CN104865508B - The Recognition of Partial Discharge quantified based on packet - Google Patents
The Recognition of Partial Discharge quantified based on packet Download PDFInfo
- Publication number
- CN104865508B CN104865508B CN201510309495.0A CN201510309495A CN104865508B CN 104865508 B CN104865508 B CN 104865508B CN 201510309495 A CN201510309495 A CN 201510309495A CN 104865508 B CN104865508 B CN 104865508B
- Authority
- CN
- China
- Prior art keywords
- pulse
- partial discharge
- packet
- recognition
- shelf depreciation
- 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.)
- Active
Links
Landscapes
- Testing Relating To Insulation (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
A kind of Recognition of Partial Discharge quantified based on packet, is comprised the following steps：Step 1：Gather Partial Discharge Data；Step 2：The Partial Discharge Data collected is pre-processed, burst length sequence is obtained；Step 3：Burst length sequence to acquisition carries out group quantization, counts the pulse data of each quantized interval；Step 4：FFT is carried out to packet pulse data, and calculating obtains the frequency values of amplitude one；Step 5：According to amplitude frequency values, the energy under different frequency is calculated；Step 6：Judge whether occur shelf depreciation according to energy accounting, complete the identification of shelf depreciation.Dependence instant invention overcomes conventional method to phase information and artificial threshold value, can monitor whether occur shelf depreciation at any time in the case where different light makes an uproar background.
Description
Technical field
The present invention relates to shelf depreciation flash spotting field, the shelf depreciation identification more particularly to quantified based on packet
Method.
Background technology
The high-tension apparatuses such as power transformer occupy an important position in power system.Shelf depreciation can be led in actual motion
Send a telegraph power inside transformer and produce insulation degradation, can if things go on like this influence the normal work of equipment and bring potential safety hazard.Play a game
The detection and identification of portion's electric discharge are one of important means of current monitoring transformer station high-voltage side bus situation.Shelf depreciation can be supervened
Electricity, optical phenomenon, therefore conventional detection method for local discharge is main on this basis, is divided into electrical measuring method and flash spotting.
Electrical measuring method is by detecting impedance or current sensor, and the transient electrical pulses produced during detection shelf depreciation pass through meter
Calculation obtains Apparent discharge magnitude to weigh shelf depreciation.This method is serious by live electromagnetic interference, it is impossible to be effectively applied to
Line is detected, and detection sensitivity is low, and detection error is big.
Flash spotting is mainly by detecting that basis for estimation is used as in the light radiation that shelf depreciation is produced.Experiment shows, shelf depreciation
The optical wavelength of generation is distributed between 200-800nm, wherein early stage of discharging, wavelength is shorter, based on 280-400nm ultraviolet lights.
Shelf depreciation is disengaged by the incoming electrooptical device of light by optical fiber, office is analyzed by handling the electrical signal data being converted to
Discharge in portion.Because the material for receiving optical signal is non-electric material, thus electromagnetic interference can be effectively prevented from, in addition opto-electronic conversion
The high sensitivity of device also makes flash spotting have the advantages that sensitivity is high.Depend on more and manually set traditional flash spotting data processing
Determine threshold value, it is believed that the pulse that amplitude exceedes certain threshold value is discharge pulse, or think that umber of pulse exceedes necessarily in a cycle
Shelf depreciation, or acquisition phase information then occur for threshold value, by the phase for setting up related physical quantity (pulse amplitude, umber of pulse etc.)
Collection of illustrative plates completes the identification of shelf depreciation.In laboratory environments, phase information is easily obtained, and lucifuge condition is preferable, can be with
Ensureing the data of multi collect has good noise uniformity.And due to the complexity of equipment, phase in actually detected and operation
Position information is extremely difficult to be obtained, and the change such as the lucifuge condition of ambient light conditions, equipment is manually difficult to set suitable threshold than larger
Value.
It can be used in detecting shelf depreciation at any time therefore, it is necessary to find one kind, without artificial given threshold, without collection
The data processing method of phase information.
The content of the invention
It is a primary object of the present invention to provide a kind of Recognition of Partial Discharge quantified based on packet, this method
By carrying out classified statistics and FFT to the discharge data once collected, by calculating after FFT under different frequency
Energy ratio complete shelf depreciation identification, overcome dependence of the conventional method to phase information and artificial threshold value, can be not
With light make an uproar and monitor whether occur shelf depreciation under background at any time.
The present invention provides a kind of Recognition of Partial Discharge quantified based on packet, comprises the following steps：
Step 1：Gather Partial Discharge Data；
Step 2：The Partial Discharge Data collected is pre-processed, pulse-time sequence is obtained；
Step 3：Pulse-time sequence to acquisition carries out group quantization, counts the pulse data of each quantized interval；
Step 4：FFT is carried out to packet pulse data, and calculating obtains amplitude-frequency values；
Step 5：According to amplitude-frequency values, the energy under different frequency is calculated；
Step 6：Judge whether occur shelf depreciation according to energy accounting, complete the identification of shelf depreciation.
The present invention has following technique effect：
1st, the Recognition of Partial Discharge quantified based on packet that the present invention is provided, can be from once collecting
Analysis obtains result of discharging in flash spotting Partial Discharge Data；
2nd, the Recognition of Partial Discharge quantified based on packet that the present invention is provided, without relying on phase information and people
Work threshold value, can be measured and identification equipment shelf depreciation situation at any time.
3rd, the Recognition of Partial Discharge quantified based on packet that the present invention is provided, can make an uproar background in different light
Lower processing completes the identification of shelf depreciation.
Brief description of the drawings
For further illustrate the present invention technology contents, below in conjunction with accompanying drawing and case study on implementation to the detailed description of the invention such as
Under, wherein：
A kind of flow chart for Recognition of Partial Discharge quantified based on packet that Fig. 1 provides for the present invention.
Fig. 2 is carries out the flow chart of group quantization to the data that collect in the present invention.
Fig. 3 is to count 16 obtained pulse number accumulated value knots after group quantization under 3kv voltages in the embodiment of the present invention
Fruit is schemed.
Fig. 4 is to count 16 obtained pulse number accumulated values after group quantization under 4.7kv voltages in the embodiment of the present invention
Result figure.
Fig. 5 is does after 16 FFTs to the statistical result shown in Fig. 3 in the present invention, amplitude-frequency diagram of drafting.
Fig. 6 is does after 16 FFTs to the statistical result shown in Fig. 4 in the present invention, amplitude-frequency diagram of drafting.
Embodiment
Refer to shown in Fig. 1 and Fig. 2, the present invention provides a kind of Recognition of Partial Discharge quantified based on packet,
Comprise the following steps：
Step 1 (101)：Partial Discharge Data is gathered, the data are discrete voltage-time number that electrooptical device is exported
According to amplitude non-negative.
Step 2 (102)：The Partial Discharge Data collected is pre-processed, pulse-time sequence is obtained, it is described
Preprocessing process is filters out the less ground noise of amplitude first, and the partial noise derives from device or Acquisition Circuit in itself,
Noise amplitude is smaller, and quantity is intensive, and subsequent treatment can be impacted, therefore is filtered out.Filtering method is given threshold Th,
Value less than Th is set to 0, the judgement whether the threshold value Th set herein occurs not as shelf depreciation, Th regards acquisition system
Depending on sensitivity and noise size.Then data smoothing is carried out using Neighborhood Filtering, data smoothing is used for removing due to data wave
The dynamic local pseudo- maximum point caused, while continuous two identical values are avoided the occurrence of, because continuous two identical values can be right
Pulse statistics is impacted.The Size of Neighborhood taken makees minor alteration depending on the sampling interval, generally takes 5 neighborhoods.Final search owns
Local modulus maxima, a Local modulus maxima represents a pulse.While search obtains Local modulus maxima, should also
Record maximum point where position, the position as follow-up group quantization foundation.
Step 3 (103)：Pulse-time sequence to acquisition carries out group quantization, counts the umber of pulse of each quantized interval
According to.
The step 3 (103) specifically includes：
Step 31 (201)：Continuous N number of power frequency period, described N number of work are intercepted from obtained pulse-time sequence
The frequency cycle is equal to N number of power frequency period only in length, does not consider initial phase, initial phase is arbitrary value, it is desirable to this N number of cycle
It is not overlapping, it is not broken, N is sufficiently large, to ensure the validity of statistics；
Step 32 (202)：Each power frequency period average quantization is interval into K, and the length of a power frequency period is 20ms,
After quantization it is each it is interval be (20/K) ms, quantized interval K value should be appropriate, it is too small if phase separation it is not obvious enough, mistake
Statistical error can become big if big, and it is 16 that K is chosen herein.The pulse number in each interval during statistics K is interval.Due to working as
Shelf depreciation will be produced by applying when voltage is more than the starting voltage of shelf depreciation, therefore partial discharge pulse is presented one in phase
Fixed distribution, it is to utilize this phase distribution feature that quantization is carried out to the cycle.Count the pulse number of each quantized interval, arteries and veins
The distribution for rushing number embodies the phase distribution of shelf depreciation.Described pulse number physical quantity can be by pulse amplitude or pulse
Energy Equivalent is replaced；
Step 33 (203)：The pulse number statistical result being under N number of cycle in identical quantized interval is added up,
Obtain the accumulated value of K group pulse numbers.Consider partial discharge pulse's phase distribution feature, the pulse number in multiple cycles is added up
Be conducive to highlighting this feature, improve the effect of statistics.Counting obtained pulse number accumulated value can optionally normalize
To a power frequency period；
Step 4 (104)：FFT is carried out to packet pulse data, and calculating obtains amplitude-frequency values.Put according to local
The phase distribution feature of electricity, when occurring shelf depreciation, the pulse number statistical value after quantization can be in the interval for discharge pulse occur
Peak value is presented, the peak value, which is appeared in, applies alive instantaneous value more than on the quantized interval residing for firing potential, that is, applies
Shown as near the peak value of voltage positive and negative half period, in phase near 90 ° and 270 °.Initial phase is unknown during due to statistics, institute
Can not determine the particular location at two peaks, but its relative position is fixed, is 180 °, is showed in K quantized interval
For at a distance of K/2.
Statistical result to group quantization makees FFT, calculates the range value under each frequency.Pass through FFT energy
Enough judge that the pulse number after quantitative statisticses whether there is the peak of fixed frequency.
Step 5 (105)：According to amplitude-frequency values, calculate under the energy under different frequency, the calculating different frequency
Energy, is that the amplitude progress square under each frequency after FFT is obtained into energy value.
Step 6 (106)：Judge whether occur shelf depreciation according to energy accounting, it is described to be judged whether according to energy accounting
Generation shelf depreciation, be judge energy under 50Hz and 100Hz and whether more than other frequencies energy and, if being then considered as hair
Raw shelf depreciation, is otherwise considered as and does not occur shelf depreciation, and other described frequencies do not include DC component, completes shelf depreciation
Identification.
Analyzed according to above-mentioned steps, the phase profile of shelf depreciation shows at 90 ° and 270 ° discharge pulse occur,
In group quantization counts obtained pulse number data, two peaks are shown as.The relative position at two peaks is fixed as packet count K
Half, because K is the quantization length under a power frequency period, therefore after FFT, if occur shelf depreciation,
Signal amplitude at 100Hz can be big compared with the amplitude at other frequencies.
Additionally, it is contemplated that during actual discharge, 90 ° different with pulse number increasing degree at 270 °, group quantization statistics is obtained
Pulse number in two peaks size it is also different, so after FFT, in addition to 100Hz signal, 50Hz letter
Number also can be big compared with other frequency signals.Therefore, can be by 50Hz signal and 100Hz signal when judging whether to discharge
Energy be added, if both and the signal energy more than other frequencies and, show occur shelf depreciation.And part does not occur
During electric discharge, the pulse number statistical result that group quantization is obtained, without obvious peak value on each quantized interval, thus FFT
In frequency afterwards, no 100Hz and 50Hz protrusion signal.
A kind of effect of the Recognition of Partial Discharge quantified based on packet provided for the checking present invention, is carried out such as
Lower experiment：
The Partial Discharge Data of 3kv and 4.7kv under needle to board electrode is gathered, it is long that interception obtains 100 power frequency periods, will be every
Individual cycle average quantization is 16 intervals, and each interval width is 1.25ms.Count obtained pulse number accumulated value such as Fig. 3 with
Shown in Fig. 4, it can be seen that pulse number statistical result fluctuates very little on different quantized intervals during 3kv, and
Occur more obvious two peaks during 4.7kv, the result to statistics carries out 16 points of FFT, the part after conversion
Amplitude-frequency values as shown in Figure 5 and Figure 6, DC component are not shown in figure.From fig. 5, it can be seen that radio-frequency component accounting during 3kv
Larger, 50Hz and 100Hz components are smaller, and in Fig. 6 during 4.7kv 50Hz and 100Hz components and exceed high fdrequency component.It can obtain
Shelf depreciation has occurred when going out 4.7kv and has not yet occurred during 3kv.
It can be seen that the packet method for being applied to the identification of flash spotting shelf depreciation that the present invention is provided is believed independent of phase
Under breath and artificial threshold value, it can be efficiently identified out from the data once gathered and whether occur shelf depreciation.
It should be noted that above-mentioned involved physical quantity and conversion are not limited in what is mentioned in embodiment, ability
Those of ordinary skill in domain can carry out simple well known replacement to it, for example：
(1) pulse number of statistics is substituted for pulse amplitude, energy or can be by the other forms that are simply calculated
Physical quantity；
(2) FFT is substituted for other time domains to the conversion of frequency domain or wavelet field.
Particular embodiments described above, has been carried out further in detail to the purpose of the present invention, technical scheme and beneficial effect
Describe in detail it is bright, should be understood that the foregoing is only the present invention specific embodiment, be not intended to limit the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvements done etc., should be included in the guarantor of the present invention
Within the scope of shield.
Claims (8)
1. a kind of Recognition of Partial Discharge quantified based on packet, is comprised the following steps：
Step 1：Gather Partial Discharge Data；
Step 2：The Partial Discharge Data collected is pre-processed, pulse-time sequence is obtained；
Step 3：Pulse-time sequence to acquisition carries out group quantization, counts the pulse data of each quantized interval；
Wherein described step 3 is specifically included：
Step 31：Continuous N number of power frequency period is intercepted from obtained pulse-time sequence；
Step 32：Each power frequency period average quantization is interval into K, count the pulse in each interval in K interval
Number；
Step 33：The pulse number being under N number of cycle in identical quantized interval is added up, K group pulse numbers are obtained
Statistical value；
Step 4：FFT is carried out to packet pulse data, and calculating obtains amplitude-frequency values；
Step 5：According to amplitude-frequency values, the energy under different frequency is calculated；
Step 6：Judge whether occur shelf depreciation according to energy accounting, complete the identification of shelf depreciation.
2. the Recognition of Partial Discharge according to claim 1 quantified based on packet, wherein the part gathered
Discharge data is discrete voltage-time data that electrooptical device is exported, and described preprocessing process is to filter out width first
It is worth less background noise, data smoothing, all Local modulus maximas of final search is then carried out using Neighborhood Filtering.
3. the Recognition of Partial Discharge according to claim 2 quantified based on packet, wherein obtaining office in search
While portion's maximum point, the position where maximum point should be also recorded, the Local modulus maxima represents a pulse.
4. the Recognition of Partial Discharge according to claim 1 quantified based on packet, wherein described pulse
Number is characterized by physical quantity, and the physical quantity is pulse amplitude or pulse energy.
5. the Recognition of Partial Discharge according to claim 1 quantified based on packet, wherein described N number of cycle
It is not overlapping or fracture for the continuous cycle, it is selective to normalize to a work to count obtained pulse number accumulated value
The frequency cycle.
6. the Recognition of Partial Discharge according to claim 1 quantified based on packet, wherein described calculate different
Energy accounting under frequency, is that amplitude progress square is obtained into energy, then calculates the energy accounting of different frequency.
7. the Recognition of Partial Discharge according to claim 1 quantified based on packet, wherein described according to energy
Accounting judges whether occur shelf depreciation, be judge energy under 50Hz and 100Hz and whether more than other frequencies energy and,
If being then considered as generation shelf depreciation, otherwise it is considered as and does not occur shelf depreciation.
8. the Recognition of Partial Discharge according to claim 7 quantified based on packet, wherein other described frequencies
Not comprising DC component.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510309495.0A CN104865508B (en) | 2015-06-08 | 2015-06-08 | The Recognition of Partial Discharge quantified based on packet |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510309495.0A CN104865508B (en) | 2015-06-08 | 2015-06-08 | The Recognition of Partial Discharge quantified based on packet |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104865508A CN104865508A (en) | 2015-08-26 |
| CN104865508B true CN104865508B (en) | 2017-09-29 |
Family
ID=53911476
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510309495.0A Active CN104865508B (en) | 2015-06-08 | 2015-06-08 | The Recognition of Partial Discharge quantified based on packet |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104865508B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106019094A (en) * | 2016-05-18 | 2016-10-12 | 西安交通大学 | Three-dimensional analysis method and system for partial discharge under oscillating-type impact voltage |
| CN110286302B (en) * | 2019-06-26 | 2021-08-17 | 全球能源互联网欧洲研究院 | Detection method and detection system for partial discharge signal |
| CN113552456B (en) * | 2021-07-30 | 2022-08-23 | 华能随县界山风电有限责任公司 | Switch cabinet partial discharge intensity monitoring method and device and storage medium |
| CN114089139B (en) * | 2021-12-07 | 2022-10-28 | 四川大学 | Method and device for measuring turn-to-turn insulation PDIV (pulse-induced degradation) of variable frequency motor based on frequency domain energy |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819246B (en) * | 2010-04-27 | 2012-10-17 | 重庆大学 | Discharge capacity monitoring and collecting method, device and system for ultrahigh frequency local discharge |
| CN103558517A (en) * | 2013-10-29 | 2014-02-05 | 云南电网公司楚雄供电局 | Method for positioning high-voltage apparatus electro-discharge source based on energy change |
| CN103592587A (en) * | 2013-12-02 | 2014-02-19 | 国家电网公司 | Partial discharge diagnosis method based on data mining |
| CN104215887A (en) * | 2014-09-02 | 2014-12-17 | 国家电网公司 | Portable partial discharge detector |
| CN104237750A (en) * | 2014-09-05 | 2014-12-24 | 中国西电电气股份有限公司 | GIS insulation defect partial discharge fault graph drawing method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002303648A (en) * | 2001-04-05 | 2002-10-18 | Hitachi Cable Ltd | Partial discharge measuring method |
-
2015
- 2015-06-08 CN CN201510309495.0A patent/CN104865508B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819246B (en) * | 2010-04-27 | 2012-10-17 | 重庆大学 | Discharge capacity monitoring and collecting method, device and system for ultrahigh frequency local discharge |
| CN103558517A (en) * | 2013-10-29 | 2014-02-05 | 云南电网公司楚雄供电局 | Method for positioning high-voltage apparatus electro-discharge source based on energy change |
| CN103592587A (en) * | 2013-12-02 | 2014-02-19 | 国家电网公司 | Partial discharge diagnosis method based on data mining |
| CN104215887A (en) * | 2014-09-02 | 2014-12-17 | 国家电网公司 | Portable partial discharge detector |
| CN104237750A (en) * | 2014-09-05 | 2014-12-24 | 中国西电电气股份有限公司 | GIS insulation defect partial discharge fault graph drawing method |
Non-Patent Citations (2)
| Title |
|---|
| 水轮发电机局部放电在线监测中的脉冲识别方法;万元等;《高电压技术》;20090930;第35卷(第09期);第2169-2175页 * |
| 用谐波小波包变换法提取GIS局部放电信号多尺度特征参数;唐炬等;《电工技术学报》;20150228;第31卷(第03期);第250-257页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104865508A (en) | 2015-08-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110275094B (en) | High-voltage cable partial discharge acousto-optic combined detection method and system | |
| CN104865508B (en) | The Recognition of Partial Discharge quantified based on packet | |
| CN104198898B (en) | Local discharge development process diagnosis method based on pulse-train analysis | |
| CN103197218B (en) | A kind of high-voltage cable insulation defect partial discharge electrification detection diagnostic method | |
| CN108693448B (en) | Partial discharge mode recognition system applied to power equipment | |
| CN108713151A (en) | The processing unit and method of partial discharge pulse are detected in the presence of noise signal | |
| CN107015109B (en) | A kind of recognition methods of transmission line of electricity inductive lightning and direct lightning strike based on OPGW | |
| CN105548832A (en) | High-voltage power cable fault recognition method | |
| CN104849633A (en) | Switchgear partial discharge mode recognition method | |
| CN105425106B (en) | A kind of Cable fault examination pilot system and its method of work | |
| CN105044566B (en) | A kind of GIS partial discharge fault detection method of feature based ultra-high frequency signal | |
| CN107167716A (en) | A kind of shelf depreciation default kind identification method and device | |
| CN203929975U (en) | On-Line Monitor Device is put in remote control multiband bandpass filtering office | |
| CN102792173A (en) | Instrument and method for detecting partial electrical discharges | |
| CN106249114A (en) | Multifunctional belt electric detection means based on WIFI transmission and method | |
| CN107942210A (en) | The classification of transformer pulse electric current Partial Discharge and denoising method and system | |
| CN103197206A (en) | Electrical level scanning type ultrahigh frequency partial discharge on-line monitoring system and monitoring method | |
| CN110161301A (en) | A kind of anti-interference type lightning current waveform gathering algorithm | |
| CN107422232A (en) | A kind of digital live detection instrument of power distribution network terminal equipment | |
| CN102792174A (en) | Instrument and method for detecting partial electrical discharges | |
| CN109932053A (en) | A kind of state monitoring apparatus and method for high-voltage shunt reactor | |
| CN202041614U (en) | Insulation defect local discharging signal identifying device of ultrahigh voltage cable line | |
| CN114755543A (en) | Power equipment partial discharge detection method and equipment | |
| CN104991171A (en) | Method for drawing GIS partial discharge frequency division fault spectrogram based on ultrahigh frequency signal | |
| CN103558484B (en) | A kind of substation cable is by thunderbolt interference detecting apparatus and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |