CN104569684B - A kind of fault arc detection method based on arc spectrum signal - Google Patents
A kind of fault arc detection method based on arc spectrum signal Download PDFInfo
- Publication number
- CN104569684B CN104569684B CN201510018307.9A CN201510018307A CN104569684B CN 104569684 B CN104569684 B CN 104569684B CN 201510018307 A CN201510018307 A CN 201510018307A CN 104569684 B CN104569684 B CN 104569684B
- Authority
- CN
- China
- Prior art keywords
- arc
- signal
- spectral
- fault
- mrow
- 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)
Abstract
The present invention relates to a kind of fault arc detection method based on arc spectrum signal, comprise the following steps：Spectral collector gathers spectral signal with number of times according to set time includes the wavelength and intensity of light, current signal is gathered simultaneously, and spectra re-recorded signal and current signal, spectral signal is normalized, then calculate wavelet transformation detailed information and be used as characteristic value, spectral signal is integrated by different-waveband simultaneously, the characteristic value of wavelet transformation and the spectral intensity integrated value of spectral signal different-waveband judge whether fault electric arc as the input of BP neural network according to the output of BP neural network.The present invention has that applicability is wide, safe, directly react electrical contact switch service condition, can accurate failure judgement electric arc.
Description
Technical field
The present invention relates to a kind of arc-detection technology protected for actual circuit, specifically, it is related to a kind of based on electricity
The fault arc detection method of arc light spectrum signal.
Background technology
Electric utility is fast-developing, and the daily life given people brings many facilities, but also brings frequent generation
Electrical fire accident.The electrical fire method and technology of China is delayed, and traditional detection means can not effectively detect electricity
Arc failure, and when arc fault occurs, moment, localized hyperthermia can reach 2000 to 4000 degrees Celsius.If without excision electricity in time
Arc failure, can cause electrical equipment seriously to burn, fire, the serious consequence such as personal injury.Therefore, the detection to fault electric arc is
Highly important work.
China is started late in arc-detection technical research, and the theoretical research stage (number of applying for a patent is mostly at present：
201310376133.4, denomination of invention：AC fault arc method for measuring based on wavelet transformation and time domain composite character), greatly
It is most by sensed current signal, analysis current signal waveform determines whether fault electric arc, but loading condition thousand poor ten thousand
, do not break down sometimes electric arc when and current waveform difference very little during normal work (Zou Yunfeng, Wu Weilin, the brave of Li Zhi are based on
Low voltage failure electric arc clustering [J] Chinese journal of scientific instrument of self-organizing map neural network, 2010,31 (3) 571-576.),
It is difficult to determine whether fault electric arc accordingly.
There are more ripe fault electric arc detection product in foreign countries, but voltage environment is different both at home and abroad, and arc characteristic has
Change, existing procucts can not be applied directly, therefore, and it is very necessary that research, which is adapted to domestic fault electric arc detection product,.
The content of the invention
For above-mentioned defect of the prior art, a kind of fault electric arc inspection based on arc spectrum signal proposed by the present invention
Survey method.
To realize above-mentioned purpose, the technical solution adopted by the present invention is：
The present invention provides a kind of fault arc detection method based on arc spectrum signal, and this method passes through arclight first
Spectrum collector gathers spectral signal and current signal, the normalized based on power is then done to spectral signal, and carry out small
Wave conversion, will calculate the energy of obtained three first layers wavelet transformation detail signal as the characteristic value of wavelet transformation；Spectrum is believed
Number by light different colours wave-length coverage disjunction processing, to every section of spectral signal Integral Processing, spectrum of the record per segment is strong
Spend integrated value；The input value of intensity integrated value using the characteristic value of wavelet transformation and per segment spectrum as BP neural network；Root
Fault electric arc is judged whether according to the output valve of BP neural network.
The specific implementation step of the present invention is as follows：
Step 1：Set spectra collection number of times and time interval.Spectra collection equipment gathers electricity in a switching process and connect
Tactile arc spectrum signal, obtains the wavelength and strength information (λ of spectrumi, Ii), while Hall current sensor collection electric current letter
Cease Ia i；
Step 2：Judge whether spectra collection number of times reaches given threshold, if reaching, start perform step 3, otherwise after
Continuous step 1；
Step 3：Spectral signal is normalized, and carries out wavelet transformation, three first layers wavelet transformation details is believed
Number energy as wavelet transformation characteristic value；
Step 4：By wave-length coverage segment processing of the spectral signal by the different colours of light, at every section of spectral signal integration
Reason, spectral intensity integrated value of the record per segment；
Step 5：Using the characteristic value of wavelet transformation and per segment spectral intensity integrated value as BP neural network input
Value；
Step 6：Fault electric arc is judged whether according to the output valve of BP neural network.
Described carries out wavelet transformation to the spectral signal after normalization, obtains the spy of three first layers wavelet transformation details energy
Value indicative, it is specific as follows：Spectral signal X after normalization carries out wavelet transform, and obtained approximation component is filtered twice
Scale coefficient.
Described BP neural network, method for building up is as follows：
The details energy eigenvalue of wavelet transformation and the spectral intensity integrated value of different colours are used as the defeated of BP neural network
Enter value, BP neural network is set up with MATLAB softwares, BP neural network is from input layer to hidden layer and hidden layer is to output layer
Between transmission function be respectively adopted logsig and tansig functions, training function uses traingd functions.
The described output valve according to BP networks judges whether occur arc fault, is specially：The spectrum letter repeatedly obtained
Breath training BP neural network, the output valve of neutral net refers to no arc fault when being 0, when being 1 arc fault occurs for output valve.
Compared with prior art, the beneficial effects of the invention are as follows：
The present invention is a kind of brand-new method that fault electric arc is detected using spectral signal, analyzes the ripple of spectral signal
The sizes such as length, intensity can obtain the material that burning produces electric arc, so as to understand the use state of electrical contact switch.Present invention tool
Have that applicability is wide, safe, direct reaction electrical contact switch service condition, can accurate failure judgement electric arc.
Brief description of the drawings
Fig. 1 is the fault electric arc identification process figure of a preferred embodiment of the invention；
Fig. 2 is the fault electric arc identifying device schematic diagram of a preferred embodiment of the invention.
Embodiment
Technical scheme is further described below, the following description is only to understand technical solution of the present invention
It is used, is not used in restriction the scope of the present invention, protection scope of the present invention is defined by claims.
As shown in figure 1, being the method flow diagram of one embodiment of the present invention：Pass through collecting fiber arc generator first
The substantial amounts of normal operation produced and the spectral information of fault electric arc, to the spectral signal normalized collected, are carried out small
Wave conversion obtains the characteristic value of three wavelet transformation details, while to spectrum is by color segments and integrates, being used as BP neural network
Learning sample, be inputted in neutral net and network be trained, assign the neutral net trained as fault electric arc
Identifier.The signal for needing to recognize is input in the neutral net trained during operation and recognized.
As shown in Fig. 2 be the fault electric arc identifying device schematic diagram of a preferred embodiment of the invention, wherein：Use Chinese mugwort ten thousand
This spectrometer is carried, spectral detection scope is 200~800nm, one point of four optical fiber.When arc generator dynamic/static contact is connected, have
Electric current passes through, and cam drives moving contact motion, separate dynamic/static contact, produce electric arc in the presence of servomotor.Photoelectric transfer
Sensor detects arc light and is converted into voltage, is used to trigger spectrometer after amplified shaping, spectrometer receives arc spectrum signal,
And further handle the spectral signal computer being collected into using software.
According to step 1, spectra collection number of times is set 1000 times, time of integration 1ms preserves the wavelength and intensity of spectrum.
Step 2：Judge whether spectra collection number of times reaches given threshold, if reaching, start to perform step 3：, otherwise after
Continuous step 1；
Step 3：Spectral signal is normalized, it is as follows
Xi=Ii/UIa i
XiFor the spectral signal after normalization, IiFor the spectral information before normalization, U is arc voltage, Ia iFor electric arc electricity
Stream.
Wavelet transformation is carried out to the spectral signal after normalization, the characteristic value of three first layers wavelet transformation details energy is obtained,
It is specific as follows：Spectral signal X after normalization carries out wavelet transform, and the yardstick system of obtained approximation component is filtered twice
Number.
Calculate three details energy eigenvalues as follows：
d1=∑ Wψ 2(J-1, k)
d2=∑ Wψ 2(J-2, k)
I refers to spectra collection number of times, with current acquisition number of times；
Wψ、Refer to the subspace of spectrum signal function；
hψ(l)、Refer to scaling function coefficient；
J refers to yardstick, and J-1 is out to out, and smallest dimension is 0；
L refers to the dummy variables in convolution, that is, substitutes two sequences of correlation；
N=2k, refers to the sequence number of spectral signal, and k refers to the signal sequence of wavelets Subspace；
d1、d2、d3For details energy eigenvalue.
Step 4：By wave-length coverage segment processing of the spectral signal by the different colours of light, at every section of spectral signal integration
Reason, spectral intensity integrated value of the record per segment；
The described different colours wavelength that spectral signal is pressed to light carries out segment processing, and integrates, specific as follows：
SjRefer to the spectral intensity integrated value of jth kind color, b, c are the upper lower limit value of this kind of color spectrum wavelength.
Step 5：Using the characteristic value of wavelet transformation and per segment spectral intensity integrated value as BP neural network input
Value；
Described BP neural network, method for building up is as follows：
The details energy eigenvalue of wavelet transformation and the spectral intensity integrated value of different colours are used as the defeated of BP neural network
Enter value, BP neural network is set up with MATLAB softwares, BP neural network is from input layer to hidden layer and hidden layer is to output layer
Between transmission function be respectively adopted logsig and tansig functions, training function uses traingd functions.
Step 6：Fault electric arc is judged whether according to the output valve of BP neural network.
The described output valve according to BP networks judges whether occur arc fault, is specially：The spectrum letter repeatedly obtained
Breath training BP neural network, the output valve of neutral net refers to no arc fault when being 0, when being 1 arc fault occurs for output valve.
The present invention detects fault electric arc using spectral signal, and analyzing the sizes such as wavelength, the intensity of spectral signal can obtain
The material of electric arc is produced to burning, so as to understand the use state of electrical contact switch.The present invention has that applicability is wide, security
Height, directly reaction electrical contact switch service condition, can accurate failure judgement electric arc.
The section Example of the present invention is the foregoing is only, any limitation not is done to the technical scope of the present invention,
Any modification made within the spirit and principles of the invention, equivalent substitution and improvement etc. should be included in the guarantor of the present invention
Within the scope of shield.
Claims (6)
1. a kind of fault arc detection method based on arc spectrum signal, it is characterised in that comprise the following steps：
Step 1：Spectra collection number of times and time interval are set, spectra collection equipment gathers what is made electrical contact with a switching process
Arc spectrum signal, obtains the wavelength and strength information (λ of spectrumi, Ii), while Hall current sensor gathers current information
Ia i, i refers to spectra collection number of times, with current acquisition number of times；
Step 2：Judge whether spectra collection number of times reaches given threshold, if reaching, start to perform step 3, otherwise continue to walk
Rapid 1；
Step 3：Spectral signal is normalized, and carries out wavelet transformation, by three first layers wavelet transformation detail signal
Energy as wavelet transformation characteristic value；
Step 4：By wave-length coverage segment processing of the spectral signal by the different colours of light, to every section of spectral signal Integral Processing,
Spectral intensity integrated value of the record per segment；
Step 5：Using the characteristic value of wavelet transformation and per segment spectral intensity integrated value as BP neural network input value；
Step 6：Fault electric arc is judged whether according to the output valve of BP neural network.
2. the fault arc detection method according to claim 1 based on arc spectrum signal, it is characterised in that described
The spectral information collected is normalized, it is specific as follows：
Xi=Ii/UIa i
XiFor the spectral signal after normalization, IiFor the spectral information before normalization, U is arc voltage, Ia iFor arc current.
3. the fault arc detection method according to claim 1 based on arc spectrum signal, it is characterised in that the step
In rapid 3：Wavelet transformation is carried out to the spectral signal after normalization, the characteristic value of three first layers wavelet transformation details energy is obtained, had
Body is as follows：
Spectral signal X after normalization carries out wavelet transform, and the scale coefficient of obtained approximation component is filtered twice：
<mrow>
<msub>
<mi>W</mi>
<mi>&psi;</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>J</mi>
<mo>-</mo>
<mn>1</mn>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mi>l</mi>
</munder>
<msub>
<mi>h</mi>
<mi>&psi;</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mi>k</mi>
<mo>-</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
<msub>
<mo>|</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>0</mn>
<mo>,</mo>
<mn>1</mn>
<mo>...</mo>
</mrow>
</msub>
</mrow>
Calculate three details energy eigenvalues as follows：
d1=∑ Wψ 2(J-1,k)
d2=∑ Wψ 2(J-2,k)
Wψ、Refer to the subspace of spectrum signal function；
hψ(l)、Refer to scaling function coefficient；
J refers to yardstick, and J-1 is out to out, and smallest dimension is 0；
L refers to the dummy variables in convolution, that is, substitutes two sequences of correlation；
N=2k, refers to the sequence number of spectral signal, and k refers to the signal sequence of wavelets Subspace；
d1、d2、d3For details energy eigenvalue.
4. the fault arc detection method according to claim 1 based on arc spectrum signal, it is characterised in that the step
In rapid 4：Spectral signal is subjected to segment processing by the different colours wavelength of light, and integrated, it is specific as follows：
<mrow>
<msub>
<mi>S</mi>
<mi>j</mi>
</msub>
<mo>=</mo>
<msubsup>
<mi>&Sigma;</mi>
<mi>b</mi>
<mi>c</mi>
</msubsup>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
</mrow>
SjRefer to the spectral intensity integrated value of jth kind color, b, c are the upper lower limit value of this kind of color spectrum wavelength, XiFor normalization
Spectral signal afterwards.
5. the fault arc detection method based on arc spectrum signal according to claim any one of 1-4, its feature exists
In in the step 5：The method for building up of BP neural network is as follows：
The details energy eigenvalue of wavelet transformation and the spectral intensity integrated value of different colours as BP neural network input value,
BP neural network is set up with MATLAB softwares, BP neural network is from input layer to hidden layer and hidden layer is between output layer
Logsig and tansig functions are respectively adopted in transmission function, and training function uses traingd functions.
6. the fault arc detection method based on arc spectrum signal according to claim any one of 1-4, its feature exists
In in the step 6：Judge whether occur arc fault according to the output valve of BP networks, be specially：The spectrum letter repeatedly obtained
Breath training BP neural network, the output valve of neutral net refers to no arc fault when being 0, when being 1 arc fault occurs for output valve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510018307.9A CN104569684B (en) | 2015-01-14 | 2015-01-14 | A kind of fault arc detection method based on arc spectrum signal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510018307.9A CN104569684B (en) | 2015-01-14 | 2015-01-14 | A kind of fault arc detection method based on arc spectrum signal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104569684A CN104569684A (en) | 2015-04-29 |
| CN104569684B true CN104569684B (en) | 2017-08-25 |
Family
ID=53086218
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510018307.9A Active CN104569684B (en) | 2015-01-14 | 2015-01-14 | A kind of fault arc detection method based on arc spectrum signal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104569684B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106679809B (en) * | 2017-02-24 | 2018-12-21 | 三峡大学 | Ring network cabinet fault distinguishing system based on spectroscopic analysis methods |
| CN108009519B (en) * | 2017-12-19 | 2023-10-31 | 中国医学科学院生物医学工程研究所 | Light irradiation information monitoring method and device |
| CN108562835A (en) * | 2018-03-19 | 2018-09-21 | 杭州拓深科技有限公司 | A kind of fault arc detection method based on BP neural network |
| CN110456234B (en) * | 2018-05-07 | 2020-11-10 | 珠海格力电器股份有限公司 | Fault arc detection method, device and system |
| CN109270384B (en) * | 2018-11-13 | 2019-06-11 | 中南民族大学 | A kind of method and system of the electric arc of electrical equipment for identification |
| CN109615070A (en) * | 2018-12-06 | 2019-04-12 | 浙江巨磁智能技术有限公司 | Electric power artificial intelligence chip and power failure recognition methods |
| CN110308377B (en) * | 2019-07-23 | 2021-06-15 | 南京航空航天大学 | Arc detection method for multi-electric-plane direct-current system |
| CN110441647A (en) * | 2019-09-06 | 2019-11-12 | 云南电网有限责任公司电力科学研究院 | Arc light assessment of risks method and device based on spectral intensity information |
| CN113702739A (en) * | 2021-08-26 | 2021-11-26 | 广东电网有限责任公司 | Electric arc detection method and device based on wavelet decomposition and neural network |
| CN114966472B (en) * | 2022-08-01 | 2022-10-21 | 云南电力试验研究院(集团)有限公司 | Electric arc spectrum identification method and device |
| CN115598470A (en) * | 2022-09-05 | 2023-01-13 | 国网江苏省电力有限公司无锡供电分公司（Cn） | Arc active early warning method and system based on multispectral frequency band |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103163430A (en) * | 2013-03-29 | 2013-06-19 | 昆明理工大学 | Resonant grounding system fault line selection method by combining complex wavelets with ANN (artificial neural network) |
| CN103543375A (en) * | 2013-08-26 | 2014-01-29 | 上海交通大学 | Method for detecting alternating-current fault arcs on basis of wavelet transformation and time-domain hybrid features |
| CN103913663A (en) * | 2014-04-21 | 2014-07-09 | 南京航空航天大学 | Online detection method and protection device for direct current system arc faults |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7391218B2 (en) * | 2005-03-11 | 2008-06-24 | Honeywell International Inc. | Method and apparatus for generalized arc fault detection |
| US9025287B2 (en) * | 2012-12-19 | 2015-05-05 | Stmicroelectronics S.R.L. | Arc fault detection equipment and method using low frequency harmonic current analysis |
-
2015
- 2015-01-14 CN CN201510018307.9A patent/CN104569684B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103163430A (en) * | 2013-03-29 | 2013-06-19 | 昆明理工大学 | Resonant grounding system fault line selection method by combining complex wavelets with ANN (artificial neural network) |
| CN103543375A (en) * | 2013-08-26 | 2014-01-29 | 上海交通大学 | Method for detecting alternating-current fault arcs on basis of wavelet transformation and time-domain hybrid features |
| CN103913663A (en) * | 2014-04-21 | 2014-07-09 | 南京航空航天大学 | Online detection method and protection device for direct current system arc faults |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104569684A (en) | 2015-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104569684B (en) | A kind of fault arc detection method based on arc spectrum signal | |
| CN104614608B (en) | A kind of low pressure parallel arc fault detection means and method | |
| CN103389430B (en) | A kind of oil-filled transformer fault detection method based on Bayesian discrimination theory | |
| CN103267937B (en) | Method for detecting electrical aging of silicone rubber composite insulator | |
| CN1834673A (en) | Insulating state on-line monitoring method of cross-linked PE cable | |
| CN102749558A (en) | Device and method for detecting cable oscillatory wave partial discharge and fault location | |
| CN108693448B (en) | Partial discharge mode recognition system applied to power equipment | |
| CN108196164B (en) | Method for extracting cable fault point discharge sound signal under strong background noise | |
| CN105403816A (en) | Identification method of DC fault electric arc of photovoltaic system | |
| CN107462560A (en) | LIF combinations LIBS edible oil quality fast analyser and method | |
| CN103512922A (en) | Electrical fire detection system and method based on electronic nose system | |
| CN106771798A (en) | A kind of fault arc detection method based on the equal difference of wavelet coefficient | |
| CN105423908A (en) | Transformer winding deformation live test method and system | |
| CN202033865U (en) | Ultraviolet and infrared composite flame detector | |
| CN114325256A (en) | Power equipment partial discharge identification method, system, equipment and storage medium | |
| CN112949497B (en) | GIS partial discharge pattern recognition method based on improved generalized regression neural network | |
| CN107703377A (en) | Fault arc detection method and device | |
| CN106100776A (en) | Frequency spectrum sensing method based on wireless station grid monitoring system | |
| CN105277510A (en) | Propiconazole discriminating method based on Terahertz theory for simulation of spectrum database | |
| CN113253063A (en) | Fault arc detection system and method based on long-time memory network deep learning | |
| Qian et al. | Research on DC arc fault detection in PV systems based on adjacent multi-segment spectral similarity and adaptive threshold model | |
| CN102937694A (en) | Device for monitoring external insulation strength of dirty insulator | |
| CN113588568B (en) | Method for detecting environment-friendly insulating gas decomposition product | |
| CN104483288B (en) | Perfluoroisopropyl hexanone extinguishing chemical recognition detection method | |
| CN103344893A (en) | Distributed cable partial discharge measuring method based on frequency conversion series resonance high-voltage holding test |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: SHANGHAI HEWU COMPOSITE MATERIAL CO., LTD. Free format text: FORMER OWNER: SHANGHAI HEWU NEW MATERIAL TECHNOLOGY CO., LTD. Effective date: 20150706 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20150706 Address after: 200240 Shanghai city Minhang District Jianchuan road 953 Lane 322 building Applicant after: Shanghai and 5 composite material Co., Ltds Address before: No. 955 Cangyuan science and Technology Park, 200240 Shanghai city Minhang District Jianchuan Road Building 9 floor, I enjoy my home Applicant before: Shanghai Hiwave Advanced Materials Technology Co., Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |