CN112036320A - Partial discharge state identification method and device, discharge simulator and identification equipment - Google Patents
Partial discharge state identification method and device, discharge simulator and identification equipment Download PDFInfo
- Publication number
- CN112036320A CN112036320A CN202010902297.6A CN202010902297A CN112036320A CN 112036320 A CN112036320 A CN 112036320A CN 202010902297 A CN202010902297 A CN 202010902297A CN 112036320 A CN112036320 A CN 112036320A
- Authority
- CN
- China
- Prior art keywords
- signal
- identified
- sdp
- discharge
- partial discharge
- 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
- 230000036961 partial effect Effects 0.000 title claims abstract description 98
- 238000000034 method Methods 0.000 title claims abstract description 68
- 230000004927 fusion Effects 0.000 claims abstract description 70
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 63
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 23
- 238000004458 analytical method Methods 0.000 claims abstract description 20
- 230000010355 oscillation Effects 0.000 claims abstract description 11
- 238000012549 training Methods 0.000 claims description 46
- 238000012360 testing method Methods 0.000 claims description 29
- 238000010586 diagram Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 12
- 239000000443 aerosol Substances 0.000 claims description 11
- 238000010168 coupling process Methods 0.000 claims description 9
- 238000005859 coupling reaction Methods 0.000 claims description 9
- 230000015556 catabolic process Effects 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 7
- 239000003990 capacitor Substances 0.000 claims description 6
- 230000008878 coupling Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 3
- 230000001681 protective effect Effects 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 2
- 239000003570 air Substances 0.000 description 15
- 238000009413 insulation Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 230000007547 defect Effects 0.000 description 5
- 238000001514 detection method Methods 0.000 description 5
- 239000000284 extract Substances 0.000 description 5
- 239000004020 conductor Substances 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 3
- 230000005684 electric field Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000001052 transient effect Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 239000011800 void material Substances 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 239000012080 ambient air Substances 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000003703 image analysis method Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to a partial discharge state identification method, a partial discharge state identification device, a discharge simulator, identification equipment and a storage medium, wherein the method comprises the following steps: acquiring an original transformer signal as a signal to be identified; separating an oscillation component and an impact component of a signal to be identified through resonance sparse decomposition to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified; carrying out information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method to obtain an SDP fusion image of the signal to be identified; and inputting the SDP fusion image of the signal to be recognized into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized. The method realizes efficient and accurate automatic identification of the partial discharge signal, and improves the characteristic learning effect and the state identification precision of the partial discharge signal.
Description
Technical Field
The invention relates to the technical field of transformer fault detection, in particular to a partial discharge state identification method and device, a discharge simulator, identification equipment and a storage medium.
Background
The phenomenon that a Discharge occurs only in a local region of an insulator, but does not penetrate between conductors to which a voltage is applied, may occur in the vicinity of the conductors or elsewhere, and is called a Partial Discharge (PD). Partial discharge is a significant cause of the ultimate insulation breakdown of high-voltage electrical equipment and also a significant sign of insulation degradation, and thus is one of the main features of insulation defects in power transformer systems.
Normally, partial discharges do not cause overall insulation damage in the short term, but as partial discharges develop and propagate, they can lead to insulation failure and more serious system failures. And different PD defects may cause various damages to the electric appliance. Therefore, monitoring, analysis and identification of the transformer PD are of great significance for early detection and location of equipment faults.
However, the partial discharge signal is a complex multi-modal signal, and the processing capability of the prior art is very good for complex signals with a large and close frequency component, and the extracted state features are difficult to meet the requirements of fault diagnosis. And greatly depends on the accuracy of feature extraction in the identification process. Therefore, the state recognition based on different feature extraction methods and recognition methods is limited.
Disclosure of Invention
In view of the above, the present invention provides a method and an apparatus for identifying a partial discharge state, a discharge simulator, an identification device, and a storage medium, so as to solve the defects of low processing capability and dependency on feature extraction accuracy in the partial discharge state identification in the prior art. .
In order to achieve the purpose, the invention adopts the following technical scheme:
in a first aspect, an embodiment of the present application provides a partial discharge state identification method, where the method includes:
acquiring an original transformer signal as a signal to be identified;
separating an oscillation component and an impact component of the signal to be identified through resonance sparse decomposition to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified;
carrying out information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method so as to obtain an SDP fusion image of the signal to be identified;
and inputting the SDP fusion image of the signal to be recognized to a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized.
In a second aspect, an embodiment of the present application provides a partial discharge state identification apparatus, including:
the to-be-identified signal acquisition module is used for acquiring an original transformer signal as a to-be-identified signal;
the characteristic extraction module is used for separating an oscillation component and an impact component of the signal to be identified through resonance sparse decomposition so as to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified;
the image fusion module is used for performing information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method so as to obtain an SDP fusion image of the signal to be identified;
and the discharge state identification module is used for inputting the SDP fusion image of the signal to be identified into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be identified.
In a third aspect, an embodiment of the present application provides a discharge simulator, including an ac power supply, a step-up transformer, a protection resistor, a coupling capacitor, a high-voltage bushing, a small bushing, a transformer partial discharge model, a current sensor, and a waveform discharge analyzer;
the booster transformer comprises a no-corona experimental transformer and an auto-coupling voltage regulator;
the current sensor is used for detecting a current signal and transmitting the current signal to the waveform discharge analyzer;
the waveform discharge analyzer samples the current signal to obtain a first sample signal;
the working process of the discharge simulator comprises the following steps:
adjusting the discharge simulator to a normal operation state;
measuring the voltage at the beginning of breakdown, the voltage at the beginning of discharge and background noise in the operation process;
raising the experimental voltage, and taking the voltage when the waveform discharge analyzer displays electric pulse as the initial model discharge voltage;
when the partial discharge condition meets a stable condition, recording the discharge voltage as a second sample signal;
the first sample signal and the second sample signal constitute the sample signal.
In a fourth aspect, an embodiment of the present application provides a partial discharge state identification device, including:
a processor, and a memory coupled to the processor;
the processor is used for receiving a sample signal from the discharge simulator;
the memory is configured to store a computer program, where the computer program is at least configured to execute the partial discharge state identification method according to the first aspect of the embodiment of the present application;
the processor is used for calling and executing the computer program in the memory.
In a fifth aspect, the present application provides a storage medium, where the storage medium stores a computer program, and the computer program, when executed by a processor, implements the steps in the partial discharge state identification method according to the first aspect.
By adopting the technical scheme, the high resonance component and the low resonance component of the signal can reflect different fault states of the transformer, the SDP fusion can extract the characteristics of different signals in an image mode, so that the difference between different running states of the transformer is reflected, and finally, the SDP fusion image is identified by using a CNN network. Compared with other measuring methods of partial discharge signals, the technical scheme of the embodiment of the application is that the time sequence series is measured for the same point, then an effective feature extraction method is selected, effective features in original signals are extracted, and the loss of the original signals is avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart of a partial discharge state identification method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of an SDP method suitable for use in embodiments of the present invention;
fig. 3 is a schematic structural diagram of a CNN applicable to the embodiment of the present invention;
FIG. 4 is a structural diagram of a CNN fault diagnosis model based on resonance sparse decomposition feature information fusion, which is applicable to an embodiment of the present invention;
fig. 5 is a graph of accuracy of CNN test results verified by applying a test sample signal according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of high and low resonant components of a corona discharge suitable for use in embodiments of the invention;
FIG. 7 is a schematic representation of the high and low resonant components of an air gap discharge suitable for use in embodiments of the present invention;
FIG. 8 is a schematic illustration of the high and low resonant components of an aerosol discharge suitable for use in embodiments of the present invention;
fig. 9 is a highly resonant SDP plot of three partial discharge signals suitable for use in embodiments of the present invention;
fig. 10 is a low resonant SDP plot of three partial discharge signals suitable for use in embodiments of the present invention;
fig. 11 is a schematic structural diagram of a partial discharge state identification apparatus suitable for use in an embodiment of the present invention;
FIG. 12 is a schematic diagram of a discharge simulator suitable for use in embodiments of the present invention;
FIG. 13 is a diagram of a practical application of a discharge simulator suitable for use in embodiments of the present invention;
fig. 14 is a schematic structural diagram of a partial discharge recognition apparatus applied in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without any inventive step, are within the scope of the present invention.
Fig. 1 is a flowchart of a partial discharge state identification method according to an embodiment of the present invention, where the method may be performed by a partial discharge state identification apparatus according to an embodiment of the present invention, and the apparatus may be implemented in a software and/or hardware manner. Referring to fig. 1, the method may specifically include the following steps:
and S101, acquiring an original transformer signal as a signal to be identified.
Specifically, the signal to be identified is a voltage or current signal generated in the actual operation process of the transformer, and the partial discharge state is judged by judging the real-time transformer signal, so that the insulation defect and the system fault are analyzed. For example, the original transformer signal is obtained as the signal to be identified, and the obtaining may be performed by performing signal extraction through a special signal detection device.
S102, carrying out separation of oscillation components and impact components on the signal to be identified through resonance sparse decomposition so as to extract high resonance components of the signal to be identified and low resonance components of the signal to be identified.
The resonance sparse decomposition is to distinguish different components according to the resonance attribute of the signal component, and the resonance attribute is represented by Q. And decomposing the complex signal into a high resonance component consisting of a continuous oscillation component and a low resonance component consisting of a transient impact component to realize sparse decomposition of the signal. The separation of resonance components and impact components is effectively realized, high-low resonance characteristics can be effectively extracted according to the characteristics of signals, and the characteristics of partial discharge signals are completely maintained.
S103, carrying out information fusion on the high resonance component of the signal to be recognized and the low resonance component of the signal to be recognized by applying an SDP analysis method to obtain an SDP fusion image of the signal to be recognized.
Among them, in the SDP (symmetric Dot Pattern) algorithm, x is a sampling of discrete data of a signaliAnd xi+1The amplitudes at time i and at time i +1, respectively, discrete data can be transformed into points in polar coordinate space P (r (i), θ (i), φ (i)) by the transformation formula of the symmetric point pattern (SDP) method. This makes the difference between the signals more distinctive. The SDP analysis can map the partial discharge signal into a visual expression form, so that the loss of characteristic information is avoided, an SDP fusion image after the characteristic information fusion shows the characteristics of the partial discharge signal more clearly, visually and comprehensively, and the differentiability between the same-state characteristics is greatly improved.
In a specific example, fig. 2 shows a schematic diagram of an SDP method, and the specific calculation formula of the variables is as follows:
xminand xmaxIs the maximum and minimum values of the window of the original waveform, thetasIs a designated mirror symmetry plane sthRotation angle given by θs360i/s, where s1, 2, n, n is the number of mirror symmetry planes, ξ is the gain of the plot, and ξ ≦ θ. Compared with other image analysis methods, the SDP graph enlarges the characteristics of each signal, and shows the difference of different partial discharge characteristics more clearly.
And S104, inputting the SDP fusion image of the signal to be recognized into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized.
Among them, the Convolutional Neural Network (CNN) is a feedforward Neural Network, and its artificial neurons can respond to peripheral units in a part of coverage range, and the effect is good for processing large-scale images. Specifically, a pre-constructed convolutional neural network model is trained, and when a certain condition is met, it is indicated that the trained convolutional neural network model can be applied, which is called a pre-trained convolutional neural network. At the moment, the SDP fusion image of the signal to be recognized is input to the pre-trained convolutional neural network model, and the partial discharge state corresponding to the signal to be recognized is obtained. The CNN model after parameter optimization can adaptively identify the characteristics of the SDP fusion image, obtain accurate classification and realize efficient and accurate intelligent identification of the partial discharge state.
Optionally, the partial discharge state includes: corona discharge, air gap discharge, and aerosol discharge. Among them, corona discharge refers to local self-sustaining discharge of a gas medium in a non-uniform electric field, and is one of the most common gas discharge forms. In the vicinity of the tip electrode having a small radius of curvature, the gas is ionized and excited due to the local electric field strength exceeding the ionization field strength of the gas, and thus corona discharge occurs. The corona discharge may be a relatively stable form of discharge or may be an early stage in the breakdown process of the non-uniform electric field gap. The air gap discharge refers to the situation that air between two conductors is broken down, the gas discharge is similar to the situation that a single conductor directly discharges to the air due to high ambient air humidity or high body voltage, and the air gap discharge is one of the gas discharges. The suspended particle discharge means that suspended particles in the air are deposited on the surface of a wire due to factors such as charge or airflow, and the like, and influence is caused on direct corona discharge and the electromagnetic environment of a high-voltage direct-current transmission line.
In the embodiment of the application, the high resonance component and the low resonance component of the signal can reflect different fault states of the transformer, the SDP fusion can extract the characteristics of different signals in an image mode, so that the difference between different running states of the transformer is reflected, and finally, a CNN network is used for identifying the SDP fusion image. Compared with other measuring methods of partial discharge signals, the technical scheme of the embodiment of the application is that the time sequence series is measured for the same point, then an effective feature extraction method is selected, effective features in original signals are extracted, and the loss of the original signals is avoided.
In addition, the SDP graph fusing the characteristics amplifies the distinguishability of the characteristics of local signals, and the differences among the signals can be highlighted by utilizing the graph to represent the characteristics of the signals, so that the accurate characteristics of the signals are ensured. The complexity and the non-stationarity of the signal are solved, namely, the required features can be extracted through decomposition. The resonance sparse decomposition can better strip out the required characteristic quantity.
Illustratively, the training process of the convolutional neural network model can be implemented as follows: obtaining a sample signal by using a discharge simulator; randomly dividing the sample signal into a training sample signal and a training sample signal; extracting the characteristics of the training sample signal to obtain a high resonance component of the training sample signal and a low resonance component of the training sample signal; fusing a high resonance component of the training sample signal and a low resonance component of the training sample signal by applying an SDP analysis method to obtain an SDP fusion image corresponding to the training sample signal; training a pre-constructed convolutional neural network model by using an SDP fusion image corresponding to the training sample signal and the partial discharge state of the training sample signal; and optimizing each layer of parameters of the pre-constructed convolutional neural network model until the recognition accuracy is greater than a set accuracy threshold value, so as to obtain the trained convolutional neural network model.
Specifically, in the training process of the model, a large number of training sample signals are required, and therefore, the discharge simulator is applied to generate the sample signals, and the sample signals generated by the discharge simulator are randomly divided into the training sample signals and the test sample signals. The training sample signal is used for training a pre-constructed convolutional neural network model, and then the test sample signal is used for verifying the effectiveness of the model. Illustratively, in order to ensure the accuracy of model training, the processing mode and application method of the sample signal in each link are the same as the processing method of the signal to be recognized. Therefore, according to resonance sparse decomposition, carrying out feature extraction on the sample signal to obtain a high resonance component of the sample signal and a low resonance component of the sample signal; and then fusing the high resonance component of the training sample signal and the low resonance component of the training sample signal by applying an SDP analysis method to obtain an SDP fusion image corresponding to the training sample signal. And finally, training the pre-constructed convolutional neural network model by using the SDP images of the training sample signals corresponding to the training sample signals and the known partial discharge state of the training sample signals.
In one specific example, assume 1000 training sample signals, classified into class 3 partial discharge states, corona discharge, air gap discharge, and aerosol discharge prevention. The training sample signals are then individually labeled, for example, with data for corona discharge labeled 1, air gap discharge labeled 2, and aerosol discharge labeled 3. In the training process of the model, parameters of each layer of the convolutional neural network model are gradually optimized until the recognition accuracy reaches a set accuracy threshold, the trained convolutional neural network model is obtained, and the convolutional neural network model can be directly used for recognizing the signal to be recognized of the original transformer. For example, the set accuracy threshold may be 95%.
In addition, the convolutional neural network model can be updated by applying a test sample signal, and in detail, after the trained convolutional neural network model is obtained, an SDP fusion image corresponding to the test sample signal is input to the trained convolutional neural network model for identification, so that a partial discharge state is obtained; and comparing the actual partial discharge state and the partial discharge state corresponding to the test sample signal, and updating the trained convolutional neural network model according to the comparison result.
Specifically, after a trained convolutional neural network model is obtained, the accuracy of the model is verified by using a test sample signal. Illustratively, after resonance sparse decomposition is performed on a test sample signal, an SDP fusion image corresponding to the test sample signal is input to a trained convolutional neural network model for identification, and a partial discharge state is obtained. Because the test sample signal is generated through experiments, the actual partial discharge state corresponding to the test sample signal is known, and at the moment, the identified discharge state and the actual partial discharge state are compared to determine consistency or inconsistency. And counting the percentage of the consistent cases to all cases to determine whether the convolutional neural network model needs to be updated. For example, when the percentage is less than the set percentage threshold, the model needs to be updated by continuing to optimize the model parameters.
On the basis of the technical scheme, an SDP analysis method is applied to perform information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified so as to obtain an SDP fusion image of the signal to be identified, and the method can be specifically realized in the following mode: applying an SDP method, mapping the time sequence of the high resonance component of the signal to be identified and the time sequence of the low resonance component of the signal to be identified to a polar coordinate graph, and generating a scatter diagram on the polar coordinate; and realizing information fusion according to the scatter diagram to obtain an SDP fusion image of the signal to be identified.
In practical application, the SDP algorithm maps a time series of high-resonance components and low-resonance components of a signal to be recognized onto a polar coordinate graph, and then generates a scatter plot on the polar coordinate. Therefore, information fusion is realized by each scatter diagram of the pattern, and an SDP fusion image of the signal to be identified can be obtained.
Optionally, the convolutional neural network includes an input layer, a convolutional layer, a sampling layer, a full-link layer, and an output layer. Illustratively, the convolutional neural network applied in the embodiment of the present application includes an input layer, a convolutional layer including a convolutional layer 1 and a convolutional layer 2, and a sampling layer including a downsampling layer 1 and a downsampling layer 2, a full connection layer, and an output layer.
The CNN essentially extracts and processes the features hidden in the input image data layer by layer, simulates feature differentiation through convolution, reduces the magnitude of network parameters through weight sharing, namely pooling of the convolution, and finally completes classification tasks through a traditional neural network. In a specific example, fig. 3 shows a schematic diagram of a CNN structure, where Input is an Input layer, Feature maps are Feature maps, Output is an Output layer, volumes are convolutional layers, and Subsampling is two downsampling layers, i.e., downsampling layer 1 and downsampling layer 2; fully connected is the Fully connected layer. In a specific example, the parameters of each layer of the convolutional neural network can be taken as values with reference to table 1.
TABLE 1 convolutional neural network model parameter settings Table
| Hierarchy level | Convolution kernel/step size | Characteristic diagram |
| Input layer | 128×128 | |
| Convolutional layer | 5×5/1 | 124×124×6 |
| Down- |
5×2×2 | 62×62×6 |
| Convolutional layer 2 | 5×5/1 | 58×58×16 |
| Down- |
2×2/2 | 29×29/16 |
| Full connection layer | 84 | |
| Output layer | 10 |
FIG. 4 shows a CNN fault diagnosis model structure based on resonance sparse decomposition feature information fusion. In the above process, CNN is a key part of feature learning, and the structure thereof directly affects the partial discharge state identification accuracy.
Fig. 5 shows a CNN experiment result accuracy curve verified by using a test sample signal. The horizontal axis represents the number of training times and the vertical axis represents the accuracy. It can be seen from fig. 5 that the accuracy of the model on the test sample signal is continuously improved as the training times are increased; at around 200 th, the accuracy of the model is nearly saturated.
In order to make the technical scheme of the present application clearer, the following describes the resonance sparse decomposition and SDP image fusion process.
Because the partial discharge signal mainly comprises transient impact components, some harmonic components with related quantity, external interference signals and the like, when the harmonic components, the impact components and the central frequency of background noise are overlapped, a general frequency band decomposition or linear filtering method cannot decompose the signal thoroughly, so that the signal can be decomposed through different signal resonance properties. According to the embodiment of the application, different components are distinguished according to the resonance attribute Q of the signal component, the complex signal is decomposed into a high resonance component consisting of a continuous oscillation component and a low resonance component consisting of a transient impact component, and sparse decomposition of the signal is achieved. Impact and step components in the partial discharge signals are effectively extracted through resonance sparse decomposition, and therefore the influence of interference components on the partial discharge analysis results is reduced. As shown in fig. 6, 7 and 8, after resonance sparse decomposition, low resonance component and high resonance component after original signal processing can be obtained when Q is 1 and Q is 3, respectively. Fig. 6 is a schematic diagram of high and low resonance components of corona discharge, fig. 7 is a schematic diagram of high and low resonance components of air gap discharge, and fig. 8 is a schematic diagram of high and low resonance components of aerosol discharge.
The high resonance component and the low resonance component of the signal can reflect different fault states of the transformer, and the SDP graph can extract the characteristics of different signals in an image mode, so that the difference between different running states of the transformer is reflected. Therefore, the partial discharge signals in 3 states are firstly subjected to resonance sparse decomposition, 5 samples of each state are randomly selected, 15 samples are obtained in total, and the selected samples are subjected to SDP feature analysis. The SDP map for high resonance is shown in fig. 9, and the SDP map for low resonance is shown in fig. 10.
The three partial discharge signals are Corona discharge, Air gap discharge and Aerosol discharge, respectively, that is, Corona discharge, Air gap discharge and Aerosol discharge.
It is found from fig. 9 that when Q is 3, that is, the high resonance component, the SDP map features of the three partial discharge signals are distinguishable to some extent, but the differences are not obvious enough, the petal structures and shapes are similar, and misdiagnosis is easily caused on the discharge types. As shown in fig. 10, when Q is 1, that is, the low resonance component, the obtained SDP map features have respective shape features, and the difference is mainly expressed in the concentration region of the petal points: the point of corona discharge is mainly concentrated in the edge area of the petals; the air gap discharge point tiling range is large, and the whole petal area is filled; the point of aerosol discharge is mainly concentrated in the central area of the petals and the points are relatively concentrated. Compared with the SDP graph when Q is 3, the SDP graph when Q is 1 has obvious difference, and the manual identification can basically distinguish the discharge type. And carrying out SDP map fusion on the high and low resonance components so as to make up for the defect of single characteristic component information missing. The low resonance in corona discharge, air gap discharge and aerosol discharge presents a situation with similar characteristics, but the SDP fusion profiles of the three states differ significantly.
The SDP graph with the fused features improves the distinguishability of the partial discharge signal features to a greater extent, and the difference between signals can be highlighted by utilizing the graph to represent the features of the signals. However, the manual recognition of image features requires considerable expertise, and it is difficult to recognize small differences between images, and the demand for intelligent recognition cannot be satisfied. Therefore, the CNN-based partial discharge signal SDP information fusion image identification method is provided, and an intelligent partial discharge state identification model is established.
Fig. 11 is a schematic structural diagram of a partial discharge state identification apparatus according to an embodiment of the present invention, which is suitable for executing a partial discharge state identification method according to an embodiment of the present invention. As shown in fig. 11, the apparatus may specifically include: a signal to be recognized acquisition module 1101, a feature extraction module 1102, an image fusion module 1103, and a discharge state recognition module 1104.
The to-be-identified signal acquiring module 1101 is configured to acquire an original transformer signal as a to-be-identified signal; the feature extraction module 1102 is configured to separate an oscillation component and an impact component of the signal to be identified through resonance sparse decomposition, so as to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified; an image fusion module 1103, configured to perform information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by using an SDP analysis method, so as to obtain an SDP fusion image of the signal to be identified; and the discharge state identification module 1104 is configured to input the SDP fusion image of the signal to be identified to a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be identified.
In the embodiment of the application, the high resonance component and the low resonance component of the signal can reflect different fault states of the transformer, the SDP fusion can extract the characteristics of different signals in an image mode, so that the difference between different running states of the transformer is reflected, and finally, a CNN network is used for identifying the SDP fusion image. Compared with other measuring methods of partial discharge signals, the technical scheme of the embodiment of the application is that the time sequence series is measured for the same point, then an effective feature extraction method is selected, effective features in original signals are extracted, and the loss of the original signals is avoided.
Optionally, the system further comprises a model training module, configured to:
obtaining a sample signal by using a discharge simulator;
randomly dividing the sample signal into a training sample signal and a test sample signal;
performing feature extraction on the test sample signal to obtain a high resonance component of the test sample signal and a low resonance component of the test sample signal;
fusing the high resonance component of the test sample signal and the low resonance component of the test sample signal by using an SDP analysis method to obtain an SDP fusion image corresponding to the test sample signal;
training a pre-constructed convolutional neural network model by using an SDP fusion image corresponding to the test sample signal and the partial discharge state of the test sample signal;
and optimizing each layer of parameters of the pre-constructed convolutional neural network model until the recognition accuracy is greater than a set accuracy threshold value, so as to obtain the trained convolutional neural network model.
Optionally, the system further comprises a model updating module, configured to, after the trained convolutional neural network model is obtained, input the SDP fusion image corresponding to the test sample signal to the trained convolutional neural network model for identification, and obtain a partial discharge state;
and comparing the actual partial discharge state and the partial discharge state corresponding to the test sample signal, and updating the trained convolutional neural network model according to the comparison result.
Optionally, the image fusion module 1103 is specifically configured to:
applying an SDP method, mapping the time sequence of the high resonance component of the signal to be identified and the time sequence of the low resonance component of the signal to be identified to a polar coordinate graph, and generating a scatter diagram on the polar coordinate;
and realizing information fusion according to the scatter diagram to obtain an SDP fusion image of the signal to be identified.
Optionally, the partial discharge state includes: corona discharge, air gap discharge, and aerosol discharge.
Optionally, the convolutional neural network includes an input layer, a convolutional layer, a sampling layer, a full-link layer, and an output layer.
An embodiment of the present application further provides a discharge simulator, and fig. 12 shows a schematic structural diagram of the discharge simulator. Referring to fig. 12, the discharge simulator includes an ac power supply, a step-up transformer, a protection resistor, a coupling capacitor, a high-voltage bushing, a small bushing, a transformer partial discharge model, and a current sensor and a waveform discharge analyzer; the step-up transformer comprises a no-corona experimental transformer and an auto-coupling voltage regulator; the current sensor is used for detecting a current signal and transmitting the current signal to the waveform discharge analyzer; the waveform discharge analyzer samples the current signal to obtain a first sample signal.
The working process of the discharge simulator comprises the following steps: adjusting the discharge simulator to a normal operation state; measuring the voltage at the beginning of breakdown, the voltage at the beginning of discharge and background noise in the operation process; raising the experimental voltage, and taking the voltage when the waveform discharge analyzer displays electric pulse as the discharge voltage of the initial model; when the partial discharge condition meets a stable condition, recording the discharge voltage as a second sample signal; the first sample signal and the second sample signal constitute a sample signal.
Specifically, the discharge simulator comprises an Alternating Current (AC) a, a step-up transformer b, a protection resistor c, a coupling capacitor d, a high-voltage bushing e, a small bushing f, a transformer partial discharge model g, a Current sensor h and a waveform discharge analyzer i, wherein the waveform discharge analyzer i is arranged on a console. In practical applications, the coupling capacitor d is a high-voltage coupling capacitor with a capacitance of 500pF and a withstand voltage of 100KV, and is used for providing a partially discharged pulse current for the circuit in an experimental process. The step-up transformer b is formed by combining a non-corona experimental transformer and an autotransformer. In the experimental process, oil is required to be filled in an oil tank to simulate transformer oil, and a partial discharge model g of the transformer is put in the oil tank to simulate the working environment of the transformer. When the discharge starts, pulse current appears on a grounded line, the current signal can be detected by using a current sensor h for inhibiting the detection frequency band from 500KHz to 16MHz, the current signal is transmitted to a waveform discharge analyzer i through a cable for display, and the waveform discharge analyzer i samples the current signal to obtain a first sample signal.
In addition, the working process of the discharge simulator is specifically as follows: the test equipment was first connected and assembled as per fig. 10 and the discharge simulator was adjusted to a normal operating condition. Measuring the voltage at the beginning of breakdown, the voltage at the beginning of discharge and background noise in the operation process; after the model is connected, the voltage of the experiment is slowly increased, and the voltage is recorded as the background noise when the voltage is added and the experiment voltage is lower. Continuing to increase the voltage until the waveform discharge analyzer i can see the discharge pulse, and recording the voltage as the initial model discharge voltage; the voltage continues to be slowly raised and when the partial discharge is gradually smoothed, i.e. when a stable condition is met, the recording of the discharge voltage with the detection means as a second sample signal is started. The first sample signal and the second sample signal constitute a sample signal as data required for subsequent analysis. Through the discharge simulator, a plurality of groups of partial discharge signals are obtained, and complex signals which are strong in modulation signal and multiple and close in frequency component are superposed, so that training sample signals are enriched, the accuracy of model training is further improved, and the accuracy of model application is further improved.
Fig. 13 shows a practical application diagram of a discharge simulator, and referring to fig. 13, HV electrode is a high voltage electrode, oil is transformer oil, void is void, and Metal pieces are Metal pieces.
An embodiment of the present invention further provides a partial discharge identification device, please refer to fig. 14, fig. 14 is a schematic structural diagram of a partial discharge identification device, as shown in fig. 12, the partial discharge identification device includes: a processor 141, and a memory 142 connected to the processor 141; the memory 142 is used for storing a computer program for at least performing a partial discharge state identification method in the embodiment of the present invention; processor 141 is configured to invoke and execute the computer program in the memory; the partial discharge state identification method at least comprises the following steps: acquiring an original transformer signal as a signal to be identified; separating an oscillation component and an impact component of a signal to be identified through resonance sparse decomposition to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified; carrying out information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method to obtain an SDP fusion image of the signal to be identified; and inputting the SDP fusion image of the signal to be recognized into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized.
The embodiment of the present invention further provides a storage medium, where the storage medium stores a computer program, and when the computer program is executed by a processor, the method implements the steps in the method for identifying a partial discharge state according to the embodiment of the present invention: acquiring an original transformer signal as a signal to be identified; separating an oscillation component and an impact component of a signal to be identified through resonance sparse decomposition to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified; carrying out information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method to obtain an SDP fusion image of the signal to be identified; and inputting the SDP fusion image of the signal to be recognized into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized.
It is understood that the same or similar parts in the above embodiments may be mutually referred to, and the same or similar parts in other embodiments may be referred to for the content which is not described in detail in some embodiments.
It should be noted that the terms "first," "second," and the like in the description of the present invention are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Further, in the description of the present invention, the meaning of "a plurality" means at least two unless otherwise specified.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (10)
1. A partial discharge state identification method is characterized in that:
acquiring an original transformer signal as a signal to be identified;
separating an oscillation component and an impact component of the signal to be identified through resonance sparse decomposition to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified;
carrying out information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method so as to obtain an SDP fusion image of the signal to be identified;
and inputting the SDP fusion image of the signal to be recognized to a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be recognized.
2. The method of claim 1, wherein the training process of the convolutional neural network model comprises:
obtaining a sample signal by using a discharge simulator;
randomly dividing the sample signal into a training sample signal and a test sample signal;
performing feature extraction on the training sample signal to obtain a high resonance component of the training sample signal and a low resonance component of the training sample signal;
fusing the high resonance component of the training sample signal and the low resonance component of the training sample signal by using the SDP analysis method to obtain an SDP fusion image corresponding to the training sample signal;
training a pre-constructed convolutional neural network model by applying an SDP fusion image corresponding to the training sample signal and the partial discharge state of the training sample signal;
and optimizing each layer of parameters of the pre-constructed convolutional neural network model until the recognition accuracy is greater than a set accuracy threshold value, so as to obtain the trained convolutional neural network model.
3. The method of claim 2, after obtaining the trained convolutional neural network model, further comprising:
inputting the SDP fusion image corresponding to the test sample signal to the trained convolutional neural network model for identification to obtain a partial discharge state;
and comparing the actual partial discharge state corresponding to the test sample signal with the partial discharge state, and updating the trained convolutional neural network model according to the comparison result.
4. The method of claim 1, wherein applying SDP analysis to perform information fusion on the high-resonance component of the signal to be identified and the low-resonance component of the signal to be identified to obtain an SDP fusion image of the signal to be identified comprises:
applying an SDP method, mapping the time sequence of the high resonance component of the signal to be identified and the time sequence of the low resonance component of the signal to be identified to a polar coordinate graph, and generating a scatter diagram on the polar coordinate;
and realizing information fusion according to the scatter diagram to obtain an SDP fusion image of the signal to be identified.
5. The method of claim 1, wherein the partial discharge state comprises: corona discharge, air gap discharge, and aerosol discharge.
6. The method of any one of claims 1-5, wherein the convolutional neural network comprises an input layer, a convolutional layer, a sampling layer, a fully-connected layer, and an output layer.
7. A partial discharge state recognition apparatus characterized in that:
the to-be-identified signal acquisition module is used for acquiring an original transformer signal as a to-be-identified signal;
the characteristic extraction module is used for separating an oscillation component and an impact component of the signal to be identified through resonance sparse decomposition so as to extract a high resonance component of the signal to be identified and a low resonance component of the signal to be identified;
the image fusion module is used for performing information fusion on the high resonance component of the signal to be identified and the low resonance component of the signal to be identified by applying an SDP analysis method so as to obtain an SDP fusion image of the signal to be identified;
and the discharge state identification module is used for inputting the SDP fusion image of the signal to be identified into a pre-trained convolutional neural network model to obtain a partial discharge state corresponding to the signal to be identified.
8. A discharge simulator is characterized by comprising an alternating current power supply, a boosting transformer, a protective resistor, a coupling capacitor, a high-voltage bushing, a small bushing, a transformer partial discharge model, a current sensor and a waveform discharge analyzer;
the booster transformer comprises a no-corona experimental transformer and an auto-coupling voltage regulator;
the current sensor is used for detecting a current signal and transmitting the current signal to the waveform discharge analyzer;
the waveform discharge analyzer samples the current signal to obtain a first sample signal;
the working process of the discharge simulator comprises the following steps:
adjusting the discharge simulator to a normal operation state;
measuring the voltage at the beginning of breakdown, the voltage at the beginning of discharge and background noise in the operation process;
raising the experimental voltage, and taking the voltage when the waveform discharge analyzer displays electric pulse as the initial model discharge voltage;
when the partial discharge condition meets a stable condition, recording the discharge voltage as a second sample signal;
the first sample signal and the second sample signal constitute the sample signal.
9. A partial discharge recognition apparatus, characterized by comprising:
a processor and a memory coupled to the processor, the processor for receiving a sample signal from a discharge simulator;
the memory is configured to store a computer program for performing at least the partial discharge state identification method of any one of claims 1 to 6;
the processor is used for calling and executing the computer program in the memory.
10. A storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the steps of the partial discharge state identification method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010902297.6A CN112036320A (en) | 2020-09-07 | 2020-09-07 | Partial discharge state identification method and device, discharge simulator and identification equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010902297.6A CN112036320A (en) | 2020-09-07 | 2020-09-07 | Partial discharge state identification method and device, discharge simulator and identification equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112036320A true CN112036320A (en) | 2020-12-04 |
Family
ID=73587089
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010902297.6A Pending CN112036320A (en) | 2020-09-07 | 2020-09-07 | Partial discharge state identification method and device, discharge simulator and identification equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112036320A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113191313A (en) * | 2021-05-20 | 2021-07-30 | 国能大渡河沙坪发电有限公司 | Video stream discharge identification method and device based on hydraulic power plant and computer equipment |
| CN113238128A (en) * | 2021-05-11 | 2021-08-10 | 广州智丰电气科技有限公司 | Circulating progressive partial discharge discrimination method |
| CN116647946A (en) * | 2023-07-27 | 2023-08-25 | 济宁九德半导体科技有限公司 | Semiconductor-based heating control system and method thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103091345A (en) * | 2013-01-15 | 2013-05-08 | 湖南省电力公司检修公司 | Nuclear-magnetic-resonance-technology-based method for detecting transformer oil ageing state parameters |
| CN108573225A (en) * | 2018-03-30 | 2018-09-25 | 国网天津市电力公司电力科学研究院 | A kind of local discharge signal mode identification method and system |
| CN109102012A (en) * | 2018-07-30 | 2018-12-28 | 上海交通大学 | A kind of defect identification method and system of local discharge signal |
| 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 |
| CN111025100A (en) * | 2019-12-10 | 2020-04-17 | 中国电力科学研究院有限公司 | Transformer ultrahigh frequency partial discharge signal mode identification method and device |
-
2020
- 2020-09-07 CN CN202010902297.6A patent/CN112036320A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103091345A (en) * | 2013-01-15 | 2013-05-08 | 湖南省电力公司检修公司 | Nuclear-magnetic-resonance-technology-based method for detecting transformer oil ageing state parameters |
| CN108573225A (en) * | 2018-03-30 | 2018-09-25 | 国网天津市电力公司电力科学研究院 | A kind of local discharge signal mode identification method and system |
| CN109102012A (en) * | 2018-07-30 | 2018-12-28 | 上海交通大学 | A kind of defect identification method and system of local discharge signal |
| 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 |
| CN111025100A (en) * | 2019-12-10 | 2020-04-17 | 中国电力科学研究院有限公司 | Transformer ultrahigh frequency partial discharge signal mode identification method and device |
Non-Patent Citations (1)
| Title |
|---|
| 周沛: "基于深度学习的旋转机械振动故障诊断方法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113238128A (en) * | 2021-05-11 | 2021-08-10 | 广州智丰电气科技有限公司 | Circulating progressive partial discharge discrimination method |
| CN113191313A (en) * | 2021-05-20 | 2021-07-30 | 国能大渡河沙坪发电有限公司 | Video stream discharge identification method and device based on hydraulic power plant and computer equipment |
| CN116647946A (en) * | 2023-07-27 | 2023-08-25 | 济宁九德半导体科技有限公司 | Semiconductor-based heating control system and method thereof |
| CN116647946B (en) * | 2023-07-27 | 2023-10-13 | 济宁九德半导体科技有限公司 | Semiconductor-based heating control system and method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112036320A (en) | Partial discharge state identification method and device, discharge simulator and identification equipment | |
| Duan et al. | Identification of partial discharge defects based on deep learning method | |
| CN105548832B (en) | High voltage power cable fault recognition method | |
| Sahoo et al. | Trends in partial discharge pattern classification: a survey | |
| CN109633368B (en) | VMD and DFA-based power quality disturbance detection method for power distribution network with distributed power supplies | |
| Ferreira et al. | A method based on independent component analysis for single and multiple power quality disturbance classification | |
| CN103076547B (en) | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines | |
| EP3699614A1 (en) | Method and system of partial discharge recognition for diagnosing electrical networks | |
| Rosle et al. | Partial discharges classification methods in xlpe cable: A review | |
| Do et al. | Convolutional-neural-network-based partial discharge diagnosis for power transformer using UHF sensor | |
| CN110007197B (en) | Series arc fault detection system and detection method | |
| CN107907799A (en) | The recognition methods of shelf depreciation defect type based on convolutional neural networks and system | |
| CN111679166A (en) | Switch cabinet partial discharge fault multi-source information fusion detection early warning system and method based on wireless transmission technology | |
| CN109102508B (en) | Method for identifying insulation defects of alternating current cable based on partial discharge image characteristics | |
| CN107179486A (en) | A kind of GIS device monitors ultrahigh-frequency signal noise-reduction method on-line | |
| CN112014700A (en) | GIS insulator defect identification method and system based on partial discharge multi-information fusion | |
| Hunter et al. | Comparison of two partial discharge classification methods | |
| CN112763871A (en) | Partial discharge classification identification method | |
| Su et al. | Classification of partial discharge events in GILBS using probabilistic neural networks and the fuzzy c-means clustering approach | |
| Al-geelani et al. | A review on hybrid wavelet regrouping particle swarm optimization neural networks for characterization of partial discharge acoustic signals | |
| CN105785236A (en) | GIS local discharge detection external interference signal elimination method | |
| Mitiche et al. | Classification of EMI discharge sources using time–frequency features and multi-class support vector machine | |
| CN114217164B (en) | Cable fault distance measurement method and system based on discharge waveform intelligent identification | |
| Butdee et al. | Pattern recognition of partial discharge faults using convolutional neural network (CNN) | |
| Chang et al. | Assessment of the insulation status aging in power cable joints using support vector machine |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201204 |