise. Based on this, both the PRPD spectrum and the time domain waveform can be used as the current mainstream insulation defect detection spectrum forms.

[0122] like Figure 4 As shown in the figure, for different insulation defect types, the representations in the PRPD spectrum and time domain waveform are different. Figure 4 The PRPD spectrum and time domain waveform corresponding to the internal air gap discharge, the PRPD spectrum and time domain waveform corresponding to the surface discharge, the PRPD spectrum and time domain waveform corresponding to the semiconductor cracking, and the PRPD spectrum and time domain waveform corresponding to the buffer layer ablation defect are shown in turn. Figure 5 As shown, the PRPD spectrum and time domain waveform under various typical noise interferences are shown. Figure 5 The PRPD spectrum and time domain waveform under motor interference, the PRPD spectrum and time domain waveform under power system / power supply interference, the PRPD spectrum and time domain waveform under mobile phone / radio interference, and the PRPD spectrum and time domain waveform under corona discharge are shown in turn. Figure 4 and Figure 5 From the PRPD spectrum in the figure, we can see that the PRPD spectrum of insulation defects has origin symmetry at a phase angle of 180 degrees. Figure 4 and Figure 5 All of them are signal graphs without negation processing. In the calculation process of this application, it is necessary to negate.

[0123] In addition, by Figure 3 and Figure 4 From the spectrum, we can see that no matter it is a typical insulation defect signal or noise interference, there is a certain relationship between its waveform and PRPD. In particular, the pulse waveform of the insulation defect signal forms a corresponding relationship with the cluster signal of different phase segments.

[0124] Therefore, the embodiment of the present application proposes a method for global cross-correlation judgment of insulation defect detection spectrum. It mainly solves the problems of insufficient global correlation in existing insulation defect judgment methods, resulting in high misjudgment rate and lack of comparative analysis of time domain waveform spectra, and then solidifies this method into a set of computer executable programs, which automatically judges and issues an alarm when an insulation defect signal is found.

[0125] The workflow of the global cross-correlation judgment method for insulation defect detection graph is as follows: Figure 6As shown, it is specifically an algorithm for analyzing and judging the association after the segmentation and transformation of the PRPD spectrum and the time domain waveform, which includes inputting data stream according to the format, drawing a gridded PRPD spectrum, dividing the spectrum into left and right parts with 180° as the dividing line, searching whether there is a signal cluster in the left spectrum, searching the time domain waveform W1 corresponding to the signal cluster, judging the similarity between the waveform and the marked data, and starting to search for the signal cluster of the right PRPD spectrum if the similarity is high (otherwise returning to searching the left spectrum), if there is a signal cluster, searching the corresponding time domain waveform W2, and comparing the correlation with W1 after "taking the opposite number" of the waveform. If the correlation is greater than 80%, it is found that the insulation defect signal causes an alarm.

[0126] like Figure 6 As shown, the specific implementation process of the insulation defect identification method includes:

[0127] (1) Format of input data stream: The input data stream is required to be a two-dimensional data set .

[0128] Wherein, T is the time in the working cycle, Q is the amplitude of the discharge signal, N is the discharge frequency of the signal, and n is the number of data measurements.

[0129]

[0130] (2) Draw a gridded PRPD spectrum: divide the time interval of a power frequency cycle into a number of (M) phase segments in equal proportion, and determine the number of discharges in each phase segment as , which can be calculated by the following formula:

[0131]

[0132] As shown in the above formula, M is the number of phase segments divided by the detected power frequency cycle, h d is the number of discharges in the dth phase segment. The server will format the two-dimensional data stream , filled into the grid, where T is the horizontal axis and Q is the vertical axis. The color depth of the scattered points is used to express the frequency of defect discharge. The distribution characteristics and concentrated areas of defect signals can be determined through this spectrum. Due to the different sampling rates of the device, the number of grids can be increased to 1024*1024. In order to better demonstrate the method, Figure 7 As shown, a 10*10 grid is drawn for explanation.

[0133] (3) Split the left and right parts of the PRPD spectrum: Split the spectrum obtained in (2) into two PRPD spectra according to the 180° dividing line.

[0134] (4) Search in the PRPD spectrum on the left to find a signal cluster with an area larger than 3*3 and define the phase angle f1 at which the signal cluster appears. The range of f1 is [45°~135°].

[0135] (5) If a signal cluster is found, find the time domain waveform of the data point at the corresponding time. , establish the first time domain waveform data set corresponding to the signal group , of which the first Samples are determined as labeled data, and other unlabeled data are determined as comparison data. Both the labeled data and the remaining comparison data are used for comparison and similarity calculation. The purpose of calculating similarity is to confirm whether all the signal groups belong to the same type of signal source.

[0136] (6) Calculate the similarity of the characteristic waveforms within the signal group.

[0137] Setting: Labeled Dataset , and the data set used for comparison , n is the total number of the first time domain waveform data set, p=n / 2; the calculation formula for constructing the waveform similarity C is as follows:

[0138]

[0139] Similarity measurement standard: the larger the C is, the more similar the data set is to the labeled sample; the smaller the C is, the more different the data set is from the sample.

[0140] (7) If the similarity is greater than 70%, it means that they may be defect pulse waveforms from the same source. Save the first time domain waveform data set, and search the signal cluster of the PRPD spectrum on the right, and define the search phase range f2 as:

[0141]

[0142] (8) If a signal cluster is retrieved within the defined phase range, find the corresponding data point time domain waveform, establish the corresponding waveform data and transpose it to obtain the second time domain waveform data set The meaning of transposition is to reverse the amplitude of the corresponding data point.

[0143] (9) Determine the correlation of the time domain waveform where the signal cluster is located and define the cluster correlation E. If the correlation approaches 0, it means that there is no correlation. The correlation calculation formula is as follows:

[0144]

[0145] (10) Repeat the above operations (1)-(8) to calculate how many data sets have a correlation greater than 0.8. And determine that the data set is a data set with insulation defects.

[0146] In one example, if Figure 8 As shown, this embodiment provides an on-site live detection implementation method for the global cross-correlation judgment method of the insulation defect detection spectrum. The user will first determine the object to be tested and the type of live detection sensor corresponding to the object to be tested, and collect and process the on-site data through the live detection sensor. After that, the user will input it into the judgment algorithm associated with the embodiment of the present application to obtain the judgment and recognition result, and the judgment result will be displayed in real time to the user of the on-site test for the user to judge and issue a report.

[0147] In one embodiment, Fig. 9 As shown, this embodiment provides an online monitoring implementation method for the global cross-correlation judgment method of the insulation defect detection map. After the online monitoring equipment is installed and debugged, the online monitoring equipment collects and processes the on-site detection data, and inputs the detection data into the correlation judgment algorithm of the embodiment of the present application to obtain the recognition result. If the recognition result is greater than the alarm threshold preset by the system, a sound, light or email is issued to alarm until the on-duty personnel retrieve the recognition result and monitoring data for manual retest confirmation.

[0148] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0149] Based on the same inventive concept, the embodiment of the present application also provides an insulation defect identification device for implementing the insulation defect identification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more insulation defect identification device embodiments provided below can refer to the limitations of the insulation defect identification method above, and will not be repeated here.

[0150] In an exemplary embodiment, Fig.10As shown, an insulation defect identification device 1000 is provided, comprising: a spectrum division module 1001, a first similarity determination module 1002, a second similarity determination module 1003 and an insulation defect identification module 1004, wherein:

[0151] A spectrum division module 1001 is used to obtain a PRPD spectrum corresponding to the power equipment to be detected, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum;

[0152] A first similarity determination module 1002 is used to obtain a first signal cluster in a first sub-spectrum, determine a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set; the first similarity is determined based on the marked data and the comparison data in the first time-domain waveform data set; the marked data is data corresponding to a preset number of samples configured in the first time-domain waveform data set, the comparison data is data in the first time-domain waveform data set other than the marked data, and the number of the marked data and the comparison data is the same;

[0153] A second similarity determination module 1003 is used to obtain a second signal cluster in the second sub-spectrum if the first similarity is greater than the first threshold, and determine a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal cluster is determined based on the phase angle of the first signal cluster;

[0154] The insulation defect identification module 1004 is configured to determine that an insulation defect signal exists in the power equipment to be detected if the second similarity is greater than a second threshold.

[0155] Furthermore, the first similarity determination module 1002 is specifically used to: retrieve the signal cluster contained in the first sub-spectrum, and if the area of ​​any retrieved signal cluster is larger than a preset area, determine the signal cluster as the first signal cluster, and the phase angle corresponding to the first signal cluster; based on the phase angle of the first signal cluster, determine the first time domain waveform data set corresponding to the first signal cluster, and determine the first similarity based on the marked data and comparison data in the first time domain waveform data set.

[0156] Furthermore, the first similarity determination module 1002 is further configured to: determine the first similarity using the following calculation formula:

[0157]

[0158] Among them, the first time domain waveform data set is , the labeled data is , the comparison data is , n represents the total number of elements in the first time domain waveform data set, and C represents the first similarity.

[0159] Furthermore, the second similarity determination module 1003 is specifically used to: determine the search range in the second sub-spectrum based on the phase angle corresponding to the first signal cluster; determine the second signal cluster and the second time domain waveform data set corresponding to the second signal cluster in the search range; the second time domain waveform data set is determined by taking the inverse of the data of the time domain waveform corresponding to the second signal cluster; determine the second similarity based on the first time domain waveform data set and the second time domain waveform data set.

[0160] Furthermore, the second similarity determination module 1003 is further configured to: determine the second similarity using the following calculation formula:

[0161]

[0162] Among them, the first time domain waveform data set is , the second time domain waveform data set is , n represents the total number of elements in the first time domain waveform data set, and E represents the second similarity.

[0163] Furthermore, the spectrum division module 1001 is specifically used to: determine a dividing line, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum based on the dividing line; the dividing line is a vertical line, and the dividing point of the dividing line is located at a position corresponding to 180 degrees.

[0164] Furthermore, the spectrum division module 1001 is specifically used to: obtain a local discharge signal data stream corresponding to the power equipment to be detected; the local discharge signal data stream includes moments in a working cycle and the discharge signal amplitude corresponding to each moment; for each working cycle, based on the number of discharges in each phase segment included in the working cycle and the discharge signal amplitude corresponding to each moment in the local discharge signal data stream, determine the PRPD spectrum corresponding to the power equipment to be detected.

[0165] Each module in the above insulation defect identification device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0166] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.11As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the PRPD spectrum of the power equipment to be detected, as well as a time domain waveform data set. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for identifying insulation defects is implemented.

[0167] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0168] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0169] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0170] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0171] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0172] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0173] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for identifying insulation defects, characterized in that: The method comprises: Acquire a PRPD spectrum corresponding to the power equipment to be detected, and divide the PRPD spectrum into a first sub-spectrum and a second sub-spectrum; Acquire a first signal cluster in the first sub-spectrum, determine a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set; the first similarity is determined based on the labeled data and the comparison data in the first time-domain waveform data set; the labeled data is data corresponding to a pre-set number of samples configured in the first time-domain waveform data set, the comparison data is data in the first time-domain waveform data set other than the labeled data, and the number of the labeled data and the comparison data is the same; If the first similarity is greater than a first threshold, obtaining a second signal cluster in the second sub-spectrum, and determining a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal cluster is determined based on a phase angle of the first signal cluster; If the second similarity is greater than a second threshold, it is determined that an insulation defect signal exists in the power equipment to be detected.

2. The method according to claim 1, characterized in that The obtaining of the first signal cluster in the first sub-spectrum, and determining a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set, includes: Retrieving the signal clusters contained in the first sub-spectrum, and if the area of ​​any signal cluster retrieved is larger than a preset area, determining the signal cluster as the first signal cluster, and the phase angle corresponding to the first signal cluster; Based on the phase angle of the first signal cluster, a first time domain waveform data set corresponding to the first signal cluster is determined, and based on the marked data and the comparison data in the first time domain waveform data set, a first similarity is determined.

3. The method according to claim 2, characterized in that The determining of the first similarity based on the marked data and the comparison data in the first time domain waveform data set includes: The calculation formula for determining the first similarity is as follows: Among them, the first time domain waveform data set is , the label data is , the comparison data is , n represents the total number of elements in the first time domain waveform data set, and C represents the first similarity.

4. The method according to claim 2, characterized in that: The obtaining of the second signal cluster in the second sub-spectrum, and determining a second time-domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time-domain waveform data set and the second time-domain waveform data set, includes: Determining a search range in the second sub-spectrum based on a phase angle corresponding to the first signal cluster; In the search range, a second signal group and a second time domain waveform data set corresponding to the second signal group are determined; the second time domain waveform data set is determined by taking the inverse of the data of the time domain waveform corresponding to the second signal group; A second similarity is determined based on the first time-domain waveform data set and the second time-domain waveform data set.

5. The method according to claim 4, characterized in that The determining of a second similarity based on the first time-domain waveform data set and the second time-domain waveform data set comprises: The calculation formula for determining the second similarity is as follows: Among them, the first time domain waveform data set is , the second time domain waveform data set is , n represents the total number of elements in the first time domain waveform data set, and E represents the second similarity.

6. The method according to claim 1, characterized in that The dividing the PRPD spectrum into a first sub-spectrum and a second sub-spectrum comprises: A dividing line is determined, and the PRPD spectrum is divided into a first sub-spectrum and a second sub-spectrum based on the dividing line; the dividing line is a vertical line, and the dividing point of the dividing line is located at a position corresponding to 180 degrees.

7. The method according to claim 1, characterized in that The obtaining of the PRPD spectrum corresponding to the power equipment to be detected includes: Acquire a partial discharge signal data stream corresponding to the power equipment to be detected; the partial discharge signal data stream includes moments in a working cycle and discharge signal amplitudes corresponding to each moment; For each working cycle, the PRPD spectrum corresponding to the power equipment to be detected is determined based on the number of discharges in each phase segment included in the working cycle and the discharge signal amplitude corresponding to each moment in the partial discharge signal data stream.

8. An insulation defect identification device, characterized in that: The device comprises: A spectrum division module, used for obtaining a PRPD spectrum corresponding to the power equipment to be detected, and dividing the PRPD spectrum into a first sub-spectrum and a second sub-spectrum; A first similarity determination module is used to obtain a first signal cluster in the first sub-spectrum, determine a first time-domain waveform data set corresponding to the first signal cluster, and a first similarity within the first time-domain waveform data set; the first similarity is determined based on the marked data and the comparison data in the first time-domain waveform data set; the marked data is data corresponding to a pre-set number of samples configured in the first time-domain waveform data set, the comparison data is data in the first time-domain waveform data set other than the marked data, and the number of the marked data and the comparison data is the same; A second similarity determination module, configured to obtain a second signal cluster in the second sub-spectrum if the first similarity is greater than a first threshold, and determine a second time domain waveform data set corresponding to the second signal cluster, and a second similarity between the first time domain waveform data set and the second time domain waveform data set; the second signal cluster is determined based on a phase angle of the first signal cluster; The insulation defect recognition module is used to determine that an insulation defect signal exists in the power equipment to be detected if the second similarity is greater than a second threshold.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images