CN108828377A - A kind of Diagnosis Method of Transformer Faults - Google Patents
A kind of Diagnosis Method of Transformer Faults Download PDFInfo
- Publication number
- CN108828377A CN108828377A CN201811007817.6A CN201811007817A CN108828377A CN 108828377 A CN108828377 A CN 108828377A CN 201811007817 A CN201811007817 A CN 201811007817A CN 108828377 A CN108828377 A CN 108828377A
- Authority
- CN
- China
- Prior art keywords
- transformer
- fault
- content
- gas
- ethylene
- 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
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000003745 diagnosis Methods 0.000 title claims abstract description 30
- 238000013528 artificial neural network Methods 0.000 claims abstract description 22
- 241001269238 Data Species 0.000 claims abstract description 5
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 claims description 79
- 239000007789 gas Substances 0.000 claims description 71
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 70
- VGGSQFUCUMXWEO-UHFFFAOYSA-N Ethene Chemical compound C=C VGGSQFUCUMXWEO-UHFFFAOYSA-N 0.000 claims description 44
- 239000005977 Ethylene Substances 0.000 claims description 44
- 229910002092 carbon dioxide Inorganic materials 0.000 claims description 40
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 38
- 239000001569 carbon dioxide Substances 0.000 claims description 38
- 229910002091 carbon monoxide Inorganic materials 0.000 claims description 38
- HSFWRNGVRCDJHI-UHFFFAOYSA-N alpha-acetylene Natural products C#C HSFWRNGVRCDJHI-UHFFFAOYSA-N 0.000 claims description 31
- 125000002534 ethynyl group Chemical group [H]C#C* 0.000 claims description 31
- 229910052739 hydrogen Inorganic materials 0.000 claims description 30
- 239000001257 hydrogen Substances 0.000 claims description 30
- 238000009413 insulation Methods 0.000 claims description 29
- 238000013021 overheating Methods 0.000 claims description 28
- 239000007787 solid Substances 0.000 claims description 28
- OTMSDBZUPAUEDD-UHFFFAOYSA-N Ethane Chemical compound CC OTMSDBZUPAUEDD-UHFFFAOYSA-N 0.000 claims description 26
- 150000002431 hydrogen Chemical class 0.000 claims description 24
- 230000001537 neural effect Effects 0.000 claims description 18
- 239000004215 Carbon black (E152) Substances 0.000 claims description 15
- 229930195733 hydrocarbon Natural products 0.000 claims description 15
- 150000002430 hydrocarbons Chemical class 0.000 claims description 15
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000001186 cumulative effect Effects 0.000 claims description 9
- 101100269885 Arabidopsis thaliana ANN1 gene Proteins 0.000 claims description 7
- 101100269886 Arabidopsis thaliana ANN2 gene Proteins 0.000 claims description 7
- 101100269887 Arabidopsis thaliana ANN3 gene Proteins 0.000 claims description 7
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 claims description 6
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 5
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 4
- 230000004888 barrier function Effects 0.000 claims description 4
- 229910052799 carbon Inorganic materials 0.000 claims description 4
- 238000003672 processing method Methods 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims 3
- 241000790917 Dioxys <bee> Species 0.000 claims 1
- GCNLQHANGFOQKY-UHFFFAOYSA-N [C+4].[O-2].[O-2].[Ti+4] Chemical compound [C+4].[O-2].[O-2].[Ti+4] GCNLQHANGFOQKY-UHFFFAOYSA-N 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 5
- 238000012423 maintenance Methods 0.000 abstract description 4
- 230000002159 abnormal effect Effects 0.000 description 9
- 208000024891 symptom Diseases 0.000 description 8
- 238000012360 testing method Methods 0.000 description 6
- 238000004804 winding Methods 0.000 description 5
- 238000013459 approach Methods 0.000 description 4
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical group [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 206010020843 Hyperthermia Diseases 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 150000001335 aliphatic alkanes Chemical class 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 229910002090 carbon oxide Inorganic materials 0.000 description 1
- 238000011208 chromatographic data Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000002405 diagnostic procedure Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000036031 hyperthermia Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003647 oxidation Effects 0.000 description 1
- 238000007254 oxidation reaction Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Primary Health Care (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Housings And Mounting Of Transformers (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
Abstract
The present invention provides a kind of Diagnosis Method of Transformer Faults, transformer fault diagnosis is carried out by the way of the combination of three-ratio method, BP neural network and reasoning of citing a precedent, substantially increase the accuracy of transformer fault judgement, it reduces maintenance cost, reduce maintenance risk, each professional preferably can be subjected to tissue, all kinds of status datas of equipment are effectively utilized, more accurately to determine equipment state and providing aid decision suggestion to provide technical support, also good technical foundation is established in the life-cycle management for realization power grid asset.For grid equipment is safe and stable, reliable, long period, high-quality operation provide reliable technology and management safeguard.
Description
Technical field
The present invention relates to transformer fault diagnosis technical field more particularly to a kind of Diagnosis Method of Transformer Faults.
Background technique
Transformer is the important equipment of electric system, and operating status directly affects the level of security of electric system.However
Since transformer is the closing entirety for combining the multidisciplinary technologies such as mechanical, electrical, chemistry, thermodynamics, its failure is influenced
The reason of it is intricate, fault diagnosis needs various data and knowledge, only according in a certain respect or several respects just subjectively under
Conclusion necessarily causes to judge by accident or fail to judge.At present common method for diagnosing faults be three ratio in judgement methods, overheat electric discharge figure determining method,
HAE axonometric projection determining method, characteristic gas method etc., every kind of single diagnostic method, although to certain specific fault recognition rates compared with
Height, but there is also disadvantages.By taking ratio method as an example, the right judging rate that it belongs to overheat or electric discharge type to failure is higher, general energy
Reach 90% or more.But for overheating type fault, belong to magnetic conductive loop or galvanic circle mistake further determining bottom
When hot, accuracy is just substantially reduced, only rely on certain single method for diagnosing faults determine failure be it is less reliable,
It is unable to satisfy engineering demand.And artificial neural network as it is a kind of in form close to human brain construct novel information processing system,
With powerful parallel processing capability, distributed storage ability and adaptive learning ability, thus dissolved gas content in the oil
It combines artificial neural network technology to carry out Insulation Fault on the basis of analysis result and has become research in recent years
Hot spot, and make great progress, therefore, in order to ensure safe and reliable, the flexible coordination of Hainan Power Grid operation, high-quality height
It imitates, is economic and environment-friendly, there is an urgent need for a kind of methods of energy accurate judgement transformer fault.
Summary of the invention
The purpose of the present invention is to provide a kind of Diagnosis Method of Transformer Faults, mentioned above in the background art to solve
Problem.
The present invention is achieved by the following technical solutions：A kind of Diagnosis Method of Transformer Faults, using three-ratio method, BP
The mode of the combination of neural network and reasoning of citing a precedent carries out transformer fault diagnosis, and step is：
Whether S1, the content for detecting hydrogen, acetylene and total appropriate hydrocarbon gas ∑ C in transformer respectively are exceeded, and detect total hydrocarbon
Whether gas production rate is exceeded；
S2, according to gas content result detected and total hydrocarbon gas production rate, judge whether transformer generates failure, when
When transformer does not break down, verified using the method based on rough set；
S3, when transformer breaks down, use three is than value-based algorithm, BP neural algorithm and rationalistic method of citing a precedent to transformer
Fault type carries out preliminary analysis, determines that transformer fault type is discharge fault or overheating fault；
S4, according to the Preliminary Analysis Results, continue using three than value-based algorithm, BP neural algorithm and rationalistic method pair of citing a precedent
The fault type of transformer is further analyzed, and determines the discharge fault type or overheating fault type of transformer；
S5, the fault diagnosis result according to transformer, are verified using the method based on rough set.
Preferably, in step S2, when judging to have in aforementioned four condition more than one exceeded, that is, judge transformer
Failure.
Preferably, in step S3, the transformer discharge failure includes being related to the discharge fault of solid insulation and not relating to
And the discharge fault of solid insulation, the overheat fault of transformer include overheat of conducting circuit failure and magnetic conductive loop overheat event
Barrier, the method for carrying out preliminary analysis to transformer fault type are：
S31, the content for obtaining hydrogen, acetylene, ethylene, methane, ethane gas in transformer respectively, and adopted according to the data
The fault type of transformer is tentatively judged with three ratio in judgement methods；
S32, according to the gas content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer
Data tentatively judge the fault type of transformer using BP neural algorithm；
S33, according to the content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer
Data tentatively judge the fault type of transformer using rationalistic method of citing a precedent；
S34, in summary three kinds of judging results, when there are two or more than two judging results be same fault type
When, then tentatively judge transformer for the fault type.
Preferably, in step S4, the method that the fault type of transformer is further analyzed is：
S41, according to hydrogen in transformer, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas and total
The content data of appropriate hydrocarbon gas ∑ C, the fault type of transformer is judged using three-ratio method；
S42, according to the gas content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer
Data judge the fault type of transformer using BP neural algorithm；
S43, according to the content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer
Data judge the fault type of transformer using rationalistic method of citing a precedent；
S44, in summary three kinds of judging results, when there are two or more than two judging results be same fault type
When, then judge transformer for the fault type.
Preferably, in step S4, using following expression formula：Ethylene contents/∑ C, (methane content+ethylene contents)/∑ C,
By above three ratio in judgement transformer overheat of conducting circuit event occurs for (carbon monoxide content+carbon dioxide content)/∑ C
Barrier or magnetic conductive loop overheating fault, when meeting following expression formula：0.42<=ethylene contents/carbonaceous gas content<=0.68,
And meet (methane content+ethylene contents)/carbonaceous gas content simultaneously>=0.86, and (methane content+ethylene contents)/carbon containing
Gas content<It when 30, can determine that magnetic conductive loop failure occurs for transformer, otherwise determine that thermally conductive circuit overheat event occurs for transformer
Barrier.
Preferably, in step S4, for seal transformer, when carbon dioxide content/carbon monoxide content<When=2, sentence
Determine transformer and be related to the discharge fault of solid insulation, otherwise not to be related to the discharge fault of solid insulation, for open
Transformer, work as carbon monoxide content>400 and carbon dioxide content/carbon monoxide content<=3 or carbon dioxide content>4500
And carbon dioxide content/carbon monoxide content<When=3, then determine that transformer is related to the discharge fault of solid insulation, otherwise
Not to be related to the discharge fault of solid insulation.
Preferably, in step S3 and S4, using three independent list hidden layer BP neural network ANN1, ANN2 and ANN3 to change
Depressor failure is judged, determines that discharge fault or overheating fault occur for transformer by ANN1 neural network, using ANN2 nerve
Network determines that the type of overheating fault occurs for transformer, determines that the class of discharge fault occurs for transformer using ANN3 neural network
Type.
Preferably, in step S3 and S4, when carrying out fault diagnosis using BP neural algorithm, need to hydrogen collected,
Acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas are normalized, and processing method is：
S311, to seven kinds of characteristic gas (hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide), preferentially
A kind of characteristic gas is selected, all content datas of the gas are ranked up by size, then by the gas content after sequence
Data are divided into several groups according to size；
S312, statistics fall into the frequency Δ r in each groupjWith frequency wj, wherein wj=Δ rj/ n, n are collected feature gas
Volume data number；
S313, the cumulative frequency value F for calculating i-th group of dataj：
In formula, rjFor the cumulative frequency terminated to i-th group, the calculated F of institutejInput as BP neural algorithm；
S314, it repeats the above steps, obtains the cumulative frequency value of other characteristic gas.
Preferably, in step S5, when judging the fault type of transformer using rationalistic method of citing a precedent, electric discharge and overheat need to be established
Fault signature collection, overheat of conducting circuit failure and magnetic conductive loop overheating fault feature set are related to solid insulation discharge fault feature
Collection.
Compared with prior art, what the present invention reached has the beneficial effect that：
A kind of Diagnosis Method of Transformer Faults provided by the invention, using the failure classes of a variety of method comprehensive descision transformers
Type substantially increases the accuracy of transformer fault judgement, reduces maintenance cost, reduces maintenance risk, can preferably will be each
Professional carries out tissue, and all kinds of status datas of equipment are effectively utilized, more accurately to determine equipment state and mentioning
Technical support is provided for aid decision suggestion, also good technical foundation is established in the life-cycle management for realization power grid asset.
For grid equipment is safe and stable, reliable, long period, high-quality operation provide reliable technology and management safeguard.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, the accompanying drawings in the following description is only the preferred embodiment of the present invention, for
For those of ordinary skill in the art, without any creative labor, it can also be obtained according to these attached drawings
His attached drawing.
Fig. 1 is a kind of flow chart of Diagnosis Method of Transformer Faults provided by the invention；
Fig. 2 is three-ratio method judgment rule provided by the embodiment of the present invention；
Fig. 3 is BP neural network topological structure provided by the embodiment of the present invention.
Specific embodiment
In order to be best understood from the technology of the present invention content, be provided below specific embodiment, and in conjunction with attached drawing to the present invention do into
The explanation of one step.
Referring to Fig. 1, a kind of Diagnosis Method of Transformer Faults, by the way of the combination of BP neural network and reasoning of citing a precedent into
Line transformer fault diagnosis, step are：
Whether S1, the content for detecting hydrogen, acetylene and total appropriate hydrocarbon gas ∑ C in transformer respectively are exceeded, and detect total hydrocarbon
Whether gas production rate is exceeded；
S2, according to gas content result detected and total hydrocarbon gas production rate, judge whether transformer generates failure, when
When transformer does not break down, verified using the method based on rough set；
S3, when transformer breaks down, use three is than value-based algorithm, BP neural algorithm and rationalistic method of citing a precedent to transformer
Fault type carries out preliminary analysis, determines that discharge fault type is discharge fault or overheating fault；
S4, according to the Preliminary Analysis Results, continue using three than value-based algorithm, BP neural algorithm and rationalistic method pair of citing a precedent
The fault type of transformer is further analyzed, and determines the discharge fault type or overheating fault type of transformer；
S5, the fault diagnosis result according to transformer, are verified using the method based on rough set.
Specifically, collecting containing for hydrogen in transformer, acetylene and total appropriate hydrocarbon gas ∑ C first when carrying out breakdown judge
Amount and total hydrocarbon gas production rate, when aforementioned four condition have one it is exceeded when, then can determine whether transformer generate failure.
When transformer determine generate failure when, it is first determined the fault type of transformer be transformer discharge failure still
Overheat fault of transformer, secondly, determining that discharge fault is specially to be related to the discharge fault of solid insulation and not to be related to solid exhausted
The discharge fault of edge；Determine that overheating fault specially includes overheat of conducting circuit failure and magnetic conductive loop overheating fault, and to change
Depressor fault type carries out preliminary analysis：
S31, the content for obtaining hydrogen, acetylene, ethylene, methane, ethane gas in transformer respectively, and adopted according to the data
The fault type of transformer is tentatively judged with three ratio in judgement methods, and transformer is specifically carried out using judgment rule as shown in Figure 2
Fault type judgement；
S32, according to the gas content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer
Data tentatively judge the fault type of transformer using BP neural algorithm；
S33, according to the content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer
Data tentatively judge the fault type of transformer using rationalistic method of citing a precedent；
S34, in summary three kinds of judging results, when there are two or more than two judging results be same fault type
When, then tentatively judge transformer for the fault type.
Specifically, the method that the fault type of transformer is further analyzed is：
S41, according to hydrogen in transformer, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas and total
The content data of appropriate hydrocarbon gas ∑ C, the fault type of transformer is judged using three-ratio method；
S42, according to the gas content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer
Data judge the fault type of transformer using BP neural algorithm；
S43, according to the content of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer
Data judge the fault type of transformer using rationalistic method of citing a precedent；
S44, in summary three kinds of judging results, when there are two or more than two judging results be same fault type
When, then judge transformer for the fault type.
Specifically, when judging whether transformer occurs overheating fault, since in magnetic-circuit fault, iron core failure accounts for big absolutely
It is most.Usually in iron core failure, ethylene and methane account for larger specific gravity, rule of thumb the summary of case, and ethylene accounts for the ratio of total hydrocarbon
Rate is about 45%~68%, while the content of acetylene is often lower or does not change；Further, since overheat of conducting circuit often may be used
Insulating paper can be caused to decomposite carbon monoxide, carbon dioxide, and magnetic conductive loop is mostly bare metal overheat, is not easy to cause an oxidation
Carbon, carbon dioxide content are substantially change, therefore, for using ethylene contents/∑ C, (methane content+ethylene contents)/∑ C, (one
Content of carbon oxide+carbon dioxide content)/∑ C above three ratio as judge transformer generation overheat of conducting circuit failure also
It is the foundation of magnetic conductive loop overheating fault, when meeting following expression formula：0.42<=ethylene contents/carbonaceous gas content<=
0.68, and meet (methane content+ethylene contents)/carbonaceous gas content simultaneously>=0.86, and (methane content+ethylene contents)/
Carbonaceous gas content<It when 30, can determine that magnetic conductive loop failure occurs for transformer, otherwise determine that thermally conductive circuit overheat occurs for transformer
Failure.It tests using 213 magnetic circuit overheats and 175 circuit overheat chromatographic datas as sample, after inspection result is verified
As shown in Table 1：
1 overheating fault verification result of table
| Fault type | Sample number | Correctly judge quantity | Right judging rate |
| Magnetic circuit overheated type failure | 213 | 130 | 61% |
| Circuit overheated type failure | 175 | 107 | 61% |
Specifically, when judging that the type of discharge fault occurs for transformer judgement, it is absolute based on carbon monoxide, carbon dioxide
Value content and its ratio judge whether failure has been directed to solid insulation.For seal transformer, when carbon dioxide content/
Carbon monoxide content<When=2, determine that transformer is related to the discharge fault of solid insulation, otherwise not to be related to solid insulation
Discharge fault carbon monoxide content is worked as open type transformer>400 or carbon dioxide content>4500 and carbon dioxide contain
Amount/carbon monoxide content<When=3, then determine that transformer is related to the discharge fault of solid insulation, otherwise not to be related to solid
The discharge fault of insulation.It is not related to solid insulation and 84 are related to Solid Insulator Breakdown of Electric case and test as sample using 191,
Its inspection result is as shown in table 2.
2 discharge fault verification result of table
| Fault type | Sample number | Correctly judge quantity | Right judging rate |
| It is related to solid insulation type failure | 191 | 142 | 74% |
| It is not related to solid insulation type failure | 84 | 48 | 60% |
Specifically, when carrying out transformer fault judgement using BP neural algorithm, using the error back propagation of single hidden layer
Feed-forward type BP neural network, topological structure are as shown in Figure 3.The excitation function of each neuron node usually chooses S type function.
Secondly, needing that transformer fault data are normalized, normalized is not only for extraction failure
Feature is also beneficial to the search to weight space.Therefore need to by seven kinds of characteristic gas hydrogen, acetylene, ethylene, methane, ethane,
Carbon monoxide, carbon dioxide gas content data be normalized, specific step is as follows：
S311, to seven kinds of characteristic gas (hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide), preferentially
A kind of characteristic gas is selected, all content datas of the gas are ranked up by size, then by the gas content after sequence
Data are divided into several groups according to size；
S312, statistics fall into the frequency Δ r in each groupjWith frequency wj, wherein wj=Δ rj/ n, n are collected feature gas
Volume data number；
S313, the cumulative frequency value F for calculating i-th group of dataj：
In formula, rjFor the cumulative frequency terminated to i-th group, the calculated F of institutejInput as BP neural algorithm；
S314, it repeats the above steps, obtains the cumulative frequency value of other characteristic gas.
According to above-mentioned normalization data, three independent single hidden layer BP neural network ANN1, ANN2 and ANN3 couple can be built
Transformer fault is judged, determines that discharge fault or overheating fault occur for transformer by ANN1 neural network, using ANN2 mind
Determine that the type of overheating fault occurs for transformer through network, determines that the class of discharge fault occurs for transformer using ANN3 neural network
Type, BP neural network ANN1, ANN2 and the ANN3 built be all made of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide,
The output layer node number of input quantity of the normalized value of carbon dioxide gas content as BP neural network, three networks is final
Be determined as two, and its hidden node number is identified as 15,10,10, three constructed neural network specific structures with
Parameter is as shown in table 3：
The parameter and structure of table 3BP neural network
Specifically, when judging the fault type of transformer using rationalistic method of citing a precedent, using the retrieval based on fuzzy mathematics
Algorithm finds search space where current sample.Using transformer fault as domain, corresponding sub- failure classes are as each fuzzy
Subset, the accident defect collection established are combined into electric discharge and overheating fault feature set, overheat of conducting circuit failure and magnetic conductive loop mistake
Thermal fault feature set is related to solid insulation discharge fault feature set.Compared test repeatedly, this system is selected for different defects
Corresponding retrieval character, it is specific as shown in table 4：
4 characteristic key of table is selected
After the search space and the retrieval character that determine follow-up sample, sought in this search space according to certain search method
The source example most like with follow-up sample is looked for, using the concept of approach degree in fuzzy mathematics, finds and calculates similarity.It first calculates and works as
The approach degree of each sample, selects ratio when calculating the index of approach degree in preceding sample and this search space:Acetylene/ethylene, first
Then alkane/hydrogen, ethylene/ethane, carbon dioxide/carbon monoxide select maximum three samples of approach degree according to Similarity Principle
Solution be current solution.This solution is the fail result of transformer.
According in summary three kinds of judging results, when there are two or more than two judging results be same fault type
When, then judge that transformer for overheating fault or discharge fault, and determines overheating fault type and discharge fault type.
After obtaining above-mentioned transformer diagnosis result, the method for continuing to be taken based on rough set carries out result verification, with event
Barrier sign is conditional attribute and using failure as decision attribute, for statistical analysis to the Power Transformer Faults data collected,
And with reference to more successful failure modes collection in experience before, conditional attribute set and decision attribute set such as table 5,6 institute of table are obtained
Show：
5 conditional attribute set of table
| Number | Failure symptom (conditional attribute) | Number | Failure symptom (conditional attribute) |
| C1 | Light gas movement | C14 | Oily micro-water content increases |
| C2 | Grave gas movement | C15 | Infrared detection is abnormal |
| C3 | Differential protection movement | C16 | Winding temperature increases |
| C4 | Back-up protection movement | C17 | Winding deformation test abnormality |
| C5 | The insulation resistance of winding declines | C18 | Insulating oil dielectric loss angle tangent is abnormal |
| C6 | The D.C. resistance of winding is abnormal | C19 | CO2, CO content or CO2/ CO is abnormal |
| C7 | Winding no-load voltage ratio is abnormal | C20 | Factor of created gase increases |
| C8 | DC leakage-current is exceeded | C21 | DGA shows low, medium temperature overheat |
| C9 | Iron core grounding current is exceeded | C22 | DGA shows hyperthermia and superheating |
| C10 | The decline of core insulation resistance | C23 | DGA shows low energy electric discharge |
| C11 | Short-circuit impedance is abnormal | C24 | DGA shows high-energy discharge |
| C12 | No-load test is abnormal | C25 | Air content is abnormal |
| C13 | PD test abnormality | C26 | Oil flow electrification degree is abnormal |
6 decision attribute set of table
According to historical experience, and then the fundamental relation of transformer fault and failure symptom is obtained, it is specific such as table 7, table 8, table
9, shown in table 10：
Table 7 is related to the failure symptom table of solid insulation electric discharge
The failure symptom that table 8 is not related to solid insulation electric discharge corresponds to table
The failure symptom of 9 circuit of table overheat corresponds to table
The failure symptom of 10 magnetic circuit of table overheat corresponds to table
According to the fundamental relation of above-mentioned transformer fault and failure symptom, transformer fault judging result is verified,
Finally obtain accurate transformer fault type.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the present invention.
Claims (9)
1. a kind of Diagnosis Method of Transformer Faults, which is characterized in that using three-ratio method, the knot of BP neural network and reasoning of citing a precedent
The mode of conjunction carries out transformer fault diagnosis, and step is：
Whether S1, the content for detecting hydrogen, acetylene and total appropriate hydrocarbon gas ∑ C in transformer respectively are exceeded, and detect total hydrocarbon and produce gas
Whether rate is exceeded；
S2, according to gas content result detected and total hydrocarbon gas production rate, judge whether transformer generates failure, work as transformation
When device does not break down, verified using the method based on rough set；
S3, when transformer breaks down, use three is than value-based algorithm, BP neural algorithm and rationalistic method of citing a precedent to transformer fault
Type carries out preliminary analysis, determines that transformer fault type is discharge fault or overheating fault；
S4, according to the Preliminary Analysis Results, continue using three than value-based algorithm, BP neural algorithm and rationalistic method of citing a precedent to transformation
The fault type of device is further analyzed, and determines the discharge fault type or overheating fault type of transformer；
S5, the fault diagnosis result according to transformer, are verified using the method based on rough set.
2. a kind of Diagnosis Method of Transformer Faults according to claim 1, which is characterized in that in step S2, when in judgement
State have in four conditions more than one it is exceeded when, that is, judge that transformer breaks down.
3. a kind of Diagnosis Method of Transformer Faults according to claim 1, which is characterized in that in step S3, the transformation
Device discharge fault includes being related to the discharge fault of solid insulation and not being related to the discharge fault of solid insulation, the transformer mistake
Thermal fault includes overheat of conducting circuit failure and magnetic conductive loop overheating fault, carries out preliminary analysis to transformer fault type
Method is：
S31, the content for obtaining hydrogen, acetylene, ethylene, methane, ethane gas in transformer respectively, and three are used according to the data
Ratio in judgement method tentatively judges the fault type of transformer；
S32, according to the gas content data of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer,
The fault type of transformer is tentatively judged using BP neural algorithm；
S33, according to the content data of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer,
The fault type of transformer is tentatively judged using rationalistic method of citing a precedent；
S34, in summary three kinds of judging results, when there are two or more than two judging results be same fault type when, then
Tentatively judge transformer for the fault type.
4. a kind of Diagnosis Method of Transformer Faults according to claim 3, which is characterized in that in step S4, to transformer
The method that is further analyzed of fault type be：
S41, according to hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas and total hydrocarbon gas in transformer
The content data of body ∑ C, the fault type of transformer is judged using three-ratio method；
S42, according to the gas content data of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide in transformer,
The fault type of transformer is judged using BP neural algorithm；
S43, according to the content data of hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide gas in transformer,
The fault type of transformer is judged using rationalistic method of citing a precedent；
S44, in summary three kinds of judging results, when there are two or more than two judging results be same fault type when, then
Judge transformer for the fault type.
5. a kind of Diagnosis Method of Transformer Faults according to claim 4, which is characterized in that in step S4, use is following
Expression formula：Ethylene contents/∑ C, (methane content+ethylene contents)/∑ C, (carbon monoxide content+carbon dioxide content)/∑ C,
Overheat of conducting circuit failure or magnetic conductive loop overheating fault occurs by above three ratio in judgement transformer, it is following when meeting
Expression formula：0.42<=ethylene contents/carbonaceous gas content<=0.68, and meet (methane content+ethylene contents)/carbon containing simultaneously
Gas content>=0.86, and (methane content+ethylene contents)/carbonaceous gas content<When 30, it can determine that magnetic conduction occurs for transformer
Otherwise loop fault determines that thermally conductive circuit overheating fault occurs for transformer.
6. a kind of Diagnosis Method of Transformer Faults according to claim 5, which is characterized in that in step S4, for closing
Formula transformer, when carbon dioxide content/carbon monoxide content<When=2, determine that transformer is related to the electric discharge event of solid insulation
Barrier, otherwise not to be related to the discharge fault of solid insulation, for open type transformer, works as carbon monoxide content>400 and titanium dioxide
Carbon content/carbon monoxide content<=3 or carbon dioxide content>4500 and carbon dioxide content/carbon monoxide content<When=3,
Then determine that transformer is related to the discharge fault of solid insulation, otherwise not to be related to the discharge fault of solid insulation.
7. a kind of Diagnosis Method of Transformer Faults according to claim 3, which is characterized in that in step S3 and S4, use
Three independent list hidden layer BP neural network ANN1, ANN2 and ANN3 judge transformer fault, by ANN1 neural network
Determine that discharge fault or overheating fault occur for transformer, determines that the class of overheating fault occurs for transformer using ANN2 neural network
Type determines that the type of discharge fault occurs for transformer using ANN3 neural network.
8. a kind of Diagnosis Method of Transformer Faults according to claim 7, which is characterized in that in step S3 and S4, adopting
It, need to be to hydrogen collected, acetylene, ethylene, methane, ethane, carbon monoxide, dioxy when carrying out fault diagnosis with BP neural algorithm
Change carbon gas to be normalized, processing method is：
S311, to seven kinds of characteristic gas (hydrogen, acetylene, ethylene, methane, ethane, carbon monoxide, carbon dioxide), it is preferential selected
All content datas of the gas are ranked up by a kind of characteristic gas by size, then by the gas content data after sequence
It is divided into several groups according to size；
S312, statistics fall into the frequency Δ r in each groupjWith frequency wj, wherein wj=Δ rj/ n, n are collected characteristic gas number
According to number；
S313, the cumulative frequency value F for calculating i-th group of dataj：
In formula, rjFor the cumulative frequency terminated to i-th group, the calculated F of institutejInput as BP neural algorithm；
S314, it repeats the above steps, obtains the cumulative frequency value of other characteristic gas.
9. a kind of Diagnosis Method of Transformer Faults according to claim 1, which is characterized in that in step S5, using citing a precedent
When rationalistic method judges the fault type of transformer, electric discharge and overheating fault feature set, overheat of conducting circuit failure need to be established and led
Magnetic loop overheating fault feature set is related to solid insulation discharge fault feature set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811007817.6A CN108828377A (en) | 2018-08-31 | 2018-08-31 | A kind of Diagnosis Method of Transformer Faults |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811007817.6A CN108828377A (en) | 2018-08-31 | 2018-08-31 | A kind of Diagnosis Method of Transformer Faults |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108828377A true CN108828377A (en) | 2018-11-16 |
Family
ID=64151857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811007817.6A Pending CN108828377A (en) | 2018-08-31 | 2018-08-31 | A kind of Diagnosis Method of Transformer Faults |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108828377A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111060221A (en) * | 2019-12-31 | 2020-04-24 | 云领电气智能科技(苏州)有限公司 | Transformer overheating fault early warning method based on cyclic neural network |
| CN114325495A (en) * | 2021-12-20 | 2022-04-12 | 山东汇能电气有限公司 | Operation protection method for distribution transformer based on loss comparison |
| CN114924209A (en) * | 2022-04-18 | 2022-08-19 | 云南电网有限责任公司电力科学研究院 | Transformer winding deformation monitoring system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103513125A (en) * | 2012-06-29 | 2014-01-15 | 上海翔骋电气设备有限公司 | Integrated intelligent diagnosis system and method of above-220KV transformers |
| CN104730378A (en) * | 2015-02-13 | 2015-06-24 | 国家电网公司 | Internal transformer composite-defect fuzzy diagnostic method based on gas dissolved in oil |
| CN105259435A (en) * | 2015-09-23 | 2016-01-20 | 国网山东莒县供电公司 | Transformer monitoring device and fault diagnosis method |
-
2018
- 2018-08-31 CN CN201811007817.6A patent/CN108828377A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103513125A (en) * | 2012-06-29 | 2014-01-15 | 上海翔骋电气设备有限公司 | Integrated intelligent diagnosis system and method of above-220KV transformers |
| CN104730378A (en) * | 2015-02-13 | 2015-06-24 | 国家电网公司 | Internal transformer composite-defect fuzzy diagnostic method based on gas dissolved in oil |
| CN105259435A (en) * | 2015-09-23 | 2016-01-20 | 国网山东莒县供电公司 | Transformer monitoring device and fault diagnosis method |
Non-Patent Citations (2)
| Title |
|---|
| 张炜: "基于状态评价及风险评估的变电设备状态检修方法研究", 《中国优秀硕士学位论文全文数据库（电子期刊）》 * |
| 武国亮等: "山西电网输变电状态监测中心的设计实现", 《山西电网》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111060221A (en) * | 2019-12-31 | 2020-04-24 | 云领电气智能科技(苏州)有限公司 | Transformer overheating fault early warning method based on cyclic neural network |
| CN114325495A (en) * | 2021-12-20 | 2022-04-12 | 山东汇能电气有限公司 | Operation protection method for distribution transformer based on loss comparison |
| CN114325495B (en) * | 2021-12-20 | 2022-09-13 | 山东汇能电气有限公司 | Operation protection method for distribution transformer based on loss comparison |
| CN114924209A (en) * | 2022-04-18 | 2022-08-19 | 云南电网有限责任公司电力科学研究院 | Transformer winding deformation monitoring system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Optimal dissolved gas ratios selected by genetic algorithm for power transformer fault diagnosis based on support vector machine | |
| Ghoneim et al. | A new approach of DGA interpretation technique for transformer fault diagnosis | |
| Senoussaoui et al. | Combining and comparing various machine‐learning algorithms to improve dissolved gas analysis interpretation | |
| Abu-Siada et al. | A new fuzzy logic approach for consistent interpretation of dissolved gas-in-oil analysis | |
| Lin et al. | Grey clustering analysis for incipient fault diagnosis in oil-immersed transformers | |
| CN108875783A (en) | A kind of extreme learning machine Diagnosis Method of Transformer Faults towards unbalanced dataset | |
| CN108828377A (en) | A kind of Diagnosis Method of Transformer Faults | |
| CN107132310B (en) | Transformer equipment health status method of discrimination based on gauss hybrid models | |
| Flores et al. | Expert system for the assessment of power transformer insulation condition based on type-2 fuzzy logic systems | |
| CN107907783A (en) | Transformer fault integrated diagnostic system and diagnostic method based on fuzzy association rules | |
| Guo et al. | Transformer failure diagnosis using fuzzy association rule mining combined with case‐based reasoning | |
| CN104655948A (en) | Novel multistage fault diagnosis method for power transformer | |
| CN109490685B (en) | Early defect early warning method of transformer based on-line monitoring of dissolved gas in oil | |
| WO2012029154A1 (en) | Internal abnormality diagnosis method, internal abnormality diagnosis system, and decision tree generating method for internal abnormality diagnosis of oil-filled electric apparatus utilizing gas concentration in oil | |
| CN110188309A (en) | Oil-immersed power transformer defect method for early warning based on Hidden Markov Model | |
| Zhao et al. | A novel approach to transformer fault diagnosis using IDM and naive credal classifier | |
| Gouda et al. | Enhancing the diagnostic accuracy of DGA techniques based on IEC-TC10 and related databases | |
| Dhote et al. | Improvement in transformer diagnosis by DGA using fuzzy logic | |
| CN107103361B (en) | Transformer fault diagnosis method and system based on rough set and answer set | |
| CN112183610A (en) | Transformer fault prediction method and transformer fault prediction device | |
| Bakar et al. | Improvement of DGA interpretation using scoring index method | |
| CN110059773A (en) | A kind of compound diagnostic method of transformer fault | |
| CN108845074B (en) | State evaluation method for oil-immersed transformer | |
| CN108508319A (en) | Transformer fault kind identification method based on linked character between fault characteristic gases | |
| MansourLakouraj et al. | Event classification in active distribution grids using physics-informed graph neural network and PMU measurements |
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: 20181116 |
|
| RJ01 | Rejection of invention patent application after publication |