CN107167716A - A kind of shelf depreciation default kind identification method and device - Google Patents
A kind of shelf depreciation default kind identification method and device Download PDFInfo
- Publication number
- CN107167716A CN107167716A CN201710560684.4A CN201710560684A CN107167716A CN 107167716 A CN107167716 A CN 107167716A CN 201710560684 A CN201710560684 A CN 201710560684A CN 107167716 A CN107167716 A CN 107167716A
- Authority
- CN
- China
- Prior art keywords
- discharge
- pulse
- confidence level
- defect type
- accounting
- 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
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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Relating To Insulation (AREA)
Abstract
The present invention provides a kind of shelf depreciation default kind identification method, including step：A, obtain local discharge signal；B, the electric discharge statistical nature for extracting local discharge signal；C, calculating sampling zones accounting；D, input quantity of statistical nature and each sampling zones accounting of discharging as first nerves network classifier and nervus opticus network classifier, first, second neural network classifier each export discharge defect type and its confidence level；E, synthetic determination final discharge defect type and its confidence level.The present invention had both considered the statistical nature of multiple partial discharge pulse's signals, it is further contemplated that multimodal and Vibration Condition in single discharge pulse signal, by neural network classifier, and according to the final discharge defect type of certain regular synthetic determination, the adverse effect that multimodal and oscillating waveform are brought to identification is solved, the accuracy and reliability of final recognition result is improved.
Description
Technical field
The present invention relates to a kind of shelf depreciation default kind identification method and device.
Background technology
Failure and defect may be produced in the apparatus insulated structure running of high-tension electricity, can be caused after longtime running
The failure of insulation breakdown and whole power equipment, so that the reliability of power system is influenceed, it is therefore, apparatus insulated to high-tension electricity
Defect needs to be identified and assess, and Partial Discharge Detection is proved to be to disclose high-tension apparatus defect and assesses its order of severity
Effective means.
The local discharge signal form of expression is single or continuous electrically pulse, and impulse waveform not only contains defect
Discharge mechanism information, and defect order of severity information is also contains, therefore, one of main task of Partial Discharge Detection is exactly
Defect type is identified by partial discharge pulse's waveform, so as to the defect order of severity is diagnosed and formulated rationally
Plant maintenance measure and strategy.
Its existing regularity of partial discharge pulse's waveform, also there is randomness.Its partial discharge pulse of different types of defect
The operating frequency phase of appearance has the certain regularity of itself, and partial discharge pulse's waveform of same type defect is in unimodal pulse
The characterisitic parameters such as its rise time, fall time, pulse width have similitude, and these rules provide for the identification of defect type
Facility.However, due to the complexity of shelf depreciation mechanism, existing so far still without complete thoroughly research, particularly defect
Unimodal pulse not only occurs under certain voltage, multimodal pulse and oscillating impulse also occurs, the opportunity of appearance has randomness,
This brings difficulty for the identification of defect type.
Extensive engineer applied is had in the identification technology of defect type based on phase-resolved statistical recognition method, but
It is due to that this method is based primarily upon the regularity of distribution of the partial discharge pulse in operating frequency phase, the influence being interfered is larger, often results in
The recognition result of mistake, and can not accurately judge the order of severity of defect.And based on partial discharge pulse's waveform characteristic parameter
The method of identification of (such as rise time), either time domain method or frequency domain method, it is more effective for unimodal pulse, to multimodal or shake
Swinging pulse, but effect is poor.Therefore, for multimodal and oscillating impulse, it is necessary to explore a kind of new method to defect is effective to know
Not.
The content of the invention
The purpose of the present invention is in view of the shortcomings of the prior art, to propose a kind of shelf depreciation default kind identification method and dress
Put, while considering unimodal, multimodal and vibratory impulse, improve the accuracy and reliability of final recognition result.
The present invention is achieved through the following technical solutions：
A kind of shelf depreciation default kind identification method, comprises the following steps：
A, the original local discharge signal in the collection multiple power frequency periods of power equipment simultaneously carry out noise reduction process to it, obtain
Local discharge signal；
B, the electric discharge statistical nature for extracting local discharge signal；
C, single partial discharge pulse is extracted from local discharge signal, by dividing single partial discharge pulse's amplitude
And the sampling period, single partial discharge pulse's waveform is divided into multiple region windows, the accounting of each region window is calculated as adopting
Sample subregion accounting；
D, will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier and nervus opticus
The input quantity of network classifier, first, second neural network classifier each exports discharge defect type and its confidence level；
E, the possibility for judging according to each confidence level its corresponding discharge defect type, and combine possibility and confidence level,
Synthetic determination final discharge defect type and its confidence level.
Further, the step C specifically includes following steps：
C1, the maximum sampled point quantity interception for needing local discharge signal according to waveform integrality, obtain single part
Discharge pulse；
C2, single partial discharge pulse's amplitude is normalized, it is [- 1,1] to make its amplitude range；
C3, single partial discharge pulse is divided into n equal subcycles, by the width after normalization the corresponding sampling period
Value is divided into m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region window, wherein, n >=16, m >=
16；
C4, by all effective sampling points sums in the window of region divided by all sampled point sums in the sampling period, be the area
The accounting of domain window.
Further, the step A specifically includes following steps：
A1, the original local discharge signal by being produced in multiple power frequency periods in sensor continuous acquisition power equipment；
A2, original local discharge signal is subjected to signal amplification and analog-to-digital conversion process；
A3, using discrete wavelet noise reduction algorithm, to being handled through step A2 after original discharge signal in white noise and narrow
Band interference is suppressed or eliminated to obtain local discharge signal.
Further, the step B specifically includes following steps：
B1, the local discharge signal collected in multiple power frequency periods is overlapped, obtains discharge capacity-phase-pulse
Several three-dimensional collection of illustrative plates, the acquisition of wherein phase value is by 0°To 360°It is divided into multiple phase intervals at equal intervals, umber of pulse is each phase
The interval actual pulse number in position；
B2, extract discharge capacity-phase, the two-dimensional map of umber of pulse-phase respectively from three-dimensional collection of illustrative plates, and from the two two
Tie up and electric discharge statistical nature is obtained in collection of illustrative plates.
Further, first, second neural network classifier described in the step D is BP neural network classification
Device.
Further, the step E specifically includes following steps：
E1, the possibility for judging according to each confidence level its corresponding discharge defect type, be specially：During confidence level >=9, its
The possibility of corresponding discharge defect type is height, and during 4 ＜ confidence level ＜ 9, the possibility of its corresponding discharge defect type is
In, during confidence level≤4, the possibility of its corresponding discharge defect type is low；
E2, according to certain regular synthetic determination final discharge defect type and its confidence level.
Further, positive-negative half-cycle electric discharge amplitude degree of skewness, positive-negative half-cycle discharge frequency degree of skewness, positive-negative half-cycle electric discharge width
It is worth the degree of asymmetry of kurtosis, positive-negative half-cycle discharge frequency kurtosis and positive-negative half-cycle.
The present invention is also achieved through the following technical solutions：
A kind of shelf depreciation Classifcation of flaws device, including：
Signal acquisition and pretreatment module：For gathering the original local discharge signal in the multiple power frequency periods of power equipment
And noise reduction process is carried out to it, obtain local discharge signal；
Statistical nature extraction module：Electric discharge statistical nature for extracting local discharge signal；
Sampling zones accounting computing module：For extracting single partial discharge pulse from local discharge signal, by drawing
Divide single partial discharge pulse's amplitude and sampling period, single partial discharge pulse's waveform is divided into multiple region windows, count
The accounting of each region window is calculated as sampling zones accounting；
Determination module：For will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier
With the input quantity of nervus opticus network classifier, first, second neural network classifier each export discharge defect type and its
Confidence level；The possibility of its corresponding discharge defect type is judged according to each confidence level, and combines possibility and confidence level, it is comprehensive
Judge final discharge defect type and its confidence level.
Further, the sampling is distinguished accounting computing module and included：
Single partial discharge pulse's acquisition module：For the maximum that local discharge signal needs according to waveform integrality to be adopted
Sampling point quantity is intercepted, and obtains single partial discharge pulse；
Normalize module：For single partial discharge pulse's amplitude to be normalized, make its amplitude range for [-
1,1]；
Region window division module：For single partial discharge pulse to be divided into the corresponding sampling period into n equal son weeks
Phase, the amplitude after normalization is divided into m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region
Window, wherein, n >=16, m >=16；
Accounting computing module：For by all effective sampling points sums divided by all sampled points in the sampling period in the window of region
Sum, is the accounting of the region window.
Further, the statistical nature extraction module includes：
Three-dimensional spectrum acquisition module：For the local discharge signal collected in multiple power frequency periods to be overlapped, obtain
To the three-dimensional collection of illustrative plates of discharge capacity-phase-umber of pulse, the acquisition of wherein phase value is by 0°To 360°It is divided into multiple phases at equal intervals
Interval, umber of pulse is the actual pulse number of each phase interval；
Statistical nature acquisition module：For extracted respectively from three-dimensional collection of illustrative plates discharge capacity-phase, umber of pulse-phase two
Collection of illustrative plates is tieed up, and obtains from the two two-dimensional maps electric discharge statistical nature.
The present invention has the advantages that：
The present invention is by electric discharge statistical nature and sampling zones accounting respectively as first nerves network classifier and the second god
Input quantity through network classifier, judges that final electric discharge lacks further according to the output integrated of first, second neural network classifier
Type and its confidence level are fallen into, the statistical nature of multiple partial discharge pulse's signals had both been considered, it is further contemplated that single discharge pulse signal
In multimodal and Vibration Condition, by neural network classifier, and according to the final discharge defect of certain regular synthetic determination
Type, solves the adverse effect that multimodal and oscillating waveform are brought to identification, improves the accuracy and reliability of final recognition result.
Brief description of the drawings
The present invention is described in further details below in conjunction with the accompanying drawings.
Fig. 1 is flow chart of the invention.
Embodiment
As shown in figure 1, a kind of shelf depreciation default kind identification method, comprises the following steps：
A, the original local discharge signal in the collection multiple power frequency periods of power equipment simultaneously carry out noise reduction process to it, obtain
Local discharge signal, specifically includes following steps：
A1, the original local discharge signal by being produced in multiple power frequency periods in sensor continuous acquisition power equipment,
In the present embodiment, 100 power frequency periods of continuous acquisition, specifically by extra-high video sensor, transient earth voltage sensor, surpass
The local discharge signal produced in sonic sensor or High Frequency Current Sensor coupling electrical power equipment；
A2, original local discharge signal is subjected to signal amplification and analog-to-digital conversion process；
A3, using discrete wavelet noise reduction algorithm, to being handled through step A2 after original discharge signal in white noise and narrow
Band interference is suppressed or eliminated to obtain local discharge signal；
B, the electric discharge statistical nature for extracting local discharge signal, specifically include following steps：
B1, the local discharge signal collected in continuous 100 power frequency periods is overlapped, obtain discharge capacity-phase-
The three-dimensional collection of illustrative plates of umber of pulse, the acquisition of wherein phase value is by 0°To 360°It is divided into multiple phase intervals at equal intervals, umber of pulse is each
The actual pulse number of individual phase interval；
B2, extract discharge capacity-phase, the two-dimensional map of umber of pulse-phase respectively from three-dimensional collection of illustrative plates, and from the two two
Tie up and electric discharge statistical nature is obtained in collection of illustrative plates；
Electric discharge statistical nature includes：Positive-negative half-cycle electric discharge amplitude degree of skewness, positive-negative half-cycle discharge frequency degree of skewness, positive and negative half
The degree of asymmetry of Zhou Fang electricity amplitudes kurtosis, positive-negative half-cycle discharge frequency kurtosis and positive-negative half-cycle；
C, single partial discharge pulse is extracted from local discharge signal, by dividing single partial discharge pulse's amplitude
And the sampling period, single partial discharge pulse's waveform is divided into multiple region windows, the accounting of each region window is calculated as adopting
Sample subregion accounting, specifically includes following steps：
C1, the maximum sampled point quantity interception for needing local discharge signal according to waveform integrality, obtain single part
Discharge pulse；
C2, single partial discharge pulse's amplitude is normalized, it is [- 1,1] to make its amplitude range；
C3, single partial discharge pulse is divided into n equal subcycles, by the width after normalization the corresponding sampling period
Value is divided into m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region window, wherein, n >=16, m >=
16；
C4, by all effective sampling points sums in the window of region divided by all sampled point sums in the sampling period, be the area
The accounting of domain window, effective sampling points are the sampled point corresponding to the amplitude being located in single partial discharge pulse in the region window；
D, will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier and nervus opticus
The input quantity of network classifier, first, second neural network classifier each exports discharge defect type and its confidence level, electric discharge
Defect type includes corona discharge, bubble-discharge and creeping discharge, and confidence level characterizes partial discharge pulse and discharge defect type
Matching degree, confidence level it is higher represent matching degree it is higher, differentiate result it is more accurate；
In the present embodiment, first, second neural network classifier is BP neural network grader, BP neural network point
Class device is three-decker, i.e. input layer, hidden layer and output layer, and electric discharge statistical nature or sampling zones accounting are used as input layer
Neuron, output neuron is corona discharge, bubble-discharge and the corresponding confidence level of three kinds of discharge defect types of creeping discharge；
E, the possibility for judging according to each confidence level its corresponding discharge defect type, and combine possibility and confidence level,
Synthetic determination final discharge defect type and its confidence level, specifically include following steps：
E1, the possibility for judging according to each confidence level its corresponding discharge defect type, be specially：During confidence level >=9, its
The possibility of corresponding discharge defect type is height, and during 4 ＜ confidence level ＜ 9, the possibility of its corresponding discharge defect type is
In, during confidence level≤4, the possibility of its corresponding discharge defect type is low；
E2, according to certain regular synthetic determination final discharge defect type and its confidence level：If first, second nerve
The discharge defect type that network classifier is provided is consistent, then final discharge defect type is exactly the discharge defect type, confidence level
For the average value of first, second neural network classifier confidence level；If the electric discharge that first, second neural network classifier is provided lacks
Type-Inconsistencies are fallen into, but the possibility that provides of one of neural network classifier is height, then final discharge defect type is can
Energy property is the discharge defect type that high neural network classifier is provided, and confidence level is the 1/ of the neural network classifier confidence level
2；If the possibility that first, second neural network classifier is provided is not high, if now first nerves network classifier is provided
Possibility be that during the possibility that first nerves network classifier is provided is or low, then final discharge defect type is the
The discharge defect type that one neural network classifier is provided, confidence level is the 1/3 of statistical confidence；Finally discharged in the case of other
Defect type is uncertain, and confidence level is 0.
A kind of shelf depreciation Classifcation of flaws device, including：
Signal acquisition and pretreatment module：For gathering the original local discharge signal in the multiple power frequency periods of power equipment
And noise reduction process is carried out to it, obtain local discharge signal；
Statistical nature extraction module：Electric discharge statistical nature for extracting local discharge signal；
Sampling zones accounting computing module：For extracting single partial discharge pulse from local discharge signal, by drawing
Divide single partial discharge pulse's amplitude and sampling period, single partial discharge pulse's waveform is divided into multiple region windows, count
The accounting of each region window is calculated as sampling zones accounting；
Determination module：For will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier
With the input quantity of nervus opticus network classifier, first, second neural network classifier each export discharge defect type and its
Confidence level；The possibility of its corresponding discharge defect type is judged according to each confidence level, and combines possibility and confidence level, it is comprehensive
Judge final discharge defect type and its confidence level.
Wherein, sampling is distinguished accounting computing module and included：
Single partial discharge pulse's acquisition module：For the maximum that local discharge signal needs according to waveform integrality to be adopted
Sampling point quantity is intercepted, and obtains single partial discharge pulse；
Normalize module：For single partial discharge pulse's amplitude to be normalized, make its amplitude range for [-
1,1]；
Region window division module：For single partial discharge pulse to be divided into the corresponding sampling period into n equal son weeks
Phase, the amplitude after normalization is divided into m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region
Window, wherein, n >=16, m >=16；
Accounting computing module：For by all effective sampling points sums divided by all sampled points in the sampling period in the window of region
Sum, is the accounting of the region window.
Signal acquisition includes with pretreatment module：
Acquisition module：For the original part by being produced in multiple power frequency periods in sensor continuous acquisition power equipment
Discharge signal；
Pretreatment module：For original local discharge signal to be carried out into signal amplification and analog-to-digital conversion process；
Noise reduction module：For using discrete wavelet noise reduction algorithm, in the original discharge signal after step A2 processing
White noise and arrowband interference are suppressed or eliminated to obtain local discharge signal.
Statistical nature extraction module includes：
Three-dimensional spectrum acquisition module：For the local discharge signal collected in multiple power frequency periods to be overlapped, obtain
To the three-dimensional collection of illustrative plates of discharge capacity-phase-umber of pulse, the acquisition of wherein phase value is by 0°To 360°It is divided into multiple phases at equal intervals
Interval, umber of pulse is the actual pulse number of each phase interval；
Statistical nature acquisition module：For extracted respectively from three-dimensional collection of illustrative plates discharge capacity-phase, umber of pulse-phase two
Collection of illustrative plates is tieed up, and obtains from the two two-dimensional maps electric discharge statistical nature.
The foregoing is only a preferred embodiment of the present invention, therefore the scope that the present invention is implemented can not be limited with this, i.e.,
The equivalent changes and modifications made according to scope of the present invention patent and description, all should still belong to what patent of the present invention covered
In the range of.
Claims (10)
1. a kind of shelf depreciation default kind identification method, it is characterised in that：Comprise the following steps：
A, the original local discharge signal in the collection multiple power frequency periods of power equipment simultaneously carry out noise reduction process to it, obtain part
Discharge signal；
B, the electric discharge statistical nature for extracting local discharge signal；
C, single partial discharge pulse is extracted from local discharge signal, by dividing single partial discharge pulse's amplitude and adopting
In the sample cycle, single partial discharge pulse's waveform is divided into multiple region windows, calculates the accounting of each region window as sampling point
Area's accounting；
D, will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier and nervus opticus network
The input quantity of grader, first, second neural network classifier each exports discharge defect type and its confidence level；
E, the possibility for judging according to each confidence level its corresponding discharge defect type, and possibility and confidence level are combined, it is comprehensive
Judge final discharge defect type and its confidence level.
2. a kind of shelf depreciation default kind identification method according to claim 1, it is characterised in that：The step C tools
Body comprises the following steps：
C1, the maximum sampled point quantity interception for needing local discharge signal according to waveform integrality, obtain single shelf depreciation
Pulse；
C2, single partial discharge pulse's amplitude is normalized, it is [- 1,1] to make its amplitude range；
C3, single partial discharge pulse is divided into n equal subcycles, by the amplitude after normalization point the corresponding sampling period
For m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region window, wherein, n >=16, m >=16；
C4, by all effective sampling points sums in the window of region divided by all sampled point sums in the sampling period, be the region window
Accounting.
3. a kind of shelf depreciation default kind identification method according to claim 1, it is characterised in that：The step A tools
Body comprises the following steps：
A1, the original local discharge signal by being produced in multiple power frequency periods in sensor continuous acquisition power equipment；
A2, original local discharge signal is subjected to signal amplification and analog-to-digital conversion process；
A3, using discrete wavelet noise reduction algorithm, to being handled through step A2 after original discharge signal in white noise and arrowband it is dry
Disturb and suppressed or eliminated to obtain local discharge signal.
4. a kind of shelf depreciation default kind identification method according to claim 1 or 2 or 3, it is characterised in that：The step
Rapid B specifically includes following steps：
B1, the local discharge signal collected in multiple power frequency periods is overlapped, obtains discharge capacity-phase-umber of pulse
The acquisition of three-dimensional collection of illustrative plates, wherein phase value is by 0°To 360°It is divided into multiple phase intervals at equal intervals, umber of pulse is each phase region
Between actual pulse number；
B2, extract discharge capacity-phase, the two-dimensional map of umber of pulse-phase respectively from three-dimensional collection of illustrative plates, and from the two X-Y schemes
Electric discharge statistical nature is obtained in spectrum.
5. a kind of shelf depreciation default kind identification method according to claim 1 or 2 or 3, it is characterised in that：The step
First, second neural network classifier described in rapid D is BP neural network grader.
6. a kind of shelf depreciation default kind identification method according to claim 1 or 2 or 3, it is characterised in that：The step
Rapid E specifically includes following steps：
E1, the possibility for judging according to each confidence level its corresponding discharge defect type, be specially：During confidence level >=9, its correspondence
Discharge defect type possibility for height, during 4 ＜ confidence level ＜ 9, during the possibility of its corresponding discharge defect type is, put
During reliability≤4, the possibility of its corresponding discharge defect type is low；
E2, according to certain regular synthetic determination final discharge defect type and its confidence level.
7. a kind of shelf depreciation default kind identification method according to claim 4, it is characterised in that：The electric discharge statistics
Feature includes：Positive-negative half-cycle electric discharge amplitude degree of skewness, positive-negative half-cycle discharge frequency degree of skewness, positive-negative half-cycle electric discharge amplitude kurtosis,
The degree of asymmetry of positive-negative half-cycle discharge frequency kurtosis and positive-negative half-cycle.
8. a kind of shelf depreciation Classifcation of flaws device, it is characterised in that：Including：
Signal acquisition and pretreatment module：For gathering the original local discharge signal in the multiple power frequency periods of power equipment and right
It carries out noise reduction process, obtains local discharge signal；
Statistical nature extraction module：Electric discharge statistical nature for extracting local discharge signal；
Sampling zones accounting computing module：, should by dividing for extracting single partial discharge pulse from local discharge signal
Single partial discharge pulse's amplitude and sampling period, single partial discharge pulse's waveform is divided into multiple region windows, calculates every
The accounting of individual region window is used as sampling zones accounting；
Determination module：For will electric discharge statistical nature and each sampling zones accounting respectively as first nerves network classifier and the
The input quantity of two neural network classifiers, first, second neural network classifier each exports discharge defect type and its confidence
Degree；The possibility of its corresponding discharge defect type is judged according to each confidence level, and combines possibility and confidence level, synthetic determination
Final discharge defect type and its confidence level.
9. a kind of shelf depreciation Classifcation of flaws device according to claim 8, it is characterised in that：The sampling is distinguished
Accounting computing module includes：
Single partial discharge pulse's acquisition module：For the maximum sampled point for needing local discharge signal according to waveform integrality
Quantity is intercepted, and obtains single partial discharge pulse；
Normalize module：For single partial discharge pulse's amplitude to be normalized, it is [- 1,1] to make its amplitude range；
Region window division module：For single partial discharge pulse to be divided into the corresponding sampling period into n equal subcycles, incited somebody to action
Amplitude after normalization is divided into m equal subregions, i.e., single partial discharge pulse's waveform is divided into n × m region window, its
In, n >=16, m >=16；
Accounting computing module：For all sampled points by all effective sampling points sums in the window of region divided by the sampling period it
With, be the region window accounting.
10. a kind of shelf depreciation Classifcation of flaws device according to claim 8 or claim 9, it is characterised in that：The statistics
Characteristic extracting module includes：
Three-dimensional spectrum acquisition module：For the local discharge signal collected in multiple power frequency periods to be overlapped, put
The three-dimensional collection of illustrative plates of electricity-phase-umber of pulse, the acquisition of wherein phase value is by 0°To 360°It is divided into multiple phase intervals at equal intervals,
Umber of pulse is the actual pulse number of each phase interval；
Statistical nature acquisition module：X-Y scheme for extracting discharge capacity-phase, umber of pulse-phase respectively from three-dimensional collection of illustrative plates
Spectrum, and obtain from the two two-dimensional maps electric discharge statistical nature.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710560684.4A CN107167716A (en) | 2017-07-11 | 2017-07-11 | A kind of shelf depreciation default kind identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710560684.4A CN107167716A (en) | 2017-07-11 | 2017-07-11 | A kind of shelf depreciation default kind identification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107167716A true CN107167716A (en) | 2017-09-15 |
Family
ID=59823648
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710560684.4A Pending CN107167716A (en) | 2017-07-11 | 2017-07-11 | A kind of shelf depreciation default kind identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107167716A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107831412A (en) * | 2017-10-11 | 2018-03-23 | 国网上海市电力公司 | Local discharge of electrical equipment localization method and system, storage medium and terminal device |
| CN108053136A (en) * | 2017-12-27 | 2018-05-18 | 上海储翔信息科技有限公司 | A kind of car dealer's analysis on Achievements expert system based on neutral net |
| CN108169643A (en) * | 2018-01-30 | 2018-06-15 | 西南石油大学 | A kind of method and system for cable local discharge pattern-recognition |
| 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 |
| CN110208660A (en) * | 2019-06-05 | 2019-09-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of training method and device for power equipment shelf depreciation defect diagonsis |
| CN110310260A (en) * | 2019-06-19 | 2019-10-08 | 北京百度网讯科技有限公司 | Sub-material decision-making technique, equipment and storage medium based on machine learning model |
| CN110470960A (en) * | 2019-09-05 | 2019-11-19 | 国网北京市电力公司 | The analysis method and device of cable local discharge, storage medium and processor |
| CN110531234A (en) * | 2019-09-26 | 2019-12-03 | 武汉三相电力科技有限公司 | A kind of identification extracting method of transmission line of electricity discharge pulse |
| CN110794264A (en) * | 2019-10-14 | 2020-02-14 | 浙江浙能技术研究院有限公司 | Generator partial discharge type identification method based on time domain pulse characteristics |
| CN112034309A (en) * | 2020-07-28 | 2020-12-04 | 北京交通大学 | Partial discharge identification method for high-voltage switch cabinet |
| JP2021015098A (en) * | 2019-07-16 | 2021-02-12 | 株式会社東芝 | Partial discharge diagnostic device, partial discharge diagnostic method, learning device, learning method, partial discharge diagnostic system, and computer program |
| CN112734012A (en) * | 2021-01-07 | 2021-04-30 | 北京灵汐科技有限公司 | Impulse neural network training method, data processing method, electronic device, and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102854445A (en) * | 2012-10-18 | 2013-01-02 | 上海市电力公司 | Method for extracting waveform feature of local discharge pulse current |
| CN102955108A (en) * | 2012-10-25 | 2013-03-06 | 清华大学 | Method for measuring converter transformer paper oil insulation partial discharge characteristic |
| CN103076547A (en) * | 2013-01-24 | 2013-05-01 | 安徽省电力公司亳州供电公司 | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines |
| CN103323749A (en) * | 2013-05-16 | 2013-09-25 | 上海交通大学 | Multi-classifier information fusion partial discharge diagnostic method |
| CN103698669A (en) * | 2013-12-12 | 2014-04-02 | 广东电网公司东莞供电局 | Local discharge pulse separating method for high-voltage power equipment |
| CN106199351A (en) * | 2016-06-27 | 2016-12-07 | 国网北京市电力公司 | The sorting technique of local discharge signal and device |
-
2017
- 2017-07-11 CN CN201710560684.4A patent/CN107167716A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102854445A (en) * | 2012-10-18 | 2013-01-02 | 上海市电力公司 | Method for extracting waveform feature of local discharge pulse current |
| CN102955108A (en) * | 2012-10-25 | 2013-03-06 | 清华大学 | Method for measuring converter transformer paper oil insulation partial discharge characteristic |
| CN103076547A (en) * | 2013-01-24 | 2013-05-01 | 安徽省电力公司亳州供电公司 | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines |
| CN103323749A (en) * | 2013-05-16 | 2013-09-25 | 上海交通大学 | Multi-classifier information fusion partial discharge diagnostic method |
| CN103698669A (en) * | 2013-12-12 | 2014-04-02 | 广东电网公司东莞供电局 | Local discharge pulse separating method for high-voltage power equipment |
| CN106199351A (en) * | 2016-06-27 | 2016-12-07 | 国网北京市电力公司 | The sorting technique of local discharge signal and device |
Non-Patent Citations (2)
| Title |
|---|
| 司文荣等: "局部放电脉冲信号ICA 提取技术的初步研究", 《高电压技术》 * |
| 季洪鑫等: "冲击电压下气体绝缘开关设备悬浮缺陷放电特征", 《电工技术学报》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107831412A (en) * | 2017-10-11 | 2018-03-23 | 国网上海市电力公司 | Local discharge of electrical equipment localization method and system, storage medium and terminal device |
| CN107831412B (en) * | 2017-10-11 | 2020-02-21 | 国网上海市电力公司 | Electrical equipment partial discharge positioning method and system, storage medium and terminal equipment |
| CN108053136A (en) * | 2017-12-27 | 2018-05-18 | 上海储翔信息科技有限公司 | A kind of car dealer's analysis on Achievements expert system based on neutral net |
| CN108169643A (en) * | 2018-01-30 | 2018-06-15 | 西南石油大学 | A kind of method and system for cable local discharge pattern-recognition |
| 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 |
| CN110208660A (en) * | 2019-06-05 | 2019-09-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of training method and device for power equipment shelf depreciation defect diagonsis |
| CN110208660B (en) * | 2019-06-05 | 2021-07-27 | 国网江苏省电力有限公司电力科学研究院 | Training method and device for diagnosing partial discharge defects of power equipment |
| CN110310260A (en) * | 2019-06-19 | 2019-10-08 | 北京百度网讯科技有限公司 | Sub-material decision-making technique, equipment and storage medium based on machine learning model |
| CN110310260B (en) * | 2019-06-19 | 2021-10-15 | 北京百度网讯科技有限公司 | Material distribution decision method, equipment and storage medium based on machine learning model |
| JP7297575B2 (en) | 2019-07-16 | 2023-06-26 | 株式会社東芝 | Partial discharge diagnosis device, partial discharge diagnosis method, partial discharge diagnosis system, and computer program |
| JP2021015098A (en) * | 2019-07-16 | 2021-02-12 | 株式会社東芝 | Partial discharge diagnostic device, partial discharge diagnostic method, learning device, learning method, partial discharge diagnostic system, and computer program |
| CN110470960A (en) * | 2019-09-05 | 2019-11-19 | 国网北京市电力公司 | The analysis method and device of cable local discharge, storage medium and processor |
| CN110531234A (en) * | 2019-09-26 | 2019-12-03 | 武汉三相电力科技有限公司 | A kind of identification extracting method of transmission line of electricity discharge pulse |
| CN110794264A (en) * | 2019-10-14 | 2020-02-14 | 浙江浙能技术研究院有限公司 | Generator partial discharge type identification method based on time domain pulse characteristics |
| CN110794264B (en) * | 2019-10-14 | 2021-11-19 | 浙江浙能技术研究院有限公司 | Generator partial discharge type identification method based on time domain pulse characteristics |
| CN112034309B (en) * | 2020-07-28 | 2021-11-23 | 北京交通大学 | Partial discharge identification method for high-voltage switch cabinet |
| CN112034309A (en) * | 2020-07-28 | 2020-12-04 | 北京交通大学 | Partial discharge identification method for high-voltage switch cabinet |
| CN112734012A (en) * | 2021-01-07 | 2021-04-30 | 北京灵汐科技有限公司 | Impulse neural network training method, data processing method, electronic device, and medium |
| CN112734012B (en) * | 2021-01-07 | 2024-03-05 | 北京灵汐科技有限公司 | Impulse neural network training method, data processing method, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107167716A (en) | A kind of shelf depreciation default kind identification method and device | |
| CN104198898B (en) | Local discharge development process diagnosis method based on pulse-train analysis | |
| CN103645425B (en) | High-voltage cable insulation defect partial discharge on-line monitoring diagnosis method | |
| CN104749468B (en) | A kind of GIS fault diagnosis systems and its method | |
| CN105548832B (en) | High voltage power cable fault recognition method | |
| CN102597790B (en) | Device and method for detecting and processing signals relating to partial electrical discharges | |
| CN102539527B (en) | GIS (gas insulated substation) partial discharge mode identification method based on ultrasonic testing | |
| CN101907437B (en) | Wavelet difference algorithm-based cable fault localization method | |
| CN103197218B (en) | A kind of high-voltage cable insulation defect partial discharge electrification detection diagnostic method | |
| CN107329079A (en) | A kind of primary cut-out on-line monitoring and synthetic fault diagnosis system | |
| CN103091612B (en) | Separation and recognition algorithm for transformer oiled paper insulation multiple partial discharging source signals | |
| EP1094323A2 (en) | Method and system for identifying cause of partial discharge | |
| CN108693448B (en) | Partial discharge mode recognition system applied to power equipment | |
| CN105044566B (en) | A kind of GIS partial discharge fault detection method of feature based ultra-high frequency signal | |
| CN105676088A (en) | Device and method for testing fault arc detection apparatus | |
| CN107942210A (en) | The classification of transformer pulse electric current Partial Discharge and denoising method and system | |
| CN102792173A (en) | Instrument and method for detecting partial electrical discharges | |
| CN112668612A (en) | Partial discharge signal clustering analysis method based on grids | |
| CN102792174B (en) | Instrument and method for detecting partial electrical discharges | |
| CN116280628A (en) | Partial discharge test method for high-voltage power equipment withstand voltage test | |
| CN101802627B (en) | A method and an apparatus for monitoring an activity of partial electrical discharges in an electrical apparatus powered with direct voltage | |
| CN106680574B (en) | A kind of perception of substation equipment overvoltage and data processing method | |
| CN114113943B (en) | Transformer partial discharge detection system, method and equipment based on current and ultrasonic signals | |
| EP4001936B1 (en) | Automatic partial discharge and noise signals separation using arithmetic coding in time domain and magnitude distributions in frequency domain | |
| CN102478618A (en) | Online monitoring method for partial discharging of 500 KV cross-linked cable |
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 |
Application publication date: 20170915 |
|
| RJ01 | Rejection of invention patent application after publication |