CN114325256A - Power equipment partial discharge identification method, system, equipment and storage medium - Google Patents
Power equipment partial discharge identification method, system, equipment and storage medium Download PDFInfo
- Publication number
- CN114325256A CN114325256A CN202111411547.7A CN202111411547A CN114325256A CN 114325256 A CN114325256 A CN 114325256A CN 202111411547 A CN202111411547 A CN 202111411547A CN 114325256 A CN114325256 A CN 114325256A
- Authority
- CN
- China
- Prior art keywords
- partial discharge
- type
- prpd
- prpd spectrogram
- spectrogram
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000003860 storage Methods 0.000 title claims abstract description 21
- 238000005070 sampling Methods 0.000 claims abstract description 46
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000013528 artificial neural network Methods 0.000 claims abstract description 33
- 238000000605 extraction Methods 0.000 claims abstract description 14
- 238000004364 calculation method Methods 0.000 claims description 41
- 238000004422 calculation algorithm Methods 0.000 claims description 22
- 238000007621 cluster analysis Methods 0.000 claims description 14
- 238000001228 spectrum Methods 0.000 claims description 14
- 238000000926 separation method Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 abstract description 3
- 238000001514 detection method Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 12
- 238000009413 insulation Methods 0.000 description 9
- 238000009826 distribution Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 230000007547 defect Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000012423 maintenance Methods 0.000 description 5
- 239000004020 conductor Substances 0.000 description 4
- 238000003909 pattern recognition Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000032683 aging Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 239000012212 insulator Substances 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 125000004432 carbon atom Chemical group C* 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000004138 cluster model Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Abstract
A power equipment partial discharge identification method, a system, equipment and a storage medium comprise: sampling data of the power equipment within a certain time is collected by adopting a broadband high-speed sampling mode; converting the sampling data into a PRPD spectrogram, and extracting and classifying signal characteristics of the PRPD spectrogram; substituting statistical characteristics in signal characteristics extracted from each type of PRPD spectrogram into a pre-trained partial discharge recognition model corresponding to the type to obtain whether partial discharge exists in the power equipment and the type of the partial discharge; the partial discharge recognition model is obtained by training the statistical characteristics extracted from the PRPD spectrogram of each type and the discharge type to which the statistical characteristics belong by adopting a neural network. According to the invention, sampling data are acquired by adopting broadband high-speed sampling, a complete partial discharge waveform is acquired, the PRPD spectrogram is subjected to signal characteristic extraction and classification, and a pre-trained partial discharge identification model is adopted for partial discharge identification, so that the partial discharge monitoring precision of the equipment is improved, and false alarm and missing report are reduced.
Description
Technical Field
The invention relates to the field of signal identification, in particular to a method, a system, equipment and a storage medium for identifying partial discharge of power equipment.
Background
Partial Discharge (PD): it is meant that the discharge occurs in only a localized area in the insulation system without breakdown, i.e. the discharge does not penetrate between the conductors to which the voltage is applied. Partial discharges may occur inside the insulator, at the interface of the insulator and the conductor, and at the surface of the insulator. With gas around the conductor, the PD at the edge of the conductor is also called corona.
With the development of social science and technology, the economic level is continuously improved, and the demand for electric power is continuously increased. The power distribution network as the last link in the power network is directly oriented to the terminal users and directly related to the power supply reliability and the power utilization quality of the majority of users, so that the power distribution network has very important significance in the state overhaul and operation and maintenance management of power equipment in the daily operation process of the power network. Partial discharge (partial discharge for short) is not only a cause of insulation aging of the power equipment, but also an important expression form of insulation aging and internal defects of the equipment, so that the partial discharge identification is an important means for identifying the state of the power equipment.
The partial discharge identification is an important means for identifying the defects and states of the power equipment, can effectively reflect the potential problem of equipment insulation, and is a history of decades of researches on partial discharge and detection technology by domestic and foreign experts and scholars.
Partial discharge detection: based on the detection of various physical quantities generated when partial discharge occurs, such as electricity, sound, light, etc., detection methods such as an electrical detection method, an acoustic detection method, a light detection method, etc., have been proposed according to the change in these physical quantities.
Partial discharge detection is considered to be the most important and effective insulation state assessment method for electrical equipment, and the partial discharge is used as the main expression form of the early insulation fault of the electrical equipment, namely the main cause of insulation aging and the main characteristic parameter for representing the insulation condition, so the partial discharge detection is one of the most important insulation state detection means.
The discharge signal obtained by the partial discharge detection needs to be identified in the discharge type, a discharge pattern identification method which needs to be assisted by various signal analysis and intelligent algorithms is an important technical field of partial discharge, a PRPD spectrogram statistical analysis method is a partial discharge signal feature extraction method which is widely applied at present, and a neural network algorithm is a signal electric feature identification and classification method which is widely applied.
Partial discharge phase distribution (PRPD) spectrum: and describing a spectrogram of the relation among a power frequency phase (0-360 degrees) of partial discharge, a discharge amplitude q and a discharge frequency n.
The following describes a principle of extracting a feature of a partial discharge signal by using a PRPD spectrogram in a partial discharge detection technique, taking the PRPD spectrogram as an example. As shown in fig. 1, the PRPD spectrogram is a two-dimensional or three-dimensional spectrogram which counts and describes the amplitude, frequency and phase relationship of a partial discharge signal within a period of time, spectrograms with different discharge types having different characteristics are similar to each other, and the type to which the discharge belongs can be identified through the PRPD spectrogram by using a certain algorithm.
Neural networks (neural networks) are a typical representation of machine learning in data mining by simulating human thinking through mathematical algorithms. Neural networks are abstract computational models of the human brain. In short, the neural network algorithm is to train a large amount of data to finally obtain a desired input/output model F.
A neural network algorithm is used in the detection, classification and identification of the partial discharge signals, and a model F is finally obtained by training a large number of electric signal characteristics and corresponding discharge types, so that the maximum probability of the model F meets the condition that F (electric signal characteristic X) is the discharge type corresponding to the electric signal characteristic X. Currently, commonly used neural networks include Back Propagation (BP) neural network, Radial Basis Function (RBF) neural network, and the like.
The conventional partial discharge detection faces the following two problems: 1) the local discharge signal is weak, and the extraction of the electrical characteristic quantity becomes very difficult due to the field electromagnetic interference, but the current mainstream anti-interference method is a narrow-band filtering method, but after the filtering method is adopted, a complete pulse waveform cannot be obtained, partial discharge signals can be lost while partial interference signals are separated, and the characteristic extraction and accurate identification of the partial discharge signals are influenced. 2) The existing neural network algorithm for identifying and classifying partial discharge signals is not high in working efficiency and convergence accuracy.
Disclosure of Invention
In order to solve the problems that the partial discharge signals are easy to annihilate due to electromagnetic interference in the prior art, and the working efficiency and convergence accuracy of a neural network algorithm for identifying and classifying the partial discharge signals in the prior art are not high, the invention provides a partial discharge identification method for power equipment, which comprises the following steps:
sampling data of the power equipment within a certain time is collected by adopting a broadband high-speed sampling mode;
converting the sampling data into a PRPD spectrogram, and extracting and classifying signal features of the PRPD spectrogram;
substituting the statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has partial discharge and the partial discharge type of the power equipment in the certain time;
the trained partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
Preferably, the training of the partial discharge recognition model includes:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
Preferably, the cluster analysis method includes: a K-means algorithm and a modified K-means algorithm.
Preferably, the converting the sampling data into a PRPD spectrogram comprises:
determining first and last sampling data from the sampling data according to the time sequence, wherein the first and last sampling data are respectively used as a starting point and an end point;
taking a point located in the middle of the starting point and the end point as an intermediate point;
finding a time position with the amplitude value of 0 or closest to 0 forwards and backwards respectively by taking the middle point as a center;
a PRPD spectrum is constructed by the waveform data between the time positions with the amplitude of 0 or the closest 0, which is called the main peak of a pulse waveform.
Preferably, the signal feature extraction and classification of the PRPD spectrogram comprises:
calculating the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the equivalent bandwidth of the main peak respectively by adopting a skewness calculation formula, a steepness calculation formula, a time expected value calculation formula, a frequency expected value calculation formula, an equivalent duration calculation formula and an equivalent bandwidth calculation formula based on the pulse waveform in the PRPD spectrogram;
and performing cluster analysis on the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the time domain feature vector of the equivalent frequency width of the main peak by using a cluster analysis method to obtain a PRPD spectrogram which rejects the noise and retains the partial discharge signal and a PRPD spectrogram which only contains the noise.
Preferably, the skewness calculation formula is as follows:
in the formula, SkfIs the degree of skewness, xiFor the point-by-point data of the frequency domain waveform, μ is all xiThe average value of (1) is the number, W is the total number, and σ is the corresponding standard deviation.
Preferably, the steepness calculation is as follows:
in the formula, KufIs steep.
Preferably, the time expectation value calculation formula and the frequency expectation value calculation formula are as follows:
in the formula, tjAnd gi(tj) Respectively representing time and amplitude in the time domain of a single pulse signal, m being the total number of times, gi 2(tj) Representing the square of the time domain amplitude of a single pulse signal; f. ofjAnd gi(fj) Respectively representing the frequency and amplitude of a single pulse signal, gi 2(fj) Representing the square of the frequency domain amplitude of a single pulse signal, j representing the index number, T0 iTo a desired value of time, F0 iIs the frequency expected value.
Preferably, the equivalent duration calculation formula and the equivalent bandwidth calculation formula are as follows:
in the formula, TiTo equivalent duration, FiIs the equivalent bandwidth.
Preferably, the bringing the statistical features in the signal features extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has partial discharge within the certain time and the partial discharge type to which the power equipment belongs includes:
and substituting the statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type, if the statistical characteristics are consistent with phase characteristic signals in the partial discharge recognition model, recognizing the partial discharge and outputting a partial discharge type, otherwise, recognizing the partial discharge type as noise.
In another aspect, the present application further provides a power device partial discharge identification system, including:
the data acquisition module is used for acquiring sampling data of the power equipment within a certain time by adopting a broadband high-speed sampling mode;
the separation and classification module is used for converting the sampling data into a PRPD spectrogram and extracting and classifying signal features of the PRPD spectrogram;
the partial discharge identification module is used for substituting statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge identification model corresponding to the type to obtain whether the power equipment has partial discharge within the certain time and the partial discharge type of the power equipment;
the partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
Preferably, the training of the partial discharge recognition model includes:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
Preferably, the separation and classification module includes:
a conversion submodule for determining first and last sampled data from the sampled data according to a time sequence, as a start point and an end point, respectively; taking a point located in the middle of the starting point and the end point as an intermediate point; finding a time position with the amplitude value of 0 or closest to 0 forwards and backwards respectively by taking the middle point as a center; constructing a PRPD spectrogram by using waveform data between the time positions with the amplitude value of 0 or the time position closest to 0 as a main peak of a pulse waveform;
a feature extraction submodule, configured to calculate a skewness, a steepness, a time expected value, a frequency expected value, an equivalent duration calculation formula, and an equivalent bandwidth calculation formula of the main peak based on a pulse waveform in the PRPD spectrogram by using the skewness calculation formula, the steepness calculation formula, the time expected value, the frequency expected value, the equivalent duration, and the equivalent bandwidth calculation formula, respectively;
and the classification submodule is used for performing cluster analysis on the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the time domain feature vector of the equivalent frequency width of the main peak by using a cluster analysis method to obtain a PRPD spectrogram which eliminates the noise and retains the partial discharge signal and a PRPD spectrogram which only contains the noise.
In yet another aspect, the present application further provides a computer device, including:
one or more processors;
a processor for executing one or more programs;
when the one or more programs are executed by the one or more processors, a power equipment partial discharge identification method as described above is implemented.
In still another aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the method for identifying partial discharge of an electric power device is implemented.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a power equipment partial discharge identification method, which comprises the following steps: sampling data of the power equipment within a certain time is collected by adopting a broadband high-speed sampling mode; converting the sampling data into a PRPD spectrogram, and extracting and classifying signal features of the PRPD spectrogram; substituting the statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has partial discharge and the partial discharge type of the power equipment in the certain time; the partial discharge recognition model is obtained by training the statistical characteristics extracted from the PRPD spectrogram of each type and the discharge type to which the statistical characteristics belong by adopting a neural network. The invention adopts a broadband high-speed sampling mode to acquire sampling data, can acquire complete partial discharge waveforms, avoids signal loss, adopts a pre-trained partial discharge identification model to perform partial discharge identification, can effectively improve the partial discharge monitoring precision of equipment and reduce the situations of misinformation and missing report.
Drawings
Figure 1 is a prior art PRPD spectrum:
FIG. 2 is a flow chart of a method for identifying partial discharge of an electrical device according to the present invention;
FIG. 3 is a general flowchart of a partial discharge identification method for power equipment according to the present invention;
FIG. 4 is a phase distribution spectrum of the present invention;
FIG. 5 is a diagram of a partial discharge pulse waveform according to the present invention;
FIG. 6 is a flow chart of a cluster analysis based on a multi-mean algorithm according to the present invention;
FIG. 7 is a BP neural network architecture framework of the present invention;
FIG. 8 is a BP neural network programming framework of the present invention;
FIG. 9 is a schematic diagram of the present invention for continuously acquiring 10 second data;
FIG. 10 is a schematic of a 2 microsecond length pulse of the present invention;
FIG. 11 is a PRPD spectrum of a full pulse in an embodiment of the present invention;
FIG. 12 is a PRPD spectrum of class 1 after waveform feature extraction in an embodiment of the present invention;
fig. 13 is a PRPD spectrogram of classification 2 through waveform feature extraction in the embodiment of the present invention.
Detailed Description
The invention can solve the problem of signal loss in the traditional partial discharge signal acquisition, and improves the identification and classification of the partial discharge signals based on an improved and optimized neural network algorithm. The invention adopts a broadband high-speed sampling method to collect partial discharge signals so as to obtain complete partial discharge waveforms, avoid signal loss and provide a good identification basis for partial discharge signal identification; extracting waveform time domain characteristics of the collected partial discharge signals, and providing an automatic classification and separation method; and aiming at each classified PRPD spectrogram, extracting statistical characteristic quantity of the PRPD spectrogram, and performing pattern recognition by adopting a neural network algorithm, so that the partial discharge signal of the equipment can be accurately recognized.
The core method provided by the invention has the advantages that the partial discharge signals are acquired by a broadband high-speed sampling method, the loss of the partial discharge signals is effectively avoided, the defect of the traditional filtering acquisition of the lost signals is overcome, meanwhile, the classification and identification of the partial discharge signals are realized by adopting a two-stage algorithm, the source factor and the hidden defect of equipment faults can be effectively found, the identification precision of the partial discharge of the equipment is improved, the false alarm and the missing alarm are reduced, operation and maintenance personnel can timely process the equipment defects, the accidents caused by further expansion of the defects are avoided, the equipment fault risk is reduced, and the operation safety, the power supply reliability and the operation and maintenance precision are improved.
Example 1
A power equipment partial discharge identification method, as shown in fig. 2, includes:
i: sampling data of the power equipment within a certain time is collected by adopting a broadband high-speed sampling mode;
II: converting the sampling data into a PRPD spectrogram, and performing feature extraction and classification on the PRPD spectrogram;
III: bringing the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has partial discharge in the certain time and the partial discharge type to which the power equipment belongs;
the trained partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
The training of the partial discharge recognition model comprises the following steps:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
The invention provides a method for identifying partial discharge of power equipment, which comprises the following steps: firstly, a two-stage partial discharge signal and noise discrimination method of 'rough classification-fine learning' is provided, and a complete partial discharge waveform can be obtained by adopting broadband high-speed sampling, so that the loss of signals is avoided; a signal feature extraction technology and a noise classification and separation technology based on partial discharge pulse waveform features are provided in a first stage; a partial discharge pattern recognition algorithm based on machine learning is provided in the second stage, a recognition model for partial discharge is established, the recognition precision of the partial discharge of the equipment can be effectively improved, and false alarms are reduced, so that the operation reliability and the operation and maintenance precision of the power distribution equipment are obviously improved. The overall process is divided into 3 stages, as shown in fig. 3:
partial discharge recognition principle: the waveform of a typical partial discharge should be a steep pulse that rises and falls rapidly, appearing as a major waveform peak, then falls rapidly in amplitude, as distinguished from ringing. According to the waveform characteristics of the signals, a large number of signals acquired on site are subjected to type discrimination, and then the types are subjected to mode identification by combining the spectrogram characteristics of the signals.
The sampling data of the power equipment in a certain time is collected by adopting a broadband high-speed sampling mode in the I, and the sampling data specifically comprises the following steps:
1. waveform signal acquisition stage
Step 1: and acquiring the partial discharge signals for a period of time (preferably, 500 power frequency periods are 10 seconds) into a discrete time sequence by a high-speed sampling system (preferably, the sampling rate is 100 MHz).
Step 2: comparing the amplitude of the acquired signal with a set pulse threshold (preferred value: 50% of the maximum value among all the acquired values); if the amplitude of the acquired signal is larger than the pulse amplitude threshold, storing the amplitude and the position information in the corresponding time sequence; otherwise, no storage is performed.
In II, converting the sampling data into a PRPD spectrogram, and performing feature extraction and classification on the PRPD spectrogram, wherein the method specifically comprises the following steps:
2. automatic separation and classification (coarse classification) stage
And step 3: reading the stored sampling points, determining a first point P of the pulse, which is larger than a pulse threshold value, determining a starting point and an end point of the pulse, (preferably, the data from the starting point to the end point are all 2 microseconds, and P is positioned in the middle position) and storing the waveform of the whole pulse, and taking the maximum value of the pulse amplitude as the peak value of the pulse; and converting the corresponding time series to determine the partial discharge phase, and drawing a phase distribution spectrogram, as shown in fig. 4.
And 4, step 4: the P points are respectively used for finding a time position with the amplitude of 0 or the closest to 0 forwards and backwards, the former is the A point, the latter is the B point, and the waveform data between the AB points is called the main peak of the pulse waveform. The main peak shape characteristics actually determine the properties of the waveform.
And 5: calculating the main peak shape characteristic value in fig. 5 (see table 1), the calculation formula is shown in formulas 1 to 5:
TABLE 1
In the formula, SkfIs the degree of skewness, xiFor frequency domain waveform point-by-point data, μ is all xiMean value ofσ is the corresponding standard deviation, i is the data number, and W is the total number of data.
In the formula, KufIs steep.
In the formula, tjAnd gi(tj) Respectively representing time and amplitude in the time domain of a single pulse signal, m being the total number of times, gi 2(tj) Representing the square of the time domain amplitude of a single pulse signal; f. ofjAnd gi(fj) Respectively representing the frequency (m/2-1 points) and amplitude, g, of a single pulse signali 2(fj) Representing the square of the frequency domain amplitude of a single pulse signal, j representing the index number, T0 iTo a desired value of time, F0 iIs the frequency expected value.
In the formula, TiTo equivalent duration, FiIs the equivalent bandwidth.
In the formula, Ti 2Is a second equivalent time length coefficient, gi 4(tj) Representing the fourth power, F, of the amplitude in the time domain of a single pulse signali 2Is the second order equivalent bandwidth coefficient.
Step 6: and (3) performing clustering analysis on the time domain feature vectors extracted in the step (5) by using a clustering method, classifying data in the PRPD spectrogram by using a K-multi-mean algorithm as shown in fig. 6, eliminating the classification to which the noise information belongs, and reserving the information of the partial discharge signals in the PRPD spectrogram for pattern recognition of each classified PRPD spectrogram subsequently. In the automatic classification and separation technology mentioned in the invention, the corresponding K-multi-mean algorithm can be replaced by a series of derivative methods which are optimized based on the K-means algorithm.
In step III, substituting the statistical features in the signal features extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power device has partial discharge within the certain time and the partial discharge type to which the power device belongs, specifically including:
3. automatic identification stage for partial discharge type
And (3) extracting the statistical characteristic quantity of the PRPD spectrogram aiming at each classified PRPD spectrogram, and performing pattern recognition by adopting a neural network algorithm as shown in fig. 7 and 8. The BP neural network method referred to in the present invention may be replaced with other neural network methods. The extracted main statistical characteristic quantities include:
TABLE 2 statistical characteristic parameters
TABLE 3 partial discharge Signal characteristic parameters
The calculation formula is shown as formula (6) to formula (12):
(1) degree of deflection
In the formula, SkFor skewness, the abscissa phase is equally divided into W phase windows (preferred value of W is 50),
represents the phase of the ith phase window,
representing the width of the phase window. Handle
As a probability density profile, piσ, μ are the probability, standard deviation and mean, respectively, of the occurrence of an event for the phase window i.
(2) Degree of steepness
In the above formula, the first and second carbon atoms are,
represents the phase of the ith phase window,
representing the width of the phase window. Handle
As a probability density distribution map, then piσ, μ are the probability, standard deviation and mean, respectively, of the occurrence of an event for the phase window i. The formula is as follows:
wherein y isiRepresents the ordinate of the spectrum, which represents the discharge quantity q or the number of discharges n.
(3) Discharge factor
In the formula, q + and q-respectively represent discharge of positive and negative half cycles, the value range of the abscissa 'phase' of a discharge spectrogram is 0 to 360 degrees, wherein 0-180 degrees is called as a positive half cycle, 180-360 degrees is called as a negative half cycle,
points in positive and negative half cycles are indicated.
(4) Cross correlation coefficient
In the above formula
The average discharge amount in the positive and negative half cycle phase windows i and the number of phase windows in which n is a positive half cycle or a negative half cycle are shown.
Fig. 9 is a pulse diagram of continuously acquiring data for 10 seconds, and fig. 10 is a pulse diagram of 2 microseconds in length. The PRPD spectrogram is used for extracting the characteristics of the partial discharge signal. From the general PRPD spectrum in fig. 11, it is difficult to distinguish the discharge type without clear phase characteristics, and the judgment will cause erroneous judgment. The discharge types here include internal discharge, surface discharge and corona discharge.
After waveform discrimination and classification by the method adopted by the invention, the PRPD spectrogram corresponding to the red cluster group of the classification 1 is shown in figure 12, and the spectrogram is a signal with fixed phase characteristic and is identified as discharge. The PRPD spectrum corresponding to the black cluster family classified in 2 is shown in fig. 13, and the spectrum is a non-fixed phase characteristic signal and is identified as noise.
The method for identifying the partial discharge of the power equipment can realize accurate and deep identification of the partial discharge type of the power equipment, thereby determining the property and the deterioration degree of the partial discharge, reducing the dependence on professionals and avoiding omission and errors caused by manual selection. Through characteristic extraction, cluster analysis and model training of typical partial discharge spectrum samples and verification of a plurality of test examples, the partial discharge waveform recognition success rate can reach more than 85%, and the spectrum recognition success rate can reach more than 95%.
The power equipment partial discharge recognition system is directly deployed on an intelligent sensor, a marginal Internet of things agent or a cloud end, or needs protection by adopting a technical framework of layered collaborative deployment of cloud edge ends, and is used for realizing partial discharge recognition.
The present invention solves the following three problems:
since the field interference is large and much, the separation and classification of the partial discharge and the noise source is the key point for the success of the field high-frequency partial discharge test. By applying the advanced separation and classification technology, the problem of false alarm caused by interference can be reduced on the basis of a threshold value alarm mode, and meanwhile, the missing report caused by partial discharge false deletion caused by noise elimination is avoided. The automatic separation and classification setting reduces the dependence on professionals, avoids omission and errors caused by manual selection, is particularly suitable for the partial discharge online monitoring system, can complete the processing, analysis and judgment of partial discharge data by operation and maintenance personnel, and solves the problem that the success rate is low because partial discharge map identification is directly carried out by adopting a neural network algorithm.
Since the main energy of the partial discharge spectrum of the power cable is concentrated in a high-frequency region, the main characteristic quantity information can be obtained by adopting high-frequency partial discharge sampling. The ultra-wideband measurement can cover low frequency, extends the measurement range to the tail end of a power cable, solves the problem of partial discharge signal attenuation, can obtain complete pulse waveforms, separates and classifies noise, and combines statistical characteristics to perform depth identification of discharge types.
Example 2
The present application further provides a power equipment partial discharge recognition system, including:
the data acquisition module is used for acquiring sampling data of the power equipment within a certain time by adopting a broadband high-speed sampling mode;
the separation and classification module is used for converting the sampling data into a PRPD spectrogram and extracting and classifying signal features of the PRPD spectrogram;
the partial discharge identification module is used for substituting statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge identification model corresponding to the type to obtain whether the power equipment has partial discharge within the certain time and the partial discharge type of the power equipment;
the partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
The training of the partial discharge recognition model comprises the following steps:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
The separation and classification module comprises:
a conversion submodule for determining first and last sampled data from the sampled data according to a time sequence, as a start point and an end point, respectively; taking a point located in the middle of the starting point and the end point as an intermediate point; finding a time position with the amplitude value of 0 or closest to 0 forwards and backwards respectively by taking the middle point as a center; constructing a PRPD spectrogram by using waveform data between the time positions with the amplitude value of 0 or the time position closest to 0 as a main peak of a pulse waveform;
a feature extraction submodule, configured to calculate a skewness, a steepness, a time expected value, a frequency expected value, an equivalent duration calculation formula, and an equivalent bandwidth calculation formula of the main peak based on a pulse waveform in the PRPD spectrogram by using the skewness calculation formula, the steepness calculation formula, the time expected value, the frequency expected value, the equivalent duration, and the equivalent bandwidth calculation formula, respectively;
and the classification submodule is used for performing cluster analysis on the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the time domain feature vector of the equivalent frequency width of the main peak by using a cluster analysis method to obtain a PRPD spectrogram which eliminates the noise and retains the partial discharge signal and a PRPD spectrogram which only contains the noise.
For convenience of description, each part of the above apparatus is separately described as each module or unit by dividing the function. Of course, the functionality of the various modules or units may be implemented in the same one or more pieces of software or hardware when implementing the present application.
Based on the same inventive concept, in yet another embodiment of the present invention, a computer apparatus is provided, which includes a processor and a memory for storing a computer program comprising program instructions, and the processor is configured to execute the program instructions stored by the computer storage medium. The Processor may be a Central Processing Unit (CPU), or may be other general-purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable gate array (FPGA) or other Programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc., which is a computing core and a control core of the terminal, and is specifically adapted to load and execute one or more instructions in a computer storage medium to implement a corresponding method flow or a corresponding function; the processor according to the embodiment of the invention can be used for executing the steps of the power equipment partial discharge identification method.
Based on the same inventive concept, in yet another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a Memory device in a computer device and is used for storing programs and data. It is understood that the computer readable storage medium herein can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer-readable storage medium may be a high-speed RAM memory, or may be a non-volatile memory (non-volatile memory), such as at least one disk memory. One or more instructions stored in the computer-readable storage medium may be loaded and executed by a processor to implement the corresponding steps of the power equipment partial discharge identification method in the above-described embodiments.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The present invention is not limited to the above embodiments, and any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention are included in the scope of the claims of the present invention which are filed as the application.
Claims (15)
1. A power equipment partial discharge identification method is characterized by comprising the following steps:
sampling data of the power equipment within a certain time is collected by adopting a broadband high-speed sampling mode;
converting the sampling data into a PRPD spectrogram, and extracting and classifying signal features of the PRPD spectrogram;
substituting the statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has partial discharge and the partial discharge type of the power equipment in the certain time;
the trained partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
2. The method of claim 1, wherein the training of the partial discharge recognition model comprises:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
3. The method of claim 2, wherein the cluster analysis method comprises: a K-means algorithm and a modified K-means algorithm.
4. The method of claim 1, wherein said converting said sampled data into a PRPD profile comprises:
determining first and last sampling data from the sampling data according to the time sequence, wherein the first and last sampling data are respectively used as a starting point and an end point;
taking a point located in the middle of the starting point and the end point as an intermediate point;
finding a time position with the amplitude value of 0 or closest to 0 forwards and backwards respectively by taking the middle point as a center;
a PRPD spectrum is constructed by the waveform data between the time positions with the amplitude of 0 or the closest 0, which is called the main peak of a pulse waveform.
5. The method of claim 1, wherein said signal feature extracting and classifying said PRPD spectrogram comprises:
calculating the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the equivalent bandwidth of the main peak respectively by adopting a skewness calculation formula, a steepness calculation formula, a time expected value calculation formula, a frequency expected value calculation formula, an equivalent duration calculation formula and an equivalent bandwidth calculation formula based on the pulse waveform in the PRPD spectrogram;
and performing cluster analysis on the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the time domain feature vector of the equivalent frequency width of the main peak by using a cluster analysis method to obtain a PRPD spectrogram which rejects the noise and retains the partial discharge signal and a PRPD spectrogram which only contains the noise.
6. The method of claim 4, wherein the skewness calculation is as follows:
in the formula, SkfIs the degree of skewness, xiFor the point-by-point data of the frequency domain waveform, μ is all xiThe average value of (1) is the number, W is the total number, and σ is the corresponding standard deviation.
7. The method of claim 6, wherein the steepness calculation is as follows:
in the formula, KufIs steep.
8. The method of claim 6, wherein the time desired value calculation and the frequency desired value calculation are as follows:
in the formula, tjAnd gi(tj) Respectively representing time and amplitude in the time domain of a single pulse signal, m being the total number of times, gi 2(tj) Representing the square of the time domain amplitude of a single pulse signal; f. ofjAnd gi(fj) Respectively representing the frequency and amplitude of a single pulse signal, gi 2(fj) Representing the square of the frequency domain amplitude of a single pulse signal, j representing the index number, T0 iTo a desired value of time, F0 iIs the frequency expected value.
9. The method of claim 8, wherein the equivalent duration calculation and equivalent bandwidth calculation are as follows:
in the formula, TiTo equivalent duration, FiIs the equivalent bandwidth.
10. The method as claimed in claim 1, wherein the bringing statistical features in the signal features extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type to obtain whether the power equipment has a partial discharge within the certain time and the partial discharge type includes:
and substituting the statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge recognition model corresponding to the type, if the statistical characteristics are consistent with phase characteristic signals in the partial discharge recognition model, recognizing the partial discharge and outputting a partial discharge type, otherwise, recognizing the partial discharge type as noise.
11. An electrical device partial discharge identification system, comprising:
the data acquisition module is used for acquiring sampling data of the power equipment within a certain time by adopting a broadband high-speed sampling mode;
the separation and classification module is used for converting the sampling data into a PRPD spectrogram and extracting and classifying signal features of the PRPD spectrogram;
the partial discharge identification module is used for substituting statistical characteristics in the signal characteristics extracted from the PRPD spectrogram of each type into a pre-trained partial discharge identification model corresponding to the type to obtain whether the power equipment has partial discharge within the certain time and the partial discharge type of the power equipment;
the partial discharge recognition model is obtained by training statistical characteristics extracted from each type of PRPD spectrogram and discharge types to which the statistical characteristics belong by adopting a neural network.
12. The system of claim 11, wherein the training of the partial discharge recognition model comprises:
constructing a training set by using the statistical characteristics extracted from each type of PRPD spectrogram and the discharge type to which the statistical characteristics belong;
inputting the statistical characteristics and the discharge types of the training sets into a neural network for training to obtain a trained partial discharge recognition model corresponding to the PRPD spectrogram of the type.
13. The system of claim 11, wherein the separation classification module comprises:
a conversion submodule for determining first and last sampled data from the sampled data according to a time sequence, as a start point and an end point, respectively; taking a point located in the middle of the starting point and the end point as an intermediate point; finding a time position with the amplitude value of 0 or closest to 0 forwards and backwards respectively by taking the middle point as a center; constructing a PRPD spectrogram by using waveform data between the time positions with the amplitude value of 0 or the time position closest to 0 as a main peak of a pulse waveform;
a feature extraction submodule, configured to calculate a skewness, a steepness, a time expected value, a frequency expected value, an equivalent duration calculation formula, and an equivalent bandwidth calculation formula of the main peak based on a pulse waveform in the PRPD spectrogram by using the skewness calculation formula, the steepness calculation formula, the time expected value, the frequency expected value, the equivalent duration, and the equivalent bandwidth calculation formula, respectively;
and the classification submodule is used for performing cluster analysis on the skewness, the steepness, the time expected value, the frequency expected value, the equivalent duration and the time domain feature vector of the equivalent frequency width of the main peak by using a cluster analysis method to obtain a PRPD spectrogram which eliminates the noise and retains the partial discharge signal and a PRPD spectrogram which only contains the noise.
14. A computer device, comprising:
one or more processors;
a processor for executing one or more programs;
the one or more programs, when executed by the one or more processors, implement a power device partial discharge identification method as recited in any of claims 1-10.
15. A computer-readable storage medium, having stored thereon a computer program which, when executed, implements a power device partial discharge identification method as claimed in any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111411547.7A CN114325256A (en) | 2021-11-25 | 2021-11-25 | Power equipment partial discharge identification method, system, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111411547.7A CN114325256A (en) | 2021-11-25 | 2021-11-25 | Power equipment partial discharge identification method, system, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114325256A true CN114325256A (en) | 2022-04-12 |
Family
ID=81046124
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111411547.7A Pending CN114325256A (en) | 2021-11-25 | 2021-11-25 | Power equipment partial discharge identification method, system, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114325256A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115186772A (en) * | 2022-09-13 | 2022-10-14 | 云智慧（北京）科技有限公司 | Method, device and equipment for detecting partial discharge of power equipment |
| CN115828148A (en) * | 2023-02-23 | 2023-03-21 | 广州智丰电气科技有限公司 | Partial discharge waveform identification method, system and storage medium |
| CN117054797A (en) * | 2023-10-12 | 2023-11-14 | 国网天津市电力公司电力科学研究院 | Method, device, apparatus and medium for defect detection of electrical device |
| CN117708685A (en) * | 2024-02-05 | 2024-03-15 | 国网江西省电力有限公司电力科学研究院 | Method and system for identifying latent partial discharge defects of switch cabinet |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202041614U (en) * | 2011-03-18 | 2011-11-16 | 山西省电力公司太原供电分公司 | Insulation defect local discharging signal identifying device of ultrahigh voltage cable line |
| CN107942210A (en) * | 2017-11-14 | 2018-04-20 | 国网上海市电力公司 | The classification of transformer pulse electric current Partial Discharge and denoising method and system |
| CN108693448A (en) * | 2018-03-28 | 2018-10-23 | 西安博源电气有限公司 | One kind being applied to power equipment PD Pattern Recognition system |
| CN109116203A (en) * | 2018-10-31 | 2019-01-01 | 红相股份有限公司 | Power equipment partial discharges fault diagnostic method based on convolutional neural networks |
| CN109212391A (en) * | 2018-09-15 | 2019-01-15 | 四川大学 | Take into account the signal processing of partial discharge method and power cable partial discharge positioning method of DISCHARGE PULSES EXTRACTION and signal denoising |
| CN109685138A (en) * | 2018-12-25 | 2019-04-26 | 东南大学 | A kind of XLPE power cable shelf depreciation kind identification method |
| CN111537850A (en) * | 2020-05-21 | 2020-08-14 | 北京传动联普科技有限公司 | Big data identification management system for intelligent separation and classification diagnosis of partial discharge signals |
| CN112763871A (en) * | 2020-12-30 | 2021-05-07 | 珠海华网科技有限责任公司 | Partial discharge classification identification method |
-
2021
- 2021-11-25 CN CN202111411547.7A patent/CN114325256A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202041614U (en) * | 2011-03-18 | 2011-11-16 | 山西省电力公司太原供电分公司 | Insulation defect local discharging signal identifying device of ultrahigh voltage cable line |
| CN107942210A (en) * | 2017-11-14 | 2018-04-20 | 国网上海市电力公司 | The classification of transformer pulse electric current Partial Discharge and denoising method and system |
| CN108693448A (en) * | 2018-03-28 | 2018-10-23 | 西安博源电气有限公司 | One kind being applied to power equipment PD Pattern Recognition system |
| CN109212391A (en) * | 2018-09-15 | 2019-01-15 | 四川大学 | Take into account the signal processing of partial discharge method and power cable partial discharge positioning method of DISCHARGE PULSES EXTRACTION and signal denoising |
| CN109116203A (en) * | 2018-10-31 | 2019-01-01 | 红相股份有限公司 | Power equipment partial discharges fault diagnostic method based on convolutional neural networks |
| CN109685138A (en) * | 2018-12-25 | 2019-04-26 | 东南大学 | A kind of XLPE power cable shelf depreciation kind identification method |
| CN111537850A (en) * | 2020-05-21 | 2020-08-14 | 北京传动联普科技有限公司 | Big data identification management system for intelligent separation and classification diagnosis of partial discharge signals |
| CN112763871A (en) * | 2020-12-30 | 2021-05-07 | 珠海华网科技有限责任公司 | Partial discharge classification identification method |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115186772A (en) * | 2022-09-13 | 2022-10-14 | 云智慧（北京）科技有限公司 | Method, device and equipment for detecting partial discharge of power equipment |
| CN115186772B (en) * | 2022-09-13 | 2023-02-07 | 云智慧（北京）科技有限公司 | Method, device and equipment for detecting partial discharge of power equipment |
| CN115828148A (en) * | 2023-02-23 | 2023-03-21 | 广州智丰电气科技有限公司 | Partial discharge waveform identification method, system and storage medium |
| CN117054797A (en) * | 2023-10-12 | 2023-11-14 | 国网天津市电力公司电力科学研究院 | Method, device, apparatus and medium for defect detection of electrical device |
| CN117054797B (en) * | 2023-10-12 | 2024-01-26 | 国网天津市电力公司电力科学研究院 | Method, device, apparatus and medium for defect detection of electrical device |
| CN117708685A (en) * | 2024-02-05 | 2024-03-15 | 国网江西省电力有限公司电力科学研究院 | Method and system for identifying latent partial discharge defects of switch cabinet |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114325256A (en) | Power equipment partial discharge identification method, system, equipment and storage medium | |
| CN111626153B (en) | Integrated learning-based partial discharge fault state identification method | |
| CN110108992B (en) | Cable partial discharge fault identification method and system based on improved random forest algorithm | |
| Peng et al. | Random forest based optimal feature selection for partial discharge pattern recognition in HV cables | |
| US20230123117A1 (en) | Method and Apparatus for Inspecting Wind Turbine Blade, And Device And Storage Medium Thereof | |
| CN102426835B (en) | Method for identifying local discharge signals of switchboard based on support vector machine model | |
| CN108693448B (en) | Partial discharge mode recognition system applied to power equipment | |
| CN104198898A (en) | Local discharge development process diagnosis method based on pulse-train analysis | |
| CN109633368A (en) | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA | |
| CN101995437B (en) | Method for extracting features of crack acoustic emission signal of drawing part | |
| CN103558519A (en) | GIS partial discharge ultrasonic signal identification method | |
| CN110543921A (en) | cable early fault identification method based on waveform learning | |
| CN216848010U (en) | Cable partial discharge online monitoring device for edge calculation | |
| CN112462355A (en) | Sea target intelligent detection method based on time-frequency three-feature extraction | |
| CN114217164B (en) | Cable fault distance measurement method and system based on discharge waveform intelligent identification | |
| CN116432112A (en) | Arc fault detection method based on wavelet packet transformation and residual convolution neural network | |
| CN108594156B (en) | Improved current transformer saturation characteristic identification method | |
| CN116628620A (en) | Non-invasive load identification calculation method | |
| CN111025100A (en) | Transformer ultrahigh frequency partial discharge signal mode identification method and device | |
| CN115128345A (en) | Power grid safety early warning method and system based on harmonic monitoring | |
| CN115902557A (en) | Switch cabinet fault diagnosis processing method and device and nonvolatile storage medium | |
| Fan et al. | Research on partial discharge identification of power transformer based on chaotic characteristics extracted by GP algorithm | |
| CN109917245B (en) | Ultrasonic detection partial discharge signal mode identification method considering phase difference | |
| CN114091593A (en) | Network-level arc fault diagnosis method based on multi-scale feature fusion | |
| CN104102834A (en) | Method for identifying sound recording locations |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |