CN106646014A - Transformer fault diagnosis method - Google Patents
Transformer fault diagnosis method Download PDFInfo
- Publication number
- CN106646014A CN106646014A CN201610837671.2A CN201610837671A CN106646014A CN 106646014 A CN106646014 A CN 106646014A CN 201610837671 A CN201610837671 A CN 201610837671A CN 106646014 A CN106646014 A CN 106646014A
- Authority
- CN
- China
- Prior art keywords
- transformer
- fault diagnosis
- fault
- state
- diagnosis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/005—Testing of electric installations on transport means
Abstract
The invention discloses a transformer fault diagnosis method comprising the following steps: (1) collecting fault cases, and sorting and classifying the cases; (2) building a fault diagnosis rule base; (3) carrying out fault diagnosis according to the state of a transformer; and (4) outputting a fault diagnosis result. According to the invention, transformer parallel fault diagnosis service based on rule reasoning is adopted, a transformer fault diagnosis algorithm based on rule reasoning is accelerated on multiple cores in parallel and encapsulated by WebService, and service is provided externally. Algorithm calling between different languages of different platforms is realized. The interface is clear and flexible. The accuracy of fault diagnosis is high. The operation time is short. The method is very suitable for diagnosing faults of transformers of various voltage grades.
Description
Technical field
The invention belongs to electrical equipment fault detection technique field, and in particular to a kind of Diagnosis Method of Transformer Faults.
Background technology
Transformer is a vitals of power system, and reliability service is huge to the safety effects of power system, its
Fault diagnosis is constantly subjected to the extensive attention of domestic and international academia and engineering circles personnel.In recent years, Chinese scholars propose many
The method for planting transformer fault diagnosis, such as artificial neural network, Bayesian network, expert system and SVMs.Public cyclopentadienyl
Method etc. proposes a kind of method for diagnosing fault of power transformer based on BP network algorithm Optimization of Fuzzy Petri networks, improves event
Barrier rate of correct diagnosis.The characteristics of having randomness and uncertainty for the information in transformer fault diagnosis, Song Gongyi etc. is carried
A kind of Diagnosis Method of Transformer Faults based on fuzzy Bayesian network is gone out.Shi Ruifeng etc. is proposed in a kind of extendible oil
The diagnosing fault of power transformer system of dissolved gas composition, the system has preferably operability and differentiates accuracy rate.Zheng Han
Rich grade is directed to Gases Dissolved in Transformer Oil, it is proposed that a kind of many classification least square method supporting vector machines and Modified particle swarm optimization
The method for diagnosing fault of power transformer for combining.But above-mentioned diagnostic method and system are mainly based on data analysis, to transformation
Relevance and inner link analysis between device running status amount is not enough, and transformer fault diagnosis have error.Therefore, we compel
It is essential and wants a kind of accurate method for judging transformer fault diagnosis.
The content of the invention
It is an object of the invention to provide a kind of Diagnosis Method of Transformer Faults, the method can accurately judge the event of transformer
Barrier.
The technical solution adopted in the present invention is:
A kind of Diagnosis Method of Transformer Faults, comprises the steps:
1), fault case is collected, and arrangement classification is carried out to case
The physical fault diagnosis case of each Utilities Electric Co.'s transformer is collected, with reference to authoritative expert's experience in industry, to case
Example carries out arrangement classification;It is that initial data is carried out to examine checking to arrange classification, and data are sorted out by fault category;
2) Failure Diagnostic Code storehouse, is set up;The rule base includes N bar diagnosis rules;
3), fault diagnosis is carried out according to transformer state
The related quantity of state of input transformer;
N bar diagnosis rules are traveled through;
N bar Failure Diagnostic Codes are divided into into L inference machine, the knowledge rule bar number K that each inference machine is realized is as follows:
Wherein,For the symbol that rounds up;
, used as single function, the related quantity of state of input transformer, L fault request is by OpenMP for each inference machine
Executed in parallel, the result of each inference machine diagnosis is merged, to match optimal rule as transformer fault diagnosis result;
The foundation of best match is quantity of state similarity highest;
4), fault diagnosis result is exported.
Step 2) in, it is the step of set up Failure Diagnostic Code storehouse:Initial data is carried out into discretization, to quantity of state and event
Barrier pattern is numbered, the causality set up between quantity of state and fault mode, so as to form Failure Diagnostic Code storehouse.
Initial data discretization is that continuous status data is carried out into discrete value according to its feature, such as patrolling and examining in
0 represents normally, and 1 represents exception;Winding frequency response by the order of severity take 0,1,2,3 respectively correspond to winding it is normal, slight, bright
Aobvious, gross distortion etc..
Step 3) in, call service, the quantity of state of input transformer to obtain fault diagnosis knot by web Service interface
Really.
Step 3) in, being calculated using multi-core parallel concurrent carries out transformer fault diagnosis, and its step is:
Rule-based reasoning is realized by multiple inference machines;The parallel computation simultaneously on multinuclear of multiple inference machines;Parallel inference is calculated
Method takes full advantage of hardware resource by OpenMP realizations, improves the speed of reasoning.Step 3) it is parallel using fault diagnosis reasoning
Algorithm carries out computing, and the algorithm refers to the rule-based reasoning matching algorithm realized based on OpenMP multi-core parallel concurrents.The number of inference machine
Determine that is, computer there are several cores by the core number of computer, just there are several inference machines;
Step 4) in, transformer fault diagnosis result is exported by web Service interface.
The present invention is by on-line monitoring, live detection, power failure test and patrols and examines the quantity of state for setting up transformer, for while depositing
In on-line monitoring, live detection, the state magnitude of the test that has a power failure, by the test that has a power failure, live detection, the sequencing monitored on-line
Assignment is carried out, so according to the reliability value of data, it is ensured that state magnitude it is accurate and effective.The value of quantity of state according to
The feature of quantity of state carries out value, such as, for 0 representing normal in patrolling and examining, 1 represents extremely;Oil chromatography presses most-often used improvement
Three-ratio method carries out value;Winding frequency response takes 0,1,2,3 and corresponds to winding respectively normally, slightly, substantially, sternly by the order of severity
Deformation etc. again.
The present invention is packaged by WebService to fault diagnosis algorithm, is externally issued in the form of services.Other
System calls service, the quantity of state of input transformer to obtain transformer fault diagnosis result by web Service interface.According to
The result of fault diagnosis, with reference to the situation of transformer field, formulates specific treatment measures, excludes transformer fault.
The beneficial effect comprise that:
The present invention can accurately carry out fault diagnosis to transformer, and diagnostic result can be real-time management and control, fortune accurately and reliably
The daily O&M provided auxiliary decision-making suggestions such as dimension maintenance, emergency first-aid repair and low-voltage improvement, in whole plan (Plan), perform
(Do) in, checking the closed loop management and control flow process of (Check) and disposal (Action), to formulate rationalization scheme reference is provided;
The method also is available for the systems such as production management, marketing, scheduling and calls;
It is of the invention can directly Instructing manufacture operation, repair based on condition of component and fault diagnosis work, greatly improve production cost and
Industry development, the huge economic benefit of generation and the societies such as managerial skills, advanced state monitoring, repair based on condition of component and intelligent O&M
Benefit;
The present invention using multi-core parallel concurrent calculate Method of Fault Diagnosis in Transformer is accelerated, diagnosis algorithm with
The mode of WebService services is externally issued, and realizes calling for algorithm cross-platform cross language, is called for other programs,
It is very simple, convenient.
Description of the drawings
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:
Fig. 1 is the concurrent fault diagnosis procedure chart of Process Based.
Specific embodiment
In order that the objects, technical solutions and advantages of the present invention become more apparent, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that specific embodiment described herein is only to explain the present invention, not
For limiting the present invention.
Referring to Fig. 1, a kind of Diagnosis Method of Transformer Faults comprises the steps:
1), the case of 280 transformer field fault diagnosises is collected, and arrangement classification is carried out to case;Arranging classification is
Initial data is carried out to examine checking, data are sorted out by fault category;
2) Failure Diagnostic Code storehouse, is set up
The case of 280 transformer field fault diagnosises is carried out into discretization, quantity of state is numbered with fault mode,
The causality set up between quantity of state and fault mode, so as to form Failure Diagnostic Code storehouse;Its step is:By initial data
Discretization is carried out, by on-line monitoring, live detection, is had a power failure and is tested and patrol and examine the quantity of state for setting up transformer, for presence simultaneously
On-line monitoring, live detection, the state magnitude of the test that has a power failure, are entered by the sequencing of the test that has a power failure, live detection, on-line monitoring
Row assignment, this reliability value according to data, it is ensured that state magnitude it is accurate and effective;
The value of quantity of state carries out value according to the feature of quantity of state, for example in patrolling and examining 0 represent it is normal, 1 represent it is different
Often;Oil chromatography carries out value by most-often used improvement three-ratio method;Winding frequency response takes 0,1,2,3 difference by the order of severity
Correspond to normal, slight, obvious, gross distortion of winding etc.;
Transformer fault diagnosis quantity of state has 104 in the case of 280 transformer field fault diagnosises, and fault mode has
32, coding schedule is as shown in table 1;
The quantity of state of table 1 is encoded
Numbering | Quantity of state | State value |
M1 | Oil chromatography | Improve three ratios |
M2 | Direct current electricity group | 0:Normally;1:It is abnormal |
M3 | Winding frequency response | 0:Normally;1:Slight deformation;2:Moderate deforms;3:Gross distortion |
M4 | Winding no-load voltage ratio | 0:Normally;1:Higher exception;-1:Low exception |
M5 | Shelf depreciation | 0:Normally;1:It is abnormal |
M6 | Iron core grounding current | 0:Normally;1:It is abnormal |
M7 | Buchholz relay protection act | 0:Normally;1:It is abnormal |
M8 | Infrared measurement of temperature | 0:Normally;1:It is abnormal |
M9 | Top-oil temperature | 0:Normally;1:It is abnormal |
... | ... | ... |
M104 | Fuel tank leakage of oil | 0:Normally;1:It is abnormal |
The fault mode of table 2 is encoded
According to the case of the 280 transformer field fault diagnosises collected, with reference to the experience of authoritative expert in industry, formed
80 knowledge rules, * represents that quantity of state is uncorrelated to fault mode in table 3;
The fault diagnosis knowledge base of table 3
Every knowledge rule is abstracted by transformer fault case, and the case of some reality is contained per rule storehouse
Example.These cases include the information such as detailed troubleshooting measure, facility information, meteorology, picture and video, by checking these
Case, can instruct user to carry out the process and analysis of failure;The knowledge base of the present invention is with expert's field diagnostic experience
Accumulate and constantly enrich, rule base also has self-learning function, can be knowing for fault diagnosis by the case automatically abstracting made a definite diagnosis
Know rule, and be added in Failure Diagnostic Code storehouse;
3), fault diagnosis is carried out according to transformer state
The related quantity of state of input transformer;
N=80 bar diagnosis rules are traveled through;
By taking the server of 16 cores as an example, N=80 bar Failure Diagnostic Codes are divided into into L=16 inference machine, each inference machine
The knowledge rule bar number K=5 of realization;
Each inference machine as single function, the related quantity of state of input transformer, L=16 fault request by
OpenMP executed in parallel, each inference machine finds out the fault diagnosis result matched with the quantity of state of input, by each inference machine
The result of diagnosis is merged, to match optimal rule as transformer fault diagnosis result;
4), transformer fault diagnosis result is exported by web Service interface.
Referring to Fig. 1, only one of which main thread when program starts, the related quantity of state of input transformer, 16 fault reasonings
Machine is by OpenMP executed in parallel, and the result of each inference machine diagnosis is merged by main program, defeated as final diagnostic result
Go out.
In order to verify the validity and feasibility of the inventive method, verified by the case library collected.Experimental situation
It is as follows:CPU:Intel (R) Xeon (R) E7-4830 2.13GHz 2.93GHz, totally 16 core;Internal memory:64G;Running environment:
Visual Studio 2013 and MyEclipse 2014.The service of transformer rule-based reasoning adopts under a windows environment C++
Write, service call adopts written in Java under Linux environment.Using neutral net, Bayesian network and the inventive method pair
280 cases collected carry out fault diagnosis, and the comparison of diagnostic result is as shown in table 4.
The comparative result of 4 three kinds of methods of table
Method | Average operating time (ms) | Accuracy rate |
Neutral net | 7324 | 83.93% |
Bayesian network | 5243 | 81.79% |
The inventive method | 1263 | 93.93% |
The time that the inventive method carries out fault diagnosis is short, and accuracy rate is high, and it is more efficient to compare other two methods, more
It is adapted to diagnose transformer fault.
Fault diagnosis is carried out to certain 10kV transformer with this method, the status data of the transformer exception is as follows:Oil temperature
Higher, infrared live detection finds sleeve pipe heating, and sleeve surface has seminess.Fault diagnosis is carried out using the inventive method
As a result it is:Fault mode is the deterioration of sleeve pipe humidified insulation.Insulaion resistance between porcelain bushing shell and ground is checked by megger, it is thus identified that be
The result that system is diagnosed automatically is consistent with actual conditions, and this method has good directive significance to the operation maintenance personnel at scene.
The present invention proposes a kind of transformer concurrent fault diagnostic service of Process Based, by OpenMP to being based on
The Method of Fault Diagnosis in Transformer of rule-based reasoning carries out multi-core parallel concurrent acceleration, and using WebService modes the algorithm is encapsulated, and
Service is externally provided;The algorithm realized between different platform different language is called, and interface is clearly flexible, fault diagnosis accuracy rate
Height, run time is short, and being especially suitable for the transformer to each electric pressure carries out fault diagnosis.
It should be appreciated that for those of ordinary skills, can according to the above description be improved or be converted,
And all these modifications and variations should all belong to the protection domain of claims of the present invention.
Claims (5)
1. a kind of Diagnosis Method of Transformer Faults, it is characterised in that comprise the steps:
1), fault case is collected, and arrangement classification is carried out to case;
2) Failure Diagnostic Code storehouse, is set up;
3), fault diagnosis is carried out according to transformer state
The related quantity of state of input transformer;
N bar diagnosis rules in Failure Diagnostic Code storehouse are traveled through;
N bar Failure Diagnostic Codes are divided into into L inference machine, the knowledge rule bar number K that each inference machine is realized is as follows:
Wherein,For the symbol that rounds up;
Each inference machine is held parallel as single function, the related quantity of state of input transformer, L inference machine by OpenMP
OK, the result of each inference machine diagnosis is merged, to match optimal rule as transformer fault diagnosis result;
4), fault diagnosis result is exported.
2. Diagnosis Method of Transformer Faults according to claim 1, it is characterised in that:Step 2) in, set up fault diagnosis
The step of rule base is:Initial data is carried out into discretization, quantity of state is numbered with fault mode, set up quantity of state with event
Causality between barrier pattern, so as to form Failure Diagnostic Code storehouse.
3. Diagnosis Method of Transformer Faults according to claim 2, it is characterised in that:Step 3) in, pass through
Web Service interface calls service, the quantity of state of input transformer to obtain fault diagnosis result.
4. Diagnosis Method of Transformer Faults according to claim 3, it is characterised in that:Step 4) in, pass through
Web Service interface exports transformer fault diagnosis result.
5. Diagnosis Method of Transformer Faults according to claim 4, it is characterised in that:Step 4) in fault diagnosis result
Fault case can be automatically stored as, in being added to Failure Diagnostic Code storehouse.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610837671.2A CN106646014A (en) | 2016-09-21 | 2016-09-21 | Transformer fault diagnosis method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610837671.2A CN106646014A (en) | 2016-09-21 | 2016-09-21 | Transformer fault diagnosis method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106646014A true CN106646014A (en) | 2017-05-10 |
Family
ID=58852466
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610837671.2A Pending CN106646014A (en) | 2016-09-21 | 2016-09-21 | Transformer fault diagnosis method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106646014A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108583115A (en) * | 2018-05-02 | 2018-09-28 | 阜阳师范学院 | A kind of manufacturing system of Mouse Embryos model |
CN108596341A (en) * | 2018-04-19 | 2018-09-28 | 中国电子科技集团公司第五十四研究所 | A kind of method for diagnosing faults based on expert system |
CN110443383A (en) * | 2019-07-18 | 2019-11-12 | 西安工程大学 | Transformer electrification method for inspecting and cruising inspection system based on Business Stream |
CN111446069A (en) * | 2020-03-24 | 2020-07-24 | 广东电网有限责任公司电力科学研究院 | Protection method, system, device and equipment for oil-immersed transformer |
CN111856170A (en) * | 2019-04-24 | 2020-10-30 | 中矿龙科能源科技(北京)股份有限公司 | Transformer fault diagnosis system based on harmonic method |
CN113220541A (en) * | 2021-06-10 | 2021-08-06 | 北京全路通信信号研究设计院集团有限公司 | Memory inspection method and system of multi-core processor |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101266279A (en) * | 2008-05-09 | 2008-09-17 | 东北大学 | Electric network failure diagnosis device and method |
CN104133981A (en) * | 2014-06-25 | 2014-11-05 | 国家电网公司 | Photovoltaic power station fault diagnosis method based on fuzzy production rule knowledge base |
CN105260778A (en) * | 2015-11-05 | 2016-01-20 | 国家电网公司 | Power transformer fault diagnosis system based on expert database |
-
2016
- 2016-09-21 CN CN201610837671.2A patent/CN106646014A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101266279A (en) * | 2008-05-09 | 2008-09-17 | 东北大学 | Electric network failure diagnosis device and method |
CN104133981A (en) * | 2014-06-25 | 2014-11-05 | 国家电网公司 | Photovoltaic power station fault diagnosis method based on fuzzy production rule knowledge base |
CN105260778A (en) * | 2015-11-05 | 2016-01-20 | 国家电网公司 | Power transformer fault diagnosis system based on expert database |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108596341A (en) * | 2018-04-19 | 2018-09-28 | 中国电子科技集团公司第五十四研究所 | A kind of method for diagnosing faults based on expert system |
CN108583115A (en) * | 2018-05-02 | 2018-09-28 | 阜阳师范学院 | A kind of manufacturing system of Mouse Embryos model |
CN108583115B (en) * | 2018-05-02 | 2021-06-18 | 阜阳师范学院 | Manufacturing system of mouse early embryo model |
CN111856170A (en) * | 2019-04-24 | 2020-10-30 | 中矿龙科能源科技(北京)股份有限公司 | Transformer fault diagnosis system based on harmonic method |
CN110443383A (en) * | 2019-07-18 | 2019-11-12 | 西安工程大学 | Transformer electrification method for inspecting and cruising inspection system based on Business Stream |
CN111446069A (en) * | 2020-03-24 | 2020-07-24 | 广东电网有限责任公司电力科学研究院 | Protection method, system, device and equipment for oil-immersed transformer |
CN111446069B (en) * | 2020-03-24 | 2021-03-12 | 广东电网有限责任公司电力科学研究院 | Protection method, system, device and equipment for oil-immersed transformer |
CN113220541A (en) * | 2021-06-10 | 2021-08-06 | 北京全路通信信号研究设计院集团有限公司 | Memory inspection method and system of multi-core processor |
CN113220541B (en) * | 2021-06-10 | 2021-09-07 | 北京全路通信信号研究设计院集团有限公司 | Memory inspection method and system of multi-core processor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106646014A (en) | Transformer fault diagnosis method | |
CN112131441B (en) | Method and system for rapidly identifying abnormal electricity consumption behavior | |
CN108073551B (en) | High-voltage switch cabinet online fault diagnosis method based on multi-Agent cooperation | |
CN103018592B (en) | The traction transformer faults diagnostic method of internal model control based PID controller | |
CN103218695A (en) | Secondary equipment intelligence state evaluation diagnostic system and method thereof | |
CN108931972A (en) | A kind of substation secondary device condition intelligent diagnostic method based on model-driven | |
CN101833324B (en) | Intelligent fault diagnosis system in tread extrusion process and diagnosis method thereof | |
CN109061391B (en) | Power grid fault diagnosis method and system based on computer vision tidal current diagram | |
CN103730894A (en) | Diagram checking method and device of energy management system | |
CN106802599A (en) | A kind of diagnosing fault of power transformer system based on expert database | |
CN115453267A (en) | Fault diagnosis system for electric power information system | |
CN110443481B (en) | Power distribution automation terminal state evaluation system and method based on hybrid K-nearest neighbor algorithm | |
CN106093636A (en) | The analog quantity check method of the secondary device of intelligent grid and device | |
CN110361609A (en) | Extra-high voltage equipment monitors system and method | |
CN110674240B (en) | GIS-based distributed multistage intelligent fault diagnosis system for power equipment | |
CN116956203A (en) | Method and system for measuring action characteristics of tapping switch of transformer | |
Zhang et al. | Power grid fault diagnosis model based on the time series density distribution of warning information | |
CN114156865B (en) | Low-voltage distribution network topology generation and fault prediction method considering state perception | |
CN108694447A (en) | Device maintenance method based on power scheduling overhaul data and device | |
CN116823230B (en) | Power distribution network fault state extraction method based on dispatching data network | |
Li et al. | State Perception Method of Intelligent Substation Secondary System Based on FCE and DCNN | |
Pengfei et al. | The condition assessment of transformer bushing based on fuzzy logic | |
Li et al. | Power grid fault detection method based on cloud platform and improved isolated forest | |
Chen et al. | A state assessment method of relay protection device based on pso optimizes svm parameters and cloud model | |
Li et al. | Evaluation method of switchgear state based on adaptive DBSCAN algorithm |
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: 20170510 |
|
RJ01 | Rejection of invention patent application after publication |