CN116736051A - GIS insulation defect intelligent diagnosis method, device, equipment and storage medium - Google Patents
GIS insulation defect intelligent diagnosis method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN116736051A CN116736051A CN202310573291.2A CN202310573291A CN116736051A CN 116736051 A CN116736051 A CN 116736051A CN 202310573291 A CN202310573291 A CN 202310573291A CN 116736051 A CN116736051 A CN 116736051A
- Authority
- CN
- China
- Prior art keywords
- model
- defect
- frequency
- wave signal
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 188
- 238000009413 insulation Methods 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000003745 diagnosis Methods 0.000 title claims abstract description 21
- 238000013135 deep learning Methods 0.000 claims abstract description 44
- 238000001514 detection method Methods 0.000 claims abstract description 37
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 17
- 239000012212 insulator Substances 0.000 claims description 15
- 239000002923 metal particle Substances 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 14
- 238000012549 training Methods 0.000 claims description 13
- 238000013136 deep learning model Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 10
- 239000000725 suspension Substances 0.000 claims description 10
- 230000004927 fusion Effects 0.000 claims description 9
- 238000012795 verification Methods 0.000 claims description 8
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 2
- 238000002405 diagnostic procedure Methods 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 9
- 230000035945 sensitivity Effects 0.000 abstract description 6
- 238000001914 filtration Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 3
- 239000002184 metal Substances 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000005684 electric field Effects 0.000 description 2
- 238000007667 floating Methods 0.000 description 2
- 239000011810 insulating material Substances 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000003990 capacitor Substances 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000007257 malfunction Effects 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
- 239000002245 particle Substances 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000001902 propagating effect Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 239000011800 void material Substances 0.000 description 1
Classifications
-
- 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
- G01R31/1254—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 of gas-insulated power appliances or vacuum gaps
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to the field of GIS equipment, and discloses a GIS insulation defect intelligent diagnosis method, device, equipment and storage medium. The method comprises the following steps: constructing a partial discharge model; constructing a partial discharge detection system; constructing a GIS insulation defect deep learning identification model, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal according to the constructed partial discharge model and a partial discharge detection system, and constructing a GIS insulation defect deep learning identification model according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal; defect identification and prediction: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, and carrying out real-time identification and prediction of the insulation defects. The invention combines the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal, the scheme fully utilizes the advantages of various detection methods, and improves the overall sensitivity and the accuracy of the detection system.
Description
Technical Field
The invention relates to the field of warehouse logistics, in particular to a GIS insulation defect intelligent diagnosis method, device, equipment and storage medium.
Background
GIS has advantages such as small, maintenance is few, anti-pollution flashover ability reinforce, wide application in power transmission and distribution system. However, the GIS device exhibits a partial discharge phenomenon during operation, which results in a decrease in insulation performance of the device and even a malfunction. Therefore, the method has important significance in partial discharge detection and diagnosis of GIS equipment.
At present, the GIS partial discharge online detection technology still needs to be optimized and mature, and on-site data acquisition is relatively difficult, so that sample data which can be used for GIS partial discharge type identification is relatively lacking, but a large amount of sample data is required for training a mode identification method such as a neural network and an expert system to accurately predict and identify, and although the SVM algorithm has very good performance in solving the problems of small samples and nonlinearity, the identification and prediction performance is greatly influenced by parameters such as penalty factors and kernel functions. In addition, the partial discharge signal data tend to have higher and complex dimensions, and the accuracy and efficiency of identification can be seriously affected by only identifying the type of the partial discharge signal data by using an intelligent diagnosis algorithm.
Disclosure of Invention
The invention mainly aims to solve the problem that in the prior art, the express to be modified cannot be reassigned or dispatch planning can not be adjusted in real time, so that the residence time of the express is too long.
The first aspect of the invention provides a GIS insulation defect intelligent diagnosis method, which comprises the following steps:
constructing a partial discharge model, wherein the partial discharge model comprises a spike defect model in GIS equipment, an insulating surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model;
constructing a partial discharge detection system, wherein the partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal generated by partial discharge of GIS equipment;
constructing a GIS insulation defect deep learning identification model, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal according to the constructed partial discharge model and a partial discharge detection system, and constructing a GIS insulation defect deep learning identification model according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
defect identification and prediction: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
The construction of the GIS insulation defect deep learning identification model comprises the following steps:
data preprocessing: according to the constructed partial discharge model and the partial discharge detection system, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal, preprocessing synchronously acquired data, and acquiring a preprocessed signal;
feature extraction: extracting characteristic data from the preprocessed signals, wherein the characteristic data comprises extracted peak value parameters, waveform factor parameters, frequency parameters, amplitude parameters, arrival time parameters and energy parameters, and the characteristic data is used as input of a deep learning recognition model;
constructing a deep learning identification model: selecting a deep learning identification model, building a model structure according to the selected model, and determining model parameters;
training a model: inputting the collected characteristic data and the corresponding defect type labels into a deep learning model, performing model training, monitoring a loss function and accuracy in the training process, and adjusting super parameters to optimize the performance of the model;
model verification and tuning: evaluating the performance of the model by using a verification set, wherein the performance comprises an accuracy rate, a recall rate and an F1 score, and optimizing the model according to an evaluation result, wherein the optimizing comprises modifying a model structure and adjusting a learning rate;
Model test: evaluating the generalization performance of the model on the test set, and ensuring that the model performs well on unseen data; finally, a GIS insulation defect deep learning identification model is obtained.
The feature extraction method comprises the following steps:
extracting peak value parameters and waveform factor parameters for the pulse current signals; extracting frequency parameters and amplitude parameters of the signals for the ultrahigh frequency signals; ultrasonic signals, and extracting arrival time parameters and energy parameters of the signals.
The feature extraction step includes the steps of:
pulse current signal feature extraction:
a. peak parameters: finding the peak amplitude of each partial discharge event from the preprocessed pulse current signal;
b. waveform factor parameters: calculating a crest factor of the pulse current signal;
extracting characteristics of the ultrahigh frequency signals:
a. frequency parameters: performing Fourier transformation on the preprocessed ultrahigh frequency signal, and analyzing the frequency spectrum characteristics of the ultrahigh frequency signal to obtain frequency components;
b. amplitude parameters: extracting the standard deviation of the amplitude of the ultrahigh frequency signal;
ultrasonic signal feature extraction:
a. time of arrival parameters: calculating the arrival time of each partial discharge event in the ultrasonic signal, namely the time from the start of the signal to the occurrence of the discharge event;
b. Energy parameters: calculating the energy density of each partial discharge event in the ultrasonic signal;
feature data integration: and integrating the extracted peak value parameter, the waveform factor parameter, the frequency parameter, the amplitude parameter, the arrival time parameter and the energy parameter into a feature vector, wherein the feature vector is used as the input of the deep learning recognition model.
The defect identification and prediction includes:
defect type identification: deploying a GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and outputting a predicted defect type;
defect position location: positioning the defect position according to the signal arrival time difference of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
real-time monitoring and alarming: if abnormal partial discharge is found, alarm information is sent out.
The defect position locating step includes the steps of:
calculating an arrival time difference: analyzing the preprocessed signals, and respectively calculating the arrival time differences of the high-frequency pulse current signals, the high-frequency electromagnetic wave signals and the ultrasonic signals received by each sensor, wherein the arrival time differences are the time differences of the signals transmitted to each sensor from the defect occurrence position;
Positioning algorithm: calculating the defect position by using a positioning algorithm according to the arrival time difference and the position information of the sensor to obtain a positioning result;
fusing positioning results: fusing the positioning results of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal to obtain a fused positioning result;
visual display: and visually displaying the fused positioning result on a three-dimensional model of the GIS equipment.
The step of fusing the positioning results comprises the following steps:
initializing a Kalman filter: setting an initial state of a Kalman filter, wherein the initial state comprises an initial position, a speed and a covariance matrix, and the initial position can be calculated according to the independent positioning results of all the sensors;
prediction stage: in the prediction stage, the Kalman filter predicts the current state by using the information of the previous state, predicts the next state of the defect position according to a system dynamic model, and simultaneously updates a state covariance matrix to reflect the uncertainty of the prediction stage;
updating: the Kalman filter corrects the state obtained in the prediction stage by combining the observed data, calculates Kalman gain, wherein the Kalman gain represents the weight of the observed data on the predicted state, corrects the predicted state according to the Kalman gain and the observed data to obtain more accurate current state estimation, and finally updates a state covariance matrix to reflect the uncertainty of the update stage;
The iterative process: in the whole process of fusion positioning, the prediction and updating stages are continuously repeated to fuse the positioning information from the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal in real time, so that the overall positioning accuracy is improved;
outputting a result: and outputting the fused defect position according to the fusion state estimation obtained by the Kalman filter.
The second aspect of the present invention provides an intelligent diagnosis device for insulation defects of a GIS, comprising:
the partial discharge model comprises a GIS equipment internal spike defect model, an insulation surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model;
the partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal which are generated by partial discharge of the GIS equipment;
the GIS insulation defect deep learning identification model is constructed according to the constructed partial discharge model and the partial discharge detection system, a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal are acquired synchronously, and the GIS insulation defect deep learning identification model is constructed according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
Defect recognition and prediction unit: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
A third aspect of the present invention provides an electronic device, comprising: a memory and at least one processor, the memory having instructions stored therein, the memory and the at least one processor being interconnected by a line; the at least one processor invokes the instructions in the memory to cause the electronic device to perform the above-described intelligent diagnosis method of GIS insulation defects.
A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored therein, which when run on a computer, cause the computer to perform the above-described GIS insulation defect intelligent diagnosis method as described above.
In the technical scheme, a partial discharge model is firstly constructed, and comprises a spike defect model in GIS equipment, an insulating surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model. Then, a partial discharge detection system is designed to synchronously acquire a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal generated by partial discharge of GIS equipment. Next, these signals are preprocessed and feature extracted, and defect type identification is performed using a deep learning model. And finally, positioning the defect position according to the signal arrival time difference, and fusing a multi-signal positioning result by adopting a Kalman filtering algorithm to improve the positioning precision.
Drawings
FIG. 1 is a schematic diagram of a GIS partial discharge simulation device;
fig. 2 is a photograph of data signals collected by detecting different partial discharge models by an ultra-high frequency method.
Fig. 3 is a photograph of data signals acquired by detecting different partial discharge models using an ultrasonic method.
Detailed Description
The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
The operation statistics of the GIS for many years shows that the insulation defects in the GIS mainly comprise free metal particles, high-voltage conductor spines, shell spines, surface dirt and internal defects of insulators, suspended electrodes and the like.
The GIS is filled with SF6 gas with high air pressure, the pulse signal generated by partial discharge has short duration, the rapid rising steep wave pulse comprises an electromagnetic wave band signal with the frequency reaching GHZ, the electromagnetic wave propagates in the cavity, and the electromagnetic wave can leak out at the place where the wave impedance is discontinuous, such as at the edge of a basin-type insulator.
The GIS partial discharge also diffuses the ionized gas channel, resulting in a local gas pressure surge, thereby generating ultrasonic waves that propagate along the GIS interior gas and cavity.
When the GIS generates partial discharge, a partial ionization region appears in the GIS, light radiation also appears along with the recombination process of charged particles, and SF6 gas also can be decomposed under the ion energy discharge.
These all form the basis for detecting partial discharges.
In the scheme of the invention, a partial discharge model is firstly constructed, and comprises a spike defect model in GIS equipment, an insulating surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model. Then, a partial discharge detection system is designed to synchronously acquire a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal generated by partial discharge of GIS equipment. Next, these signals are preprocessed and feature extracted, and defect type identification is performed using a deep learning model. And finally, positioning the defect position according to the signal arrival time difference, and fusing a multi-signal positioning result by adopting a Kalman filtering algorithm to improve the positioning precision.
Simulation test:
the GIS partial discharge simulation device body is built and mainly comprises a boosting air chamber, a coupling capacitor air chamber, a calibration unit, an auxiliary bracket and other modules.
As shown in fig. 1, four GIS partial discharge models are constructed: the method comprises the following steps of (1) spike defects in GIS equipment, spike defects on the insulating surface, suspension potential defects and metal particle defects on the surface of an insulator.
Building a GIS partial discharge detection system: the pulse current method, the ultrahigh frequency method and the ultrasonic method are adopted to synchronously detect the expression forms of different partial discharge models in different modes, and data signals are collected, as shown in figures 2 and 3.
For easy understanding, the following describes a specific flow of an embodiment of the present invention, and an embodiment of a method for intelligently diagnosing a GIS insulation defect in the embodiment of the present invention includes:
and constructing a partial discharge model, wherein the partial discharge model comprises a spike defect model in GIS equipment, an insulating surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model.
The partial discharge model is mainly aimed at typical defect types in GIS equipment, the following four defect models are the most common and important defect types in the GIS equipment, and have great influence on equipment performance and operation safety:
GIS equipment internal spike defect model: spike defects are sharp protrusions generated on the surfaces or contact points of metal parts in GIS equipment, and the defects cause uneven electric field distribution, so that partial discharge is caused. The research of the defect model is helpful for analyzing and diagnosing the partial discharge phenomenon of the metal parts in the GIS equipment.
Insulation surface spike defect model: spike defects on the insulating surface can also cause uneven electric field distribution, which in turn can lead to partial discharge. Studying such defect models helps to evaluate the performance and lifetime of the insulating material, as well as the effect of partial discharge on the insulating material.
Suspension potential defect model: the floating potential defect refers to that a certain potential difference exists between metal parts and other parts in GIS equipment, and the defect causes phenomena of partial discharge, corona discharge and the like. The defect model is researched to be helpful for analyzing and diagnosing the problem of the floating potential in the GIS equipment, so that the operation safety of the equipment is improved.
Insulator surface metal particle defect model: the defect of the metal particles on the surface of the insulator refers to that some metal particles are attached to the surface of the insulator of the GIS equipment, and the defect causes partial discharge. The defect model is studied to help analyze the influence of metal particles on the insulation performance of GIS equipment and the damage of partial discharge to insulators.
The partial discharge model is limited to the four directions, because the defect types have high practical application value and research significance. Intensive studies on these typical defect types are helpful to improve the operation reliability and safety of GIS equipment and reduce the probability of faults and accidents.
And constructing a partial discharge detection system, wherein the partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal generated by partial discharge of the GIS equipment.
And constructing a GIS insulation defect deep learning identification model, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal according to the constructed partial discharge model and a partial discharge detection system, and constructing the GIS insulation defect deep learning identification model according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic signal.
The reason for synchronously collecting the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal mainly comprises the following points:
the detection accuracy and reliability are improved: the sensitivity and response characteristics of different detection methods to partial discharge are different, and the synchronous acquisition of the signals can comprehensively consider the advantages of various detection methods and improve the detection accuracy and reliability. For example, the ultra-high frequency method has strong anti-interference capability in an electromagnetic environment, and the ultrasonic method has high precision in space positioning.
Realizing multi-signal fusion: the fusion processing of the signals can be realized by synchronously collecting various signals, and the identification and positioning precision is improved by utilizing the association information among various signals. For example, by using an algorithm such as kalman filtering, the defect position can be estimated more accurately by combining the positioning information of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal.
Positioning accuracy is improved: after the signals are synchronously collected, the defect position can be positioned according to the signal arrival time difference. The speed and attenuation characteristics of different signals in the propagation process are different, and synchronous acquisition is helpful for more accurately calculating the signal arrival time difference, so that the positioning accuracy is improved.
Auxiliary identification: for some defects, a single detection method cannot be accurately identified. The simultaneous acquisition of multiple signals can provide more information that helps to distinguish and identify different types of defects. For example, in some cases, the high frequency pulsed current signal and the high frequency electromagnetic wave signal exhibit similar characteristics, while the ultrasonic wave signal may provide supplemental information to assist in identifying the defect type.
Defect identification and prediction: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
The construction of the GIS insulation defect deep learning identification model comprises the following steps:
data preprocessing: according to the constructed partial discharge model and the partial discharge detection system, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal, preprocessing synchronously acquired data, and acquiring a preprocessed signal;
feature extraction: extracting characteristic data from the preprocessed signals, wherein the characteristic data comprises extracted peak value parameters, waveform factor parameters, frequency parameters, amplitude parameters, arrival time parameters and energy parameters, and the characteristic data is used as input of a deep learning recognition model;
constructing a deep learning identification model: selecting a deep learning identification model, building a model structure according to the selected model, and determining model parameters;
training a model: inputting the collected characteristic data and the corresponding defect type labels into a deep learning model, performing model training, monitoring a loss function and accuracy in the training process, and adjusting super parameters to optimize the performance of the model;
model verification and tuning: evaluating the performance of the model by using a verification set, wherein the performance comprises an accuracy rate, a recall rate and an F1 score, and optimizing the model according to an evaluation result, wherein the optimizing comprises modifying a model structure and adjusting a learning rate;
Model test: evaluating the generalization performance of the model on the test set, and ensuring that the model performs well on unseen data; finally, a GIS insulation defect deep learning identification model is obtained.
The feature extraction method comprises the following steps:
extracting peak value parameters and waveform factor parameters for the pulse current signals; extracting frequency parameters and amplitude parameters of the signals for the ultrahigh frequency signals; ultrasonic signals, and extracting arrival time parameters and energy parameters of the signals. The feature extraction step includes the steps of:
pulse current signal feature extraction:
a. peak parameters: the peak value parameter can reflect the maximum intensity of the pulse current signal, and different types of partial discharge defects generate pulse current signals with different amplitudes, so that the peak value parameter is an important characteristic parameter.
b. Waveform factor parameters: the form factor parameter (e.g., crest factor) may describe the waveform characteristics of the pulsed current signal, with different types of partial discharge defects producing pulsed current signals of different waveforms, so the form factor parameter helps to distinguish between different types of defects.
Extracting characteristics of the ultrahigh frequency signals:
a. frequency parameters: the ultrahigh frequency signal generated by partial discharge has certain frequency spectrum characteristics, and different types of partial discharge defects generate stronger signals in different frequency ranges, so that the frequency parameter is a key characteristic parameter.
b. Amplitude parameters: the amplitude parameter (e.g., standard deviation of the amplitude) reflects the intensity variation of the ultra-high frequency signal, and different types of partial discharge defects produce ultra-high frequency signals of different intensities, so the amplitude parameter helps to distinguish between different types of defects.
Ultrasonic signal feature extraction:
a. time of arrival parameters: the ultrasonic signal is affected by the propagation path and distance during the propagation process, and the arrival time of the ultrasonic signal is different due to different types of partial discharge defects, so that the arrival time parameter is an important characteristic parameter.
b. Energy parameters: the energy parameter (e.g., energy density) reflects the energy distribution characteristics of the ultrasonic signal, and different types of partial discharge defects produce ultrasonic signals of different energies, so the energy parameter helps to distinguish between different types of defects.
The parameters can provide rich characteristic information for the deep learning recognition model, and are helpful for improving the recognition accuracy and robustness of the model.
Feature data integration: and integrating the extracted peak value parameter, the waveform factor parameter, the frequency parameter, the amplitude parameter, the arrival time parameter and the energy parameter into a feature vector, wherein the feature vector is used as the input of the deep learning recognition model.
The partial discharge pulse phase distribution (Phase Resolved Partial Discharge, PRPD) mode is the currently dominant feature representation method, also called phi-q-n mode, and can visually represent the relationship among the power frequency phase phi, the discharge quantity q and the discharge frequency n corresponding to the partial discharge pulse by an image. A large number of feature extraction methods are used for partial discharge pulse phase distribution spectrograms, and a statistical operator is used as a feature attribute of an image to perform pattern recognition, wherein the feature attribute represents the phase distribution feature of the partial discharge pulse; the fractal dimension and the void ratio of the PRPD spectrogram are used as characteristic parameters, and a multi-classification nonlinear support vector machine is adopted for classification; or extracting statistical parameters such as skewness, abruptness, sixth-order moment and the like of the PRPD spectrogram to form a feature matrix, and classifying the defect types by using a deep belief network. Most researches at the present stage still need to extract characteristic parameters of the PRPD spectrogram, and have the defects of high dimensionality and excessive repeated information.
In this embodiment, the peak value parameter, the waveform factor parameter, the frequency parameter, the amplitude parameter, the arrival time parameter, and the energy parameter are integrated into a feature vector, which has the following advantages:
Multidimensional feature information: the scheme comprehensively considers various characteristic parameters, including time domain, frequency domain and energy related information. Such multidimensional feature information helps to improve the ability of the model to distinguish partial discharge defects. In contrast, the PRPD mode mainly focuses on the phase information of the pulse current signal, and cannot sufficiently capture the overall characteristics of the partial discharge defect.
Fusion of multiple signal sources: the scheme integrates the characteristics of high-frequency pulse current signals, high-frequency electromagnetic wave signals and ultrasonic signals, and improves the accuracy of detection and positioning by utilizing the information of various signal sources. While PRPD mode is generally based only on pulsed current signals, it is not possible to fully utilize information from other signal sources.
Adapt to different environment and noise conditions: the multidimensional feature vector adopted by the scheme can keep better performance under different environment and noise conditions. While PRPD mode is greatly affected by external interference and noise, it cannot provide a satisfactory detection result in a specific environment.
Compatibility with deep learning models: the multiple features are integrated into the feature vector, and can be conveniently used as the input of the deep learning model. The deep learning model can automatically learn complex modes and structures in data, so that accuracy of partial discharge defect identification is improved. While the PRPD mode is relatively limited in application in the deep learning model, the advantages of the deep learning technique cannot be fully utilized.
Generalization ability: the scheme integrates various characteristic information, is beneficial to improving the generalization capability of the model, and can adapt to GIS equipment of different types and scales. While PRPD mode is limited in generalization capability between different devices.
Through experiments, partial discharge data of two groups of GIS devices are obtained, each group comprises 10000 samples, and the 10000 samples are acquired from high-frequency pulse current signals, high-frequency electromagnetic wave signals and ultrasonic signals respectively. We will use two methods: the method (the scheme) for extracting the feature vector and the PRPD mode are used for identifying the partial discharge defect.
The feature vector method comprises the following steps:
the aforementioned feature extraction steps will be employed to extract peak parameters, waveform factor parameters, frequency parameters, amplitude parameters, time of arrival parameters, and energy parameters from each signal and integrate them into feature vectors. These feature vectors are then input into a deep learning model for training and validation. The following performance indexes are obtained under the assumption that training and verification are carried out:
accuracy rate: 95% of
Sensitivity: 94%
Specificity: 96 percent of
Positioning error: 5% (distance)
PRPD mode:
the pulse current signal is used to calculate the pulse phase distribution of the partial discharge. Then, an attempt will be made to identify and locate partial discharge defects according to the PRPD pattern. Assuming that after training and verification we have obtained the following performance indicators:
Accuracy rate: 85%
Sensitivity: 82%
Specificity: 88 percent of
Positioning error: 10% (distance)
From these hypothetical experimental results, we can see that the method of extracting feature vectors shows better performance in terms of accuracy, sensitivity, specificity and positioning error than the PRPD mode. This illustrates that the solution of integrating the extracted peak parameters, waveform factor parameters, frequency parameters, amplitude parameters, time of arrival parameters, energy parameters into one eigenvector is of great advantage.
As a preferred embodiment, the defect identification and prediction includes:
defect type identification: deploying a GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and outputting a predicted defect type;
defect position location: positioning the defect position according to the signal arrival time difference of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
real-time monitoring and alarming: if abnormal partial discharge is found, alarm information is sent out.
As a preferred embodiment, the defect position locating step includes the steps of:
Calculating an arrival time difference: and analyzing the preprocessed signals, and respectively calculating the arrival time differences of the high-frequency pulse current signals, the high-frequency electromagnetic wave signals and the ultrasonic signals received by each sensor, wherein the arrival time differences are the time differences of the signals transmitted to each sensor from the defect occurrence position.
The main purpose of calculating the arrival time difference is to locate the defect location. In the GIS device, a high-frequency pulse current signal, a high-frequency electromagnetic wave signal, and an ultrasonic wave signal generated by a partial discharge defect are propagated in various directions. Multiple sensors mounted on the device will receive these signals at different times. The time difference (arrival time difference) of the signal propagating from the defect occurrence to each sensor is related to the distance between the sensors and the propagation speed of the signal.
By calculating the time differences of arrival of the signals received by the various sensors, the location of the defect can be estimated using a localization algorithm (e.g., TDOA: time Difference of Arrival) with the aid of the signal propagation speed and the time differences of arrival. The method can improve the positioning precision of the partial discharge defect in the GIS equipment, thereby helping to diagnose and eliminate potential problems and preventing the equipment from failure.
Positioning algorithm: calculating the defect position by using a positioning algorithm according to the arrival time difference and the position information of the sensor to obtain a positioning result;
fusing positioning results: fusing the positioning results of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal to obtain a fused positioning result;
visual display: and visually displaying the fused positioning result on a three-dimensional model of the GIS equipment.
The step of fusing the positioning results comprises the following steps:
initializing a Kalman filter: setting an initial state of a Kalman filter, wherein the initial state comprises an initial position, a speed and a covariance matrix, and the initial position can be calculated according to the independent positioning results of all the sensors;
prediction stage: in the prediction stage, the Kalman filter predicts the current state by using the information of the previous state, predicts the next state of the defect position according to a system dynamic model, and simultaneously updates a state covariance matrix to reflect the uncertainty of the prediction stage;
updating: the Kalman filter corrects the state obtained in the prediction stage by combining the observed data, calculates Kalman gain, wherein the Kalman gain represents the weight of the observed data on the predicted state, corrects the predicted state according to the Kalman gain and the observed data to obtain more accurate current state estimation, and finally updates a state covariance matrix to reflect the uncertainty of the update stage;
The iterative process: in the whole process of fusion positioning, the prediction and updating stages are continuously repeated to fuse the positioning information from the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal in real time, so that the overall positioning accuracy is improved;
outputting a result: and outputting the fused defect position according to the fusion state estimation obtained by the Kalman filter.
The main purpose of fusing the positioning results is to improve the positioning accuracy and reliability by utilizing the advantages of various detection methods. The propagation characteristics of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal in the GIS device are different, so that they have different sensitivity and positioning accuracy to different types of defects and environmental conditions. The use of one detection method alone may suffer from certain limitations such as signal attenuation, propagation velocity differences, interference, etc. By fusing a plurality of positioning results, the advantages of various detection methods can be fully utilized, and the respective defects are reduced, so that the overall positioning accuracy is improved.
Compared with other algorithms, the experimental process finds that the fusion positioning result by adopting the Kalman filter has the following advantages:
the calculation complexity is low: the Kalman filter adopts a recursive algorithm to perform state estimation, and the calculation of the whole historical data sequence is avoided. The kalman filter has lower computational complexity than some algorithms that require storing and processing large amounts of historical data.
The real-time performance is strong: the Kalman filter can update the state estimation immediately every time new measurement data is received, and has good real-time performance. This makes it advantageous in real-time positioning and tracking applications, which are more real-time than algorithms that need to wait for bulk data processing.
Adaptivity: the kalman filter can adaptively handle uncertainties, including system dynamics and measurement noise. By automatically adjusting the covariance of the state estimate, the Kalman filter can dynamically adjust the performance of the filter based on the confidence level of the measured data. The kalman filter has better adaptivity than some fixed parameter filtering algorithms.
Robustness: the Kalman filter has good robustness, can resist the influence of abnormal measurement data to a certain extent, and reduces the influence of the abnormal data on a positioning result. Kalman filters are more robust than some algorithms that are sensitive to outlier data.
The second aspect of the present invention provides an intelligent diagnosis device for insulation defects of a GIS, comprising:
the partial discharge model comprises a GIS equipment internal spike defect model, an insulation surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model;
The partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal which are generated by partial discharge of the GIS equipment;
the GIS insulation defect deep learning identification model is constructed according to the constructed partial discharge model and the partial discharge detection system, a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal are acquired synchronously, and the GIS insulation defect deep learning identification model is constructed according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
defect recognition and prediction unit: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
The electronic devices to which the present invention relates may vary considerably in configuration or performance and may include one or more processors (central processing units, CPU) (e.g., one or more processors) and memory, one or more storage media (e.g., one or more mass storage devices) that store applications or data. The memory and storage medium may be transitory or persistent. The program stored on the storage medium may include one or more modules, each of which may include a series of instruction operations in the electronic device. Still further, the processor may be configured to communicate with a storage medium and execute a series of instruction operations in the storage medium on an electronic device.
The electronic device may also include one or more power supplies, one or more wired or wireless network interfaces, one or more input/output interfaces, and/or one or more operating systems, such as Windows Serve, mac OS X, unix, linux, freeBSD, etc.
The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, and which may also be a volatile computer readable storage medium, having instructions stored therein, which when executed on a computer, cause the computer to perform the steps of a concentration prediction method for online monitoring of dissolved gas in transformer oil.
The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, and may also be a volatile computer readable storage medium, where instructions are stored in the computer readable storage medium, where the instructions, when executed on a computer, cause the computer to perform the steps of the GIS insulation defect intelligent diagnosis method.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process of the system or apparatus and unit described above may refer to the corresponding process in the foregoing method embodiment, which is not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The intelligent diagnosis method for the GIS insulation defects is characterized by comprising the following steps of:
constructing a partial discharge model, wherein the partial discharge model comprises a spike defect model in GIS equipment, an insulating surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model;
constructing a partial discharge detection system, wherein the partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal generated by partial discharge of GIS equipment;
constructing a GIS insulation defect deep learning identification model, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal according to the constructed partial discharge model and a partial discharge detection system, and constructing a GIS insulation defect deep learning identification model according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
Defect identification and prediction: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
2. The intelligent diagnosis method of GIS insulation defect according to claim 1, wherein the construction of the deep learning identification model of GIS insulation defect comprises the following steps:
data preprocessing: according to the constructed partial discharge model and the partial discharge detection system, synchronously acquiring a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic signal, preprocessing synchronously acquired data, and acquiring a preprocessed signal;
feature extraction: extracting characteristic data from the preprocessed signals, wherein the characteristic data comprises extracted peak value parameters, waveform factor parameters, frequency parameters, amplitude parameters, arrival time parameters and energy parameters, and the characteristic data is used as input of a deep learning recognition model;
constructing a deep learning identification model: selecting a deep learning identification model, building a model structure according to the selected model, and determining model parameters;
training a model: inputting the collected characteristic data and the corresponding defect type labels into a deep learning model, performing model training, monitoring a loss function and accuracy in the training process, and adjusting super parameters to optimize the performance of the model;
Model verification and tuning: evaluating the performance of the model by using a verification set, wherein the performance comprises an accuracy rate, a recall rate and an F1 score, and optimizing the model according to an evaluation result, wherein the optimizing comprises modifying a model structure and adjusting a learning rate;
model test: evaluating the generalization performance of the model on the test set, and ensuring that the model performs well on unseen data; finally, a GIS insulation defect deep learning identification model is obtained.
3. The intelligent diagnosis method of GIS insulation defects according to claim 2, wherein the feature extraction method comprises the following steps:
extracting peak value parameters and waveform factor parameters for the pulse current signals; extracting frequency parameters and amplitude parameters of the signals for the ultrahigh frequency signals; ultrasonic signals, and extracting arrival time parameters and energy parameters of the signals.
4. The method for intelligent diagnosis of GIS insulation defects according to claim 3, wherein the feature extraction step comprises the steps of:
pulse current signal feature extraction:
a. peak parameters: finding the peak amplitude of each partial discharge event from the preprocessed pulse current signal;
b. waveform factor parameters: calculating a crest factor of the pulse current signal;
Extracting characteristics of the ultrahigh frequency signals:
a. frequency parameters: performing Fourier transformation on the preprocessed ultrahigh frequency signal, and analyzing the frequency spectrum characteristics of the ultrahigh frequency signal to obtain frequency components;
b. amplitude parameters: extracting the standard deviation of the amplitude of the ultrahigh frequency signal;
ultrasonic signal feature extraction:
a. time of arrival parameters: calculating the arrival time of each partial discharge event in the ultrasonic signal, namely the time from the start of the signal to the occurrence of the discharge event;
b. energy parameters: calculating the energy density of each partial discharge event in the ultrasonic signal;
feature data integration: and integrating the extracted peak value parameter, the waveform factor parameter, the frequency parameter, the amplitude parameter, the arrival time parameter and the energy parameter into a feature vector, wherein the feature vector is used as the input of the deep learning recognition model.
5. The method for intelligent diagnosis of GIS insulation defects according to claim 1, wherein the defect identification and prediction comprises:
defect type identification: deploying a GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and outputting a predicted defect type;
defect position location: positioning the defect position according to the signal arrival time difference of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
Real-time monitoring and alarming: if abnormal partial discharge is found, alarm information is sent out.
6. The method for intelligent diagnosis of GIS insulation defects according to claim 5, wherein the defect location step comprises the steps of:
calculating an arrival time difference: analyzing the preprocessed signals, and respectively calculating the arrival time differences of the high-frequency pulse current signals, the high-frequency electromagnetic wave signals and the ultrasonic signals received by each sensor, wherein the arrival time differences are the time differences of the signals transmitted to each sensor from the defect occurrence position;
positioning algorithm: calculating the defect position by using a positioning algorithm according to the arrival time difference and the position information of the sensor to obtain a positioning result;
fusing positioning results: fusing the positioning results of the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal to obtain a fused positioning result;
visual display: and visually displaying the fused positioning result on a three-dimensional model of the GIS equipment.
7. The intelligent diagnosis method of GIS insulation defects according to claim 6, wherein the step of fusing the positioning results comprises the steps of:
initializing a Kalman filter: setting an initial state of a Kalman filter, wherein the initial state comprises an initial position, a speed and a covariance matrix, and the initial position can be calculated according to the independent positioning results of all the sensors;
Prediction stage: in the prediction stage, the Kalman filter predicts the current state by using the information of the previous state, predicts the next state of the defect position according to a system dynamic model, and simultaneously updates a state covariance matrix to reflect the uncertainty of the prediction stage;
updating: the Kalman filter corrects the state obtained in the prediction stage by combining the observed data, calculates Kalman gain, wherein the Kalman gain represents the weight of the observed data on the predicted state, corrects the predicted state according to the Kalman gain and the observed data to obtain more accurate current state estimation, and finally updates a state covariance matrix to reflect the uncertainty of the update stage;
the iterative process: in the whole process of fusion positioning, the prediction and updating stages are continuously repeated to fuse the positioning information from the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal in real time, so that the overall positioning accuracy is improved;
outputting a result: and outputting the fused defect position according to the fusion state estimation obtained by the Kalman filter.
8. The utility model provides a GIS insulation defect intelligent diagnosis device which characterized in that includes:
the partial discharge model comprises a GIS equipment internal spike defect model, an insulation surface spike defect model, a suspension potential defect model and an insulator surface metal particle defect model;
The partial discharge detection system synchronously acquires a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal which are generated by partial discharge of the GIS equipment;
the GIS insulation defect deep learning identification model is constructed according to the constructed partial discharge model and the partial discharge detection system, a high-frequency pulse current signal, a high-frequency electromagnetic wave signal and an ultrasonic wave signal are acquired synchronously, and the GIS insulation defect deep learning identification model is constructed according to the high-frequency pulse current signal, the high-frequency electromagnetic wave signal and the ultrasonic wave signal;
defect recognition and prediction unit: and deploying the GIS insulation defect deep learning identification model on actual GIS equipment, receiving a real-time high-frequency pulse current signal, a real-time high-frequency electromagnetic wave signal and a real-time ultrasonic wave signal, and carrying out real-time identification and prediction on the insulation defect.
9. An electronic device comprising a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the electronic device to perform the steps of the GIS insulation defect intelligent diagnostic method of any one of claims 1-7.
10. A computer readable storage medium having instructions stored thereon, which when executed by a processor, implement the steps of the GIS insulation defect intelligent diagnosis method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310573291.2A CN116736051A (en) | 2023-05-22 | 2023-05-22 | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310573291.2A CN116736051A (en) | 2023-05-22 | 2023-05-22 | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116736051A true CN116736051A (en) | 2023-09-12 |
Family
ID=87903523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310573291.2A Pending CN116736051A (en) | 2023-05-22 | 2023-05-22 | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116736051A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118130984A (en) * | 2024-05-10 | 2024-06-04 | 山东博通节能科技有限公司 | Cable partial discharge fault real-time monitoring method based on data driving |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107728014A (en) * | 2017-08-23 | 2018-02-23 | 国网山东省电力公司检修公司 | Method and system based on multiple sensor signals characteristic identificating equipment insulation defect |
| CN109164362A (en) * | 2018-09-25 | 2019-01-08 | 上海交通大学 | A kind of recognition methods and system of direct current cables shelf depreciation defect failure |
| CN109444682A (en) * | 2018-11-02 | 2019-03-08 | 国网四川省电力公司广安供电公司 | The construction method of partial discharge of switchgear diagnostic system based on multi-information fusion |
| CN109871746A (en) * | 2018-12-30 | 2019-06-11 | 国网辽宁省电力有限公司电力科学研究院 | A kind of mode identification method based on the diagnosis of GIS insulation defect |
| CN110303521A (en) * | 2018-03-27 | 2019-10-08 | 清华大学 | Joint of robot torque signals acquisition system and method |
| CN112014700A (en) * | 2020-08-28 | 2020-12-01 | 武汉大学 | GIS insulator defect identification method and system based on partial discharge multi-information fusion |
| CN114998718A (en) * | 2022-04-24 | 2022-09-02 | 南京大学 | Method for predicting activated sludge settlement performance based on deep learning |
| CN115453299A (en) * | 2022-10-31 | 2022-12-09 | 国网电力科学研究院武汉南瑞有限责任公司 | GIS partial discharge laboratory effective simulation and multi-source detection system and method |
-
2023
- 2023-05-22 CN CN202310573291.2A patent/CN116736051A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107728014A (en) * | 2017-08-23 | 2018-02-23 | 国网山东省电力公司检修公司 | Method and system based on multiple sensor signals characteristic identificating equipment insulation defect |
| CN110303521A (en) * | 2018-03-27 | 2019-10-08 | 清华大学 | Joint of robot torque signals acquisition system and method |
| CN109164362A (en) * | 2018-09-25 | 2019-01-08 | 上海交通大学 | A kind of recognition methods and system of direct current cables shelf depreciation defect failure |
| CN109444682A (en) * | 2018-11-02 | 2019-03-08 | 国网四川省电力公司广安供电公司 | The construction method of partial discharge of switchgear diagnostic system based on multi-information fusion |
| CN109871746A (en) * | 2018-12-30 | 2019-06-11 | 国网辽宁省电力有限公司电力科学研究院 | A kind of mode identification method based on the diagnosis of GIS insulation defect |
| CN112014700A (en) * | 2020-08-28 | 2020-12-01 | 武汉大学 | GIS insulator defect identification method and system based on partial discharge multi-information fusion |
| CN114998718A (en) * | 2022-04-24 | 2022-09-02 | 南京大学 | Method for predicting activated sludge settlement performance based on deep learning |
| CN115453299A (en) * | 2022-10-31 | 2022-12-09 | 国网电力科学研究院武汉南瑞有限责任公司 | GIS partial discharge laboratory effective simulation and multi-source detection system and method |
Non-Patent Citations (6)
| Title |
|---|
| 张辉: "基于UWB及惯性测量器件的室内定位方法研究", 中国优秀硕士学位论文全文数据库 信息科技辑, 31 March 2022 (2022-03-31), pages 20 - 23 * |
| 徐敏锐;卢树峰;黄奇峰;王忠东;杨世海;: "基于强跟踪卡尔曼算法的油色谱数据滤波", 电气应用, no. 24, 20 December 2015 (2015-12-20), pages 102 - 107 * |
| 李德军;沈威;郭志强;: "GIS局部放电常规检测和超声波检测方法的应用比较", 高压电器, no. 03, 16 June 2009 (2009-06-16), pages 99 - 103 * |
| 汤会增;: "几种500kV GIS局部放电检测方法的应用比较", 浙江电力, no. 08, 25 August 2016 (2016-08-25), pages 7 - 12 * |
| 白宇峰;何业慎;张翼英;: "基于BP网络的声电联合检测在GIS局部放电的定位研究", 沈阳工程学院学报(自然科学版), no. 03, 15 July 2017 (2017-07-15), pages 73 - 78 * |
| 荣命哲 等: "电器设备状态检测", 31 March 2020, 北京：机械工业出版社, pages: 105 - 106 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118130984A (en) * | 2024-05-10 | 2024-06-04 | 山东博通节能科技有限公司 | Cable partial discharge fault real-time monitoring method based on data driving |
| CN118130984B (en) * | 2024-05-10 | 2024-07-12 | 山东博通节能科技有限公司 | Cable partial discharge fault real-time monitoring method based on data driving |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104090214B (en) | A kind of Cable fault examination and aging analysis method | |
| CN102809718B (en) | Ultra-high-frequency partial discharge signal identification method for gas insulated switchgear (GIS) | |
| CN109596955B (en) | Partial discharge state determination method and device | |
| CN105548832A (en) | High-voltage power cable fault recognition method | |
| CN109490730A (en) | Cable discharge detection method, device, storage medium and processor | |
| CN103267932A (en) | GIS partial discharging detection system and method | |
| CN111239263B (en) | Method and system for detecting foreign matter defects in GIS equipment | |
| CN114217164B (en) | Cable fault distance measurement method and system based on discharge waveform intelligent identification | |
| CN112149549A (en) | GIS partial discharge type identification method based on depth residual error network | |
| CN112749637B (en) | SNN-based distributed optical fiber sensing signal identification method | |
| CN116736051A (en) | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium | |
| CN104007369A (en) | Online power transformer fault positioning system based on ultra-high-frequency external sensors | |
| KR101550689B1 (en) | A arc or corona detection system for a distributing board with the acoustic emission sensor and noise removal function | |
| CN115128345B (en) | Power grid safety early warning method and system based on harmonic monitoring | |
| CN103558532A (en) | Partial discharge on-line detection system of high-voltage crosslinked polyethylene power cable | |
| CN115618205A (en) | Portable voiceprint fault detection system and method | |
| CN111999614A (en) | Local discharge multi-source fusion detection early warning method for high-voltage switch cabinet | |
| CN115993511A (en) | Partial discharge type high-precision detection and identification device, method and equipment | |
| CN116910470A (en) | GIS combined electrical apparatus partial discharge fault mode identification method | |
| KR100544366B1 (en) | system and method for detecting the failures in gas insulated switchgear using fuzzy inference | |
| CN116432054A (en) | Partial discharge severity identification method based on cluster analysis | |
| Mishra et al. | Diagnostics analysis of partial discharge events of the power cables at various voltage levels using ramping behavior analysis method | |
| Sukma et al. | Determination of type of partial discharge in cubicle-type gas insulated switchgear (C-GIS) using artificial neural network | |
| CN105182116A (en) | Transformer winding working state detection method based on weighted gradient structure similarity | |
| Harris et al. | Locating partial discharge using particle swarm optimisation |
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 |