A method for diagnosing typical partial discharge of gas insulated equipment based on high-dimensional features

Abstract

The application discloses a kind of gas insulation equipment typical partial discharge diagnosis method based on high-dimensional feature.Therein, method includes: in gas insulation equipment, multiple typical partial discharge defect tests are constructed, and the partial discharge signal set of gas insulation equipment is collected;According to the partial discharge signal set, the typical partial discharge spectrum of gas insulation equipment is generated;The partial discharge signal to be diagnosed of the gas insulation equipment to be diagnosed is collected;According to the partial discharge signal to be diagnosed and the partial discharge spectrum, the diagnosis result of the gas insulation equipment to be diagnosed is determined.The technical effect of the multiple signal fusion analysis of partial discharge is realized by the feature comparison of the newly collected partial signal to be diagnosed and the generation multiple typical partial discharge spectrum.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of partial discharge defect identification of gas insulated equipment, and more particularly, to a typical partial discharge diagnosis method for gas insulated equipment based on high-dimensional features. BACKGROUND

[0002] With the continuous development of China's economy, the demand for electricity in various industries is increasing. Under the guidance of the goal of carbon peak and carbon neutrality, electric energy has become one of the fastest growing industries in the energy industry. Compared with 2011, the number of substations / conversion stations operated by 27 provincial companies of State Grid in 2021 doubled, reaching more than 40,000. With the increasing requirements of substation intelligence, intensification, and unmanned operation, more and more substations use gas insulated equipment. Gas insulated equipment has the characteristics of small footprint and maintenance-free. However, there are still defects such as suspended potential discharge, internal air gap discharge of insulators, insulator surface discharge, sharp burr discharge, and free particle discharge in gas insulated equipment, which are important reasons for the insulation failure of gas insulated equipment. According to statistics, accidents caused by insulation failure of gas insulated equipment account for more than 30% of the total power outage in China.

[0003] Currently, partial discharge detection has been used for online monitoring and live detection of power equipment. Through external ultra-high frequency partial discharge sensors and ultrasonic partial discharge sensors, partial discharge signals occurring in gas insulated equipment can be detected, which is the main means for detecting partial discharge defects in gas insulated equipment. However, there are the following problems:

[0004] 1) Multiple signals generated by partial discharge in gas insulated equipment lack fusion analysis methods

[0005] The signals that can be detected by partial discharge are related to the cause of the discharge, the surrounding medium, and other factors. Usually, partial discharge is accompanied by multiple signals of different frequency bands and different types. In addition to external interference of gas insulated equipment, internal interference is also introduced through busbars and transformers and other equipment. However, the current methods of ultra-high frequency partial discharge detection and ultrasonic partial discharge detection are only for detecting a certain frequency band of partial discharge, and cannot diagnose and analyze the partial discharge source through the fusion of multiple signals, thus there is a risk of false detection.

[0006] 2) Analyzing signals with all comprehensive feature parameters will cause the curse of dimensionality

[0007] The original feature quantity is extracted from three aspects of statistical characteristics, fractal characteristics and image characteristics, and a principal component analysis method is used for dimension reduction processing, and finally a new feature quantity is obtained. However, research shows that the feature vector selected based on the traditional method, and then the dimension reduction or fusion of the final feature is not ideal, and when using the statistical element as the recognition feature of partial discharge, the statistical elements are strongly related to each other, and there is a certain degree of information redundancy. SUMMARY

[0008] In view of the deficiencies of the prior art, the application provides a gas insulated equipment typical partial discharge diagnosis method based on high-dimensional features.

[0009] According to one aspect of the application, a gas insulated equipment typical partial discharge diagnosis method based on high-dimensional features is provided, comprising:

[0010] Collecting a set of partial discharge signals of a plurality of typical partial discharge defect tests of the gas insulated equipment;

[0011] Generating a typical partial discharge spectrum of the gas insulated equipment according to the set of partial discharge signals;

[0012] Collecting a to-be-diagnosed partial discharge signal of the to-be-diagnosed gas insulated equipment;

[0013] According to the to-be-diagnosed partial discharge signal and the partial discharge spectrum, determining a diagnosis result of the to-be-diagnosed gas insulated equipment.

[0014] Optionally, the typical partial discharge defect test includes a suspended potential defect, an internal insulation discharge defect, an insulation surface discharge defect, a sharp tip defect and a particle discharge defect test, and the operation of constructing the typical partial discharge defect test in the gas insulated equipment comprises:

[0015] Changing the SF6 gas pressure in the gas insulated equipment from 0.4 MPa to 0.6 MPa;

[0016] Changing the defect position in the gas insulated equipment, and the defect position is changed from being close to the low-voltage side to being close to the high-voltage side;

[0017] Changing the test voltage level in the gas insulated equipment from 0.5U n to 2U n .

[0018] Optionally, the operation of collecting the set of partial discharge signals of the gas insulated equipment comprises:

[0019] The set of partial discharge signals in each typical partial discharge defect test is collected by the embedded ultrasonic partial discharge sensor, the single-photon sensor and the ultra-high frequency partial discharge sensor pre-set in the gas insulated equipment.

[0020] Optionally, the operation of generating a typical partial discharge spectrum of the gas insulated equipment according to the set of partial discharge signals comprises:

[0021] The frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity are extracted when the set of partial discharge signals is subjected to ultrasonic partial discharge, single-photon partial discharge and ultra-high frequency partial discharge.

[0022] The coarse pruning characteristic quantity combination is determined by performing coarse pruning on the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity.

[0023] The covariance matrix of the coarse pruning characteristic quantity combination is calculated, and the eigenvectors corresponding to the predetermined number of eigenvalues with the largest values in the covariance matrix are selected to form a fine pruning characteristic quantity combination.

[0024] The typical partial discharge spectrum is generated according to the fine pruning characteristic quantity combination.

[0025] Optionally, the frequency characteristic quantity comprises 27 ultrasonic partial discharge, single-photon partial discharge and ultra-high frequency partial discharge frequency characteristic quantities, namely, pulse phase positive / negative half-cycle pulse ratio R N , pulse rise time t1, pulse fall time t2 and pulse rise / fall time ratio R t , pulse frequency f, pulse amplitude H and average number of pulses per unit period and pulse number variance per unit period σ 2 .

[0026] The statistical characteristic quantity comprises 12 ultrasonic partial discharge, single-photon partial discharge and ultra-high frequency partial discharge statistical characteristic quantities, namely, skewness S k , kurtosis K u , discharge phase asymmetry A sy and phase correlation coefficient C C .

[0027] The spectrum characteristic quantity comprises 6 ultrasonic partial discharge, single-photon partial discharge and ultra-high frequency partial discharge spectrum characteristic quantities, namely, box dimension D and vacancy rate Λ.

[0028] Optionally, the operation of performing coarse pruning on the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity to determine the coarse pruning characteristic quantity combination comprises:

[0029] The correlation test is performed on the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity, and the characteristic quantities with a variation value less than a preset threshold value of the numerical value under all typical partial discharge conditions are removed to determine effective characteristic quantities.

[0030] The characteristic quantities with consistent or opposite correlations in the effective characteristic quantities are dynamically weighted to determine the coarse pruning characteristic quantity combination.

[0031] Optionally, the operation of generating a typical partial discharge pattern based on the combination of fine-pruning characteristic quantities includes:

[0032] The fine-pruned feature combination is mapped to two-dimensional matrix nodes, and the connection weights between the two-dimensional matrix nodes are initialized with the smallest random value.

[0033] Calculate the Euclidean distance between each node of the 2D matrix and the eigenvalues ​​of each dimensionality of the finely pruned branch, and select the node with the shortest Euclidean distance as the winning node.

[0034] The winning node is used as the center of the winning domain. The nodes contained in the winning domain are determined according to the pre-set domain radius. The other nodes in the winning domain are updated according to the domain function to generate a two-dimensional discrete mapping graph.

[0035] Typical partial discharge maps are generated when the learning rate of the two-dimensional discrete self-organizing feature map of the two-dimensional discrete map decays to 0 and the probability density of typical partial discharge is maximized.

[0036] Optionally, the operation of determining the diagnostic result of the gas-insulated device to be diagnosed based on the partial discharge signal and partial discharge spectrum includes:

[0037] Feature extraction and feature pruning are performed on the partial discharge signal to be diagnosed to determine the combination of finely pruned feature quantities for diagnosis.

[0038] The combination of fine-pruning features to be diagnosed is mapped onto the partial discharge spectrum, and the probability density and set of each type of partial discharge defect in the partial discharge spectrum are calculated.

[0039] Based on the probability density and set, the confidence level of each type of partial discharge defect is calculated, and the diagnostic result of the gas-insulated equipment to be diagnosed is determined based on the confidence level.

[0040] According to another aspect of the present invention, a typical partial discharge diagnostic device for gas-insulated equipment based on high-dimensional features is provided, comprising:

[0041] The first acquisition module is used to construct various typical partial discharge defect tests in the gas-insulated equipment and acquire the partial discharge signal set of the gas-insulated equipment;

[0042] A generation module is used to generate a typical partial discharge spectrum of the gas-insulated device based on the partial discharge signal set.

[0043] The second acquisition module is used to acquire the partial discharge signal of the gas-insulated equipment to be diagnosed.

[0044] The determination module is used to determine the diagnostic result of the gas-insulated device to be diagnosed based on the partial discharge signal to be diagnosed and the partial discharge spectrum.

[0045] According to yet another aspect of the present application, there is provided a computer readable storage medium storing a computer program for performing the method according to any of the preceding aspects of the present application.

[0046] According to yet another aspect of the present application, there is provided an electronic device comprising: a processor; a memory for storing instructions executable by the processor; and the processor configured to read the executable instructions from the memory and execute the instructions to implement the method according to any of the preceding aspects of the present application.

[0047] Therefore, the present application proposes a gas insulated equipment acoustic-optical-electrical combined sensing high-dimensional feature partial discharge defect diagnosis method. The partial discharge spectrum is constructed by collecting historical partial discharge signal data of each defect test of the gas insulated equipment. The partial discharge signal of the gas insulated equipment to be diagnosed is collected. The partial discharge signal to be diagnosed is diagnosed according to the pre-generated partial discharge spectrum, and the defect type of the gas insulated equipment to be diagnosed is determined. The problems of 1) lack of fusion analysis means for multiple signals of partial discharge; and 2) the signal feature extraction method based on statistical parameters will lead to too high dimension, which will cause difficulty in later partial discharge feature analysis and cause dimension disaster are solved. The gas insulated equipment typical partial discharge defect diagnosis function can be realized by the gas insulated equipment acoustic-optical-electrical combined sensing high-dimensional feature partial discharge defect diagnosis method. BRIEF DESCRIPTION OF DRAWINGS

[0048] The exemplary embodiments of the present application can be more fully understood with reference to the following drawings:

[0049] Figure 1 FIG. 1 is a flowchart of a gas insulated equipment typical partial discharge diagnosis method based on high-dimensional features according to an exemplary embodiment of the present application;

[0050] Figure 2a FIG. 3 is an acoustic-optical-electrical pulse time domain waveform according to an exemplary embodiment of the present application;

[0051] Figure 2b FIG. 4 is a single cycle acoustic-optical-electrical time domain waveform according to an exemplary embodiment of the present application;

[0052] Figure 2c FIG. 5 is a multi-cycle acoustic-optical-electrical pulse statistical feature extraction according to an exemplary embodiment of the present application;

[0053] Figure 2d FIG. 6 is a partial discharge spectrum feature extraction according to an exemplary embodiment of the present application;

[0054] Figure 3is a schematic diagram of self-organizing feature mapping provided by an exemplary embodiment of the present application;

[0055] Figure 4 is a schematic diagram of high-dimensional feature partial discharge type prediction provided by an exemplary embodiment of the present application;

[0056] Figure 5 is a structural schematic diagram of a typical partial discharge diagnosis device for gas insulated equipment based on high-dimensional features provided by an exemplary embodiment of the present application;

[0057] Figure 6 is a structure of an electronic device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0058] Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part but not all of the embodiments of the present application, and thus cannot be used to limit the scope of the present application. Various embodiments of the present application can be implemented in other specific forms without changing the technical essence or essential characteristics of the present application. Therefore, the description should not be construed as a limitation of the present application.

[0059] It should be noted that: unless otherwise specified, the relative arrangement, numerical expressions and values of the components and steps set forth in these embodiments do not limit the scope of the present application.

[0060] Those skilled in the art can understand that the terms "first", "second" and the like in the embodiments of the present application are only used to distinguish different steps, devices or modules, and do not represent any specific technical meaning, nor do they represent the inevitable logical sequence between them.

[0061] It should also be understood that in the embodiments of the present application, "a plurality of" can mean two or more, and "at least one" can mean one, two or more.

[0062] It should also be understood that for any component, data or structure mentioned in the embodiments of the present application, it can be understood as one or more in general, without explicit limitation or in the context of the opposite indication.

[0063] In addition, the term "and / or" in the present application is only a description of the association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B can mean that there are three cases of A alone, A and B together, and B alone. In addition, the character " / " in the present application generally represents an "or" relationship between the front and rear associated objects.

[0064] It should also be understood that the description of various embodiments of the present application focuses on the differences between the various embodiments, and the same or similar parts can be referred to each other, and for the sake of brevity, will not be repeated.

[0065] It should be understood, of course, that the detailed description and specific examples described below are intended for purposes of illustration only and are not intended to limit the scope of the present application.

[0066] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the scope of the application or its application or uses.

[0067] Techniques, methods, and apparatus known to those of ordinary skill in the relevant art can not be discussed in detail herein. However, where appropriate, techniques, methods, and apparatus should be considered as being encompassed by the scope of the disclosure.

[0068] It should be noted that like reference numerals and letters refer to like items throughout the attached drawings, and thus once an item is defined in one drawing, it is not necessary that it be further discussed in the remaining drawings.

[0069] Embodiments of the present application can be applied to terminal devices, computer systems, servers, and other electronic devices, which can operate with numerous other general purpose or special purpose computing system environments or configurations. Examples of well-known terminal devices, computer systems, environments, and / or configurations that can be suitable for use with terminal devices, computer systems, servers, and other electronic devices include, but are not limited to, personal computers, server computers, thin clients, thick clients, hand-held or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputers, mainframe computers, and distributed cloud computing technology environments that include any of the above systems, and the like.

[0070] Terminal devices, computer systems, servers, and other electronic devices can be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules can include routines, programs, objects, components, logic, data structures, and the like, which perform particular tasks or implement particular abstract data types. Computer systems / servers can be practiced in distributed cloud computing environments with remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules can be located in local or remote computer system storage media including memory storage devices.

[0071] Exemplary method

[0072] Figure 1 is a flowchart of a method for diagnosing typical partial discharge of gas insulated equipment based on high-dimensional features provided by an exemplary embodiment of the present application. The present embodiment can be applied to electronic devices such as Figure 1 As shown, the method 100 for diagnosing typical partial discharge of gas insulated equipment based on high-dimensional features includes the following steps:

[0073] Step 101, collecting a plurality of typical partial discharge defect test partial discharge signal sets of the gas insulated equipment;

[0074] Step 102, generating a typical partial discharge spectrum of the gas insulated equipment according to the partial discharge signal set;

[0075] Step 103, collecting a to-be-diagnosed partial discharge signal of the to-be-diagnosed gas insulated equipment;

[0076] Step 104, determining a diagnosis result of the to-be-diagnosed gas insulated equipment according to the to-be-diagnosed partial discharge signal and the partial discharge spectrum.

[0077] Specifically, in order to overcome the defects described in the background art and improve the effect of the gas insulated equipment partial discharge diagnosis method, the present application provides a gas insulated equipment acousto-optic-electric combined sensing high-dimensional feature partial discharge defect diagnosis method. The method is aimed at typical suspended electrode discharge, high-voltage guide rod spike discharge, surface discharge and free metal particle discharge defect types in the gas insulated equipment. The internal partial discharge signals of the gas insulated equipment are collected by embedded ultrasonic partial discharge sensors, single-photon sensors and ultra-high frequency partial discharge sensors. The time-frequency domain features, statistical features and spectrum features of the acousto-optic-electric partial discharge signals are extracted, the feature quantity combinations are obtained through rough pruning and fine pruning, the typical partial discharge spectrum is constructed by self-organizing mapping of the feature quantity combinations of the typical partial discharge signals, the probability density of the newly collected partial discharge signals is judged by comparing the feature quantity combinations of the newly collected partial discharge signals, and the confidence degree of the partial discharge type of the newly collected partial discharge signals is given. The method is simple to operate, economical and practical.

[0078] The specific process steps are as follows:

[0079] Step 1, constructing a typical partial discharge defect test in the gas insulated equipment, and collecting the internal partial discharge signals of the gas insulated equipment by embedded ultrasonic partial discharge sensors, single-photon sensors and ultra-high frequency partial discharge sensors.

[0080] Specifically, the typical partial discharge defects include suspended potential defects, internal insulation discharge defects, insulation surface discharge defects, sharp end defects and particle discharge defects.

[0081] Specifically, the construction of the typical partial discharge defect test in the gas insulated equipment includes: changing the SF6 gas pressure from 0.4 MPa to 0.6 MPa; changing the defect position from near the low-voltage side to near the high-voltage side; changing the test voltage level from 0.5U n to 2U n .

[0082] Specifically, the embedded ultrasonic partial discharge sensor, single photon sensor and ultra-high frequency partial discharge sensor are integrated sensors, which can jointly collect the same partial discharge signal from ultrasonic, ultraviolet and ultra-high frequency.

[0083] Specifically, the ultrasonic sensor probe collects 70-300 kHz acoustic signals, reflecting the vibration inside the gas insulated equipment caused by partial discharge defects; the single photon detection sensor probe collects electromagnetic waves with a wavelength of 100-400 nm, reflecting the ultraviolet radiation inside the gas insulated equipment caused by partial discharge defects; the ultra-high frequency partial discharge sensor probe collects electromagnetic waves with a frequency of 300M-3000MHz, reflecting the electromagnetic radiation inside the gas insulated equipment caused by partial discharge defects. Due to the diversity of partial discharge types, the uncertainty of partial discharge occurrence position and the diversity of noise sources, the collection of signals of three different frequency bands and different types can effectively avoid the missed detection and false detection of partial discharge.

[0084] Step 2, extracting ultrasonic partial discharge, single photon partial discharge and ultra-high frequency partial discharge frequency characteristics, including 27 time-frequency domain characteristics: pulse phase (acousto-optic), positive and negative half-cycle pulse ratio R N , pulse rise time t1(acousto-optic), pulse fall time t2(acousto-optic), pulse rise / fall time ratio R t (acousto-optic), pulse frequency f(acousto-optic), pulse amplitude H(acousto-optic), mean number of pulses per cycle (acousto-optic) and pulse number variance per cycle 2 (acousto-optic);

[0085] The statistical characteristics include 12 statistical characteristics of ultrasonic partial discharge, single photon partial discharge and ultra-high frequency partial discharge: skewness S k (acousto-optic), kurtosis K u (acousto-optic), discharge phase asymmetry A sy (acousto-optic), phase correlation coefficient C C (acousto-optic);

[0086] The spectrum feature quantity includes 6 spectrum feature quantities of ultrasonic partial discharge, single photon partial discharge and ultra-high frequency partial discharge: box dimension (acousto-optic) D, vacancy rate Λ(acousto-optic).

[0087] The time-frequency feature extraction method is as follows:

[0088] (1) Pulse phase = maximum amplitude pulse phase φ1, see Figure 2a ;

[0089] (2) Positive and negative half-cycle pulse ratio R N = Np / N n ; wherein, N p is the number of pulses in the positive half cycle, N n is the number of pulses in the negative half cycle, see Figure 2b ;

[0090] (3) Pulse rise time = t1, see Figure 2a ;

[0091] (4) Pulse fall time = t2, see Figure 2a ;

[0092] (5) Pulse rise / fall time ratio R t = t1 / t2, see Figure 2a ;

[0093] (6) Pulse frequency f = 1 / (t1+t2), see Figure 2a ;

[0094] (7) Pulse amplitude H, see Figure 2a ;

[0095] (8) Average number of pulses per cycle N N pi is the number of pulses in the positive half cycle of the i-th cycle, N ni is the number of pulses in the negative half cycle of the i-th cycle, n is the number of cycles, see 2c;

[0096] (9) Pulse number variance per cycle, see Figure 2c ;

[0097] The partial discharge statistical feature extraction method is as follows:

[0098] (1) Skewness wherein, q i , p i , μ, σ are the discharge amount, discharge probability, mean, and standard deviation of the i-th cycle, respectively.

[0099] (2) Kurtosis Kurtosis K u is the fourth-order central moment of data, which measures the dispersion of data.

[0100] (3) Discharge phase asymmetry Discharge phase asymmetry A sy represents the parameter ratio of positive and negative discharge cycles and discharge amount, wherein q p is the total discharge amount of the positive half cycle, q n is the total discharge amount of the negative half cycle, N p is the number of pulses in the positive half cycle, and N n is the number of pulses in the negative half cycle.

[0101] (4) Phase correlation coefficient

[0102] The local discharge feature extraction method is as follows:

[0103] (1) Box dimension Where N r =∑ i,j n r (i,j), n r (i,j) = k-l+1, a series of boxes with s x s x h size are stacked in each grid area, where the maximum value falls in the kth box and the minimum value falls in the lth box, then the number of boxes n r required to store pixel values for this grid Figure 2d ;

[0104] (2) Void rate Where,

[0105] Step 3, by performing correlation test on N full-quantity features extracted under typical partial discharge of gas insulated equipment, remove all feature quantities whose numerical change value is less than a preset threshold under typical partial discharge, where the preset threshold may be, but is not limited to, 0.05; weight the feature quantities with consistent or opposite correlation, and obtain a rough pruned feature quantity combination M.

[0106] Specifically, the correlation test step is as follows: 1) select a single feature quantity and 5 typical partial discharge conditions to construct a scatter plot from N full-quantity features; 2) determine whether the single feature quantity is subject to normal distribution; 3) calculate the correlation coefficient matrix and significance test; 4) output the correlation test result.

[0107] Specifically, the numerical change small test examines the variance of the feature quantity under 5 typical partial discharge of the gas insulated equipment, and retains the feature quantity greater than the threshold.

[0108] Specifically, for the feature quantities with consistent or opposite correlation, a dynamic weighting method is used to increase the weight of the feature quantity in the category prediction each time the category is accurately predicted.

[0109] Step 4, by calculating the covariance matrix of the M-dimensional feature quantity combination, the eigenvalue and eigenvector of the covariance matrix are obtained, and the matrix composed of the K eigenvectors corresponding to the K eigenvalues with the largest value (i.e., the largest variance) is selected. Obtain the pruned feature quantity combination K.

[0110] Specifically, the value of each node of the covariance Where a and b represent any two different feature quantities in the feature combination, m represents the number of M-dimensional feature quantities, and ai denotes the i-th eigenvalue of the a feature quantity, b i denotes the i-th eigenvalue of the b feature quantity, μ a denotes the variance of the a feature quantity, μ b denotes the variance of the b feature quantity.

[0111] Specifically, the M-dimensional feature extraction K feature combination method is as follows: 1) according to m M-dimensional composition matrix X; 2) zero mean each row of X; 3) the covariance matrix is solved 4) the eigenvalue and the corresponding eigenvector of the covariance matrix are solved; 5) the eigenvalues corresponding to the eigenvectors are arranged from top to bottom according to the size of the eigenvalues, and the first K rows of the features are taken as the fine pruning feature combination K.

[0112] Step 5, through self-organizing feature mapping, the K-dimensional feature is converted into two-dimensional discrete mapping. All two-dimensional node connection weights are initialized with small random values, a set of K-dimensional feature quantity combination is input, the similarity between K-dimensional feature quantity combination and nodes is calculated, the node with the highest similarity is selected as the winning node, the nodes contained in the winning neighborhood are determined according to the neighborhood radius r, the connection weights are updated through the neighborhood function, and finally the probability density of the K-dimensional feature quantity combination is approximated, that is, the typical partial discharge spectrum of gas insulated equipment, which includes suspended potential defect, internal discharge defect, insulating surface discharge defect, sharp tip defect and particle discharge defect defect area, and also includes unknown area.

[0113] Specifically, the K-dimensional feature conversion into two-dimensional discrete mapping can be realized through a two-dimensional rectangular node structure, as shown in Figure 3 , wherein the number of two-dimensional rectangular nodes determines the accuracy and generalization ability of the GIS typical partial discharge spectrum, and the minimum number of two-dimensional nodes is , that is , wherein w is the number of feature groups used to construct the GIS typical partial discharge spectrum.

[0114] Specifically, a set of K-dimensional feature combination is input, the two-dimensional rectangular nodes are traversed, the similarity between the K-dimensional feature combination and each node is calculated, and the similarity is expressed by the Euclidean distance. The node with the smallest distance is selected as the winning node.

[0115] Specifically, the nodes contained in the winning neighborhood are determined according to the neighborhood radius r, and the other nodes in the winning node neighborhood are updated through the neighborhood function. The update amplitude close to the winning node is large, and the update amplitude far from the winning node is small, wherein the winning node is the center of the winning field.

[0116] Specifically, whether the GIS typical partial discharge spectrum is generated is judged according to whether the two-dimensional discrete self-organizing feature mapping learning rate decays to 0 and the probability density and maximum of the typical partial discharge.

[0117] Specifically, the typical partial discharge probability density and wherein P i (j) is the probability density of the feature combination of the jth group on the ith partial discharge. When the probability density and P

[0118] Step 6: Collecting partial discharge signals in the gas insulated equipment through the embedded ultrasonic partial discharge sensor, single photon sensor and ultra-high frequency partial discharge sensor, extracting features and feature pruning to obtain K', mapping K' to the typical partial discharge spectrum of the gas insulated equipment through self-organizing feature mapping, calculating the probability density of different regions, and giving the confidence of different types of defects.

[0119] Specifically, the method for extracting features and feature pruning to obtain K' is consistent with steps 2, 3 and 4.

[0120] Specifically, referring to Figure 4 As shown, K' is mapped to the typical partial discharge spectrum of the GIS through self-organizing feature mapping, the probability density P k (x, y) of each two-dimensional node and K' is calculated, and the probability density and P i (k) of the ith type of GIS partial discharge is given, wherein P i (k) is the probability density of the ith type of GIS partial discharge.

[0121] Specifically, the confidence of different defects is The greater the confidence is, the more likely it is a certain defect.

[0122] Therefore, the present application proposes a gas insulated equipment sound-light-electricity combined sensing high-dimensional feature partial discharge defect diagnosis method, aiming to solve the problems of 1) lack of fusion analysis means for multiple signals of partial discharge; 2) the signal feature extraction method based on statistical parameters will lead to too high dimension, causing difficulty in later partial discharge feature analysis, and causing dimension disaster. Through the gas insulated equipment sound-light-electricity combined sensing high-dimensional feature partial discharge defect diagnosis method, the typical partial discharge defect diagnosis function of the gas insulated equipment can be realized.

[0123] Exemplary apparatus

[0124] Figure 5 is a structural schematic diagram of a gas insulated equipment typical partial discharge diagnosis device based on high-dimensional features provided by an exemplary embodiment of the present application. As Figure 5 shown, the device 500 comprises:

[0125] The first collecting module 510 is configured to construct multiple typical partial discharge defect tests in the gas insulated equipment, and collect a partial discharge signal set of the gas insulated equipment;

[0126] The generating module 520 is configured to generate a typical partial discharge spectrum of the gas insulated equipment according to the partial discharge signal set;

[0127] The second collecting module 530 is configured to collect a to-be-diagnosed partial discharge signal of a to-be-diagnosed gas insulated equipment;

[0128] The determining module 540 is configured to determine a diagnosis result of the to-be-diagnosed gas insulated equipment according to the to-be-diagnosed partial discharge signal and the partial discharge spectrum.

[0129] Optionally, the typical partial discharge defect tests include a floating potential defect test, an internal discharge defect test, a surface discharge defect test, a sharp-end defect test and a particle discharge defect test, and the first collecting module 510 includes:

[0130] The first changing sub-module is configured to change the SF6 gas pressure in the gas insulated equipment from 0.4 MPa to 0.6 MPa;

[0131] The second changing sub-module is configured to change the defect position in the gas insulated equipment from being close to the low-voltage side to being close to the high-voltage side;

[0132] The third changing sub-module is configured to change the test voltage level in the gas insulated equipment from 0.5U n to 2U n .

[0133] Optionally, the first collecting module 510 includes:

[0134] The collecting sub-module is configured to collect the partial discharge signal set in each typical partial discharge defect test by using the embedded ultrasonic partial discharge sensor, the single-photon sensor and the ultra-high frequency partial discharge sensor pre-set in the gas insulated equipment.

[0135] Optionally, the generating module 520 includes:

[0136] The feature extraction sub-module is configured to perform feature extraction on the ultrasonic partial discharge, the single-photon partial discharge and the ultra-high frequency partial discharge of the partial discharge signal set, to determine a frequency feature quantity, a statistical feature quantity and a spectrum feature quantity;

[0137] The coarse pruning sub-module is configured to perform coarse pruning on the frequency feature quantity, the statistical feature quantity and the spectrum feature quantity, to determine a coarse pruning feature quantity combination;

[0138] The fine pruning sub-module is configured to calculate a covariance matrix of the coarse pruning feature value combination, and select feature vectors corresponding to a predetermined number of feature values with the largest eigenvalues in the covariance matrix to form the fine pruning feature value combination;

[0139] The generating sub-module is configured to generate a typical partial discharge spectrum according to the fine pruning feature value combination.

[0140] Optionally, the frequency feature values include 27 ultrasonic partial discharge, single-photon partial discharge, and ultra-high frequency partial discharge frequency feature values: pulse phase positive and negative half-cycle pulse ratio R N , pulse rise time t1, pulse fall time t2, and pulse rise / fall time ratio R t , pulse frequency f, pulse amplitude H, and average number of pulses per unit period and pulse number variance σ 2 per unit period

[0141] The statistical feature values include 12 ultrasonic partial discharge, single-photon partial discharge, and ultra-high frequency partial discharge statistical feature values: skewness S k , kurtosis K u , discharge phase asymmetry A sy , and phase correlation coefficient C C

[0142] The spectrum feature values include 6 ultrasonic partial discharge, single-photon partial discharge, and ultra-high frequency partial discharge spectrum feature values: box dimension D and vacancy rate Λ.

[0143] Optionally, the coarse pruning sub-module includes:

[0144] The first determining unit is configured to perform correlation test on the frequency feature values, the statistical feature values, and the spectrum feature values, remove feature values with a change value less than a preset threshold value in all typical partial discharge cases, and determine effective feature values.

[0145] The second determining unit is configured to perform dynamic weighting on feature values with consistent or opposite correlations in the effective feature values, and determine the coarse pruning feature value combination.

[0146] Optionally, the generating sub-module includes:

[0147] The initializing unit is configured to map the fine pruning feature value combination to a two-dimensional matrix node, and initialize connection weights between the two-dimensional matrix nodes with minimum random values.

[0148] The winning node is determined as a node with the shortest Euclidean distance between nodes in each group of fine pruning dimensions and the two-dimensional matrix node.

[0149] ​The first generating unit generates a two-dimensional discrete mapping graph by taking the winning node as the center of the winning field, determining the nodes contained in the winning field according to a pre-set field radius, and updating other nodes of the winning field according to a field function.

[0150] The second generating unit generates a typical partial discharge spectrum under the condition that the two-dimensional discrete self-organizing feature mapping learning rate of the two-dimensional discrete mapping graph decays to 0 and the probability density and maximum of the typical partial discharge.

[0151] Optionally, the determining module 540 comprises:

[0152] The first determining sub-module is configured to perform feature extraction and feature pruning on the to-be-diagnosed partial discharge signal, and determine a to-be-diagnosed pruned feature quantity combination.

[0153] The calculating sub-module is configured to map the to-be-diagnosed pruned feature quantity combination to a partial discharge spectrum, and calculate the probability density and set of each type of partial discharge defect of the to-be-diagnosed pruned feature quantity combination to the partial discharge spectrum.

[0154] The second determining sub-module is configured to calculate the confidence of each type of partial discharge defect according to the probability density and set, and determine a diagnosis result of the to-be-diagnosed gas insulated equipment according to the confidence.

[0155] Exemplary electronic device

[0156] Figure 6 The electronic device 60 includes one or more processors 61 and memory 62. Figure 6 The processor 61 can be a central processing unit (CPU) or other form of processing unit that has data processing capability and / or instruction executing capability, and can control other components in the electronic device to perform desired functions.

[0157] The processor 61 can be a central processing unit (CPU) or other form of processing unit that has data processing capability and / or instruction executing capability, and can control other components in the electronic device to perform desired functions.

[0158] The memory 62 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), cache memory, and / or the like. The non-volatile memory, for example, can include read only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 61 can run the program instructions to implement the methods of the software programs of the various embodiments of the present application described above and / or other desired functions. In one example, the electronic device can further include an input device 63 and an output device 64, which are interconnected through a bus system and / or other forms of connection mechanisms (not shown).

[0159] In addition, the input device 63 can include, for example, a keyboard, a mouse, and / or the like.

[0160] The output device 64 can output various information to the outside. The output device 64 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and / or the like.

[0161] Of course, in order to simplify, Figure 6 In FIG. 1, only some of the components of the electronic device related to the present application are shown, and components such as buses, input / output interfaces, and / or the like are omitted. In addition, the electronic device can include any other appropriate components, depending on the specific application.

[0162] Exemplary computer program product and computer readable storage medium

[0163] In addition to the methods and devices described above, embodiments of the present application can also be a computer program product including computer program instructions that, when executed by a processor, cause the processor to perform steps of the methods according to various embodiments of the present application described in the above "Exemplary Methods" section of the specification.

[0164] The computer program product can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, and / or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, as a stand-alone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0165] In addition, an embodiment of the present application can also be a computer-readable storage medium, having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the steps of the method described in the above “Exemplary Method” section of the specification for information mining on historical change records according to various embodiments of the present application.

[0166] The computer-readable storage medium can take the form of one or more combinations of any type of readable media. The readable media can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, system, or device, or any suitable combination thereof. More specific examples (a non-exhaustive list) of the readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0167] The above describes the basic principles of the present application in combination with specific embodiments, but it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are only examples and are not limiting, and these advantages, benefits, effects, etc. cannot be considered as necessary for each embodiment of the present application. In addition, the above specific details are only for the purpose of example and for the purpose of understanding, and the above details do not limit the present application to the above specific details.

[0168] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between each embodiment can be mutually referred to. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment.

[0169] The block diagrams of the devices, systems, apparatuses, systems involved in the present application are only illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be as shown in the block diagrams. As those skilled in the art will recognize, these devices, systems, apparatuses, systems can be connected, arranged, configured in any manner. Words such as “include”, “contain”, “have”, etc. are open-ended words, mean “include but not limited to”, and can be used interchangeably. The words “or” and “and” used herein mean the word “and / or”, and can be used interchangeably unless the context clearly indicates otherwise. The word “such as” used herein means the phrase “such as but not limited to”, and can be used interchangeably.

[0170] The methods and systems of the present application can be implemented in a number of ways. For example, the methods and systems of the present application can be implemented via software, hardware, firmware, or any combination of software, hardware, and firmware. The above described order of steps for the methods is merely illustrative, and the steps of the methods of the present application are not limited to the order specifically described above unless otherwise specifically stated. Furthermore, in some embodiments, the present application can also be implemented as a program recorded on a recording medium, which includes machine readable instructions for implementing the methods according to the present application. Thus, the present application also covers recording media storing programs for executing the methods according to the present application.

[0171] It is also to be noted that in the systems, apparatuses, and methods of the present application, various components or steps can be split and / or recombined. Such splitting and / or recombining is to be considered as an equivalent of the present application. The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects without departing from the scope of the present application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0172] The above description has been presented for the purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although several example aspects and embodiments have been discussed, those skilled in the art will recognize certain variations, modifications, changes, additions, and sub-combinations thereof.

Claims

1. A method for diagnosing typical partial discharge of gas insulated equipment based on high-dimensional features, characterized in that, The method comprises the following steps: collecting a plurality of typical partial discharge defect test signals of a gas insulated device; generating a typical partial discharge spectrum of the gas insulated device according to the partial discharge signals; collecting a to-be-diagnosed partial discharge signal of a to-be-diagnosed gas insulated device; determining a diagnosis result of the to-be-diagnosed gas insulated device according to the to-be-diagnosed partial discharge signal and the partial discharge spectrum; The operation of generating a typical partial discharge spectrum of the gas insulated device according to the partial discharge signal set comprises: extracting features of ultrasonic partial discharge, single-photon partial discharge and ultra-high frequency partial discharge from the partial discharge signal set, determining frequency characteristic quantities, statistical characteristic quantities and spectrum characteristic quantities; rough pruning the frequency characteristic quantities, the statistical characteristic quantities and the spectrum characteristic quantities to determine a rough pruning characteristic quantity combination; calculating a covariance matrix of the rough pruning characteristic quantity combination, and selecting characteristic vectors corresponding to a predetermined number of characteristic values with the largest eigenvalues in the covariance matrix to form a fine pruning characteristic quantity combination; generating the typical partial discharge spectrum according to the fine pruning characteristic quantity combination; The operation of generating the typical partial discharge spectrum according to the fine pruning characteristic quantity combination comprises: mapping the fine pruning characteristic quantity combination into a two-dimensional matrix node, and initializing connection weights between the two-dimensional matrix nodes with minimum random values; respectively calculating Euclidean distances between each node of the fine pruning characteristic quantity combination and each node of the two-dimensional matrix node, and taking a node with the shortest Euclidean distance as a winning node; taking the winning node as a center of a winning field, determining nodes contained in the winning field according to a pre-set field radius, updating other nodes of the winning field according to a field function, and generating a two-dimensional discrete mapping graph; generating the typical partial discharge spectrum in the case that a two-dimensional discrete self-organizing feature mapping learning rate of the two-dimensional discrete mapping graph decays to 0 and a probability density and of the typical partial discharge are the largest; The operation of determining a diagnosis result of the to-be-diagnosed gas insulated device according to the to-be-diagnosed partial discharge signal and the partial discharge spectrum comprises: extracting features and pruning features of the to-be-diagnosed partial discharge signal to determine a to-be-diagnosed fine pruning characteristic quantity combination; mapping the to-be-diagnosed fine pruning characteristic quantity combination onto the partial discharge spectrum, and calculating a probability density and set of each type of partial discharge defect of the to-be-diagnosed fine pruning characteristic quantity combination to the partial discharge spectrum; calculating a confidence degree of each type of partial discharge defect according to the probability density and set, and determining the diagnosis result of the to-be-diagnosed gas insulated device according to the confidence degree.

2. The method of claim 1, wherein, The typical partial discharge defect test comprises a suspended potential defect test, an internal insulation discharge defect test, a dielectric surface discharge defect test, a sharp tip defect test and a micro-particle discharge defect test, and the operation of constructing the typical partial discharge defect test in the gas insulated device comprises: changing the SF6 gas pressure in the gas insulated device from 0.4 MPa to 0.6 MPa; changing a defect position in the gas insulated device from being close to a low-voltage side to being close to a high-voltage side. changing the test voltage level in the gas insulated device from 0.5 U n to 2 U n .

3. The method of claim 1, wherein, The operation of collecting the partial discharge signal set of the gas insulated equipment comprises: The partial discharge signal set in each of the typical partial discharge defect tests is collected by an embedded ultrasonic partial discharge sensor, a single-photon sensor and a UHF partial discharge sensor pre-set in the gas insulated equipment.

4. The method of claim 3, wherein, The frequency characteristic quantity includes: 27 ultrasonic partial discharge, single photon partial discharge, ultra-high frequency partial discharge time frequency characteristic quantity: pulse phase φ1 , positive and negative half cycle pulse ratio R N , pulse rise time t 1 , pulse fall time t 2 , pulse rise / fall time ratio R t , pulse frequency f , pulse amplitude H , mean value of pulse number per cycle and pulse number per cycle variance σ 2 ; The statistical characteristic quantities include 12 statistical characteristic quantities of ultrasonic partial discharge, single-photon partial discharge, and ultra-high frequency partial discharge: skewness S k , kurtosis K u , discharge phase asymmetry A sy , phase correlation coefficient C C ; The spectrum feature quantity includes 6 ultrasonic partial discharge, single photon partial discharge, spectrum feature quantity of ultra-high frequency partial discharge: box dimension D , vacancy rate Λ .

5. The method of claim 1, wherein, The operation of rough pruning the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity to determine a rough-pruned characteristic quantity combination comprises: The operation of performing correlation test on the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity to remove the characteristic quantities with a variation value less than a preset threshold under all typical partial discharge conditions to determine effective characteristic quantities; The operation of performing dynamic weighting on the characteristic quantities with consistent or opposite correlations in the effective characteristic quantities to determine the rough-pruned characteristic quantity combination.

6. A device for diagnosing typical partial discharge of gas insulated equipment based on high-dimensional features, for performing the method of any one of claims 1 to 5, characterized in that, Comprise: The first acquisition module is used for collecting partial discharge signal sets of multiple typical partial discharge defect tests of the gas insulated equipment; The generation module is used for generating a typical partial discharge spectrum of the gas insulated equipment according to the partial discharge signal sets; The second acquisition module is used for collecting a to-be-diagnosed partial discharge signal of a to-be-diagnosed gas insulated equipment; The determination module is used for determining a diagnosis result of the to-be-diagnosed gas insulated equipment according to the to-be-diagnosed partial discharge signal and the partial discharge spectrum.

7. The apparatus of claim 6, wherein, The first acquisition module comprises: The acquisition submodule is used for collecting the partial discharge signal set in each of the typical partial discharge defect tests by an embedded ultrasonic partial discharge sensor, a single-photon sensor and a UHF partial discharge sensor pre-set in the gas insulated equipment.

8. The apparatus of claim 7, wherein, The generation module comprises: The feature extraction submodule is used for performing feature extraction on the partial discharge signal set in ultrasonic partial discharge, single-photon partial discharge and UHF partial discharge to determine a frequency characteristic quantity, a statistical characteristic quantity and a spectrum characteristic quantity; The rough-pruning submodule is used for performing rough pruning on the frequency characteristic quantity, the statistical characteristic quantity and the spectrum characteristic quantity to determine a rough-pruned characteristic quantity combination; The fine-pruning submodule is used for calculating a covariance matrix of the rough-pruned characteristic quantity combination, and selecting a predetermined number of eigenvectors corresponding to the largest eigenvalues in the covariance matrix to form a fine-pruned characteristic quantity combination; The generation submodule is used for generating the typical partial discharge spectrum according to the fine-pruned characteristic quantity combination.

9. The apparatus of claim 8, wherein, The frequency characteristic quantity includes: 27 ultrasonic partial discharge, single photon partial discharge, ultra-high frequency partial discharge time frequency characteristic quantity: pulse phase φ1 , positive and negative half cycle pulse ratio R N , pulse rise time t 1 , pulse fall time t 2 , pulse rise / fall time ratio R t , pulse frequency f , pulse amplitude H , mean value of pulse number per cycle and pulse number variance per cycle σ 2 ; The statistical characteristic quantities include 12 statistical characteristic quantities of ultrasonic partial discharge, single-photon partial discharge, and ultra-high frequency partial discharge: skewness S k kurtosis K u discharge phase asymmetry A sy phase correlation coefficient C C ; The spectrum characteristic quantity includes 6 ultrasonic partial discharge, single photon partial discharge, spectrum characteristic quantity of ultra-high frequency partial discharge: box dimension D , vacancy rate Λ .

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and the computer program is used to execute the method in any one of claims 1-5.

11. An electronic device, comprising: The electronic device comprises: A processor; A memory for storing executable instructions of the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the method in any one of claims 1-5.

Images