CN104007343A - Dynamic comprehensive transformer fault diagnosis method based on Bayesian network - Google Patents

Dynamic comprehensive transformer fault diagnosis method based on Bayesian network Download PDF

Info

Publication number
CN104007343A
CN104007343A CN201410222904.9A CN201410222904A CN104007343A CN 104007343 A CN104007343 A CN 104007343A CN 201410222904 A CN201410222904 A CN 201410222904A CN 104007343 A CN104007343 A CN 104007343A
Authority
CN
China
Prior art keywords
transformer
fault
bayesian network
fault diagnosis
dynamic
Prior art date
Application number
CN201410222904.9A
Other languages
Chinese (zh)
Other versions
CN104007343B (en
Inventor
高文胜
白翠粉
程亭婷
刘通
马仪
Original Assignee
清华大学
南方电网科学研究院有限责任公司
云南电网公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 清华大学, 南方电网科学研究院有限责任公司, 云南电网公司 filed Critical 清华大学
Priority to CN201410222904.9A priority Critical patent/CN104007343B/en
Publication of CN104007343A publication Critical patent/CN104007343A/en
Application granted granted Critical
Publication of CN104007343B publication Critical patent/CN104007343B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a dynamic transformer fault diagnosis method based on the Bayesian network and relates to the technical field of electrical equipment. Under the condition that acquired proof is limited, a comprehensive fault diagnosis model is expanded forwards to the proof acquisition stage, a dynamic fault diagnosis mechanism is presented, the proof acquisition process is optimized according to a certain principle, and a state characteristic quantity which supports the transform fault condition to the maximum degree is selected preferentially. The dynamic fault diagnosis mechanism aims to preferentially select the state characteristic quantity which influences transformer operation fault diagnosis the most to serve as the input parameter of the model, other unnecessary test detection is omitted, and the number of diagnosis items is reduced and an accurate risk estimation value is obtained under the condition of limited resources.

Description

A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network
Technical field
The present invention relates to electrical equipment technical field, be specifically related to a kind of transformer dynamic fault diagnosis method based on Bayesian network.
Background technology
Mostly transformer fault is the slowly process of development, and some status flag amounts can show extremely during this time, and these abnormal amounts are important evidence of fault diagnosis.Oil chromatography is as one of important state characteristic quantity, can reflect the inner most faults of transformer, and oil chromatography data are easier to obtain with respect to other status flag amounts, so become the research emphasis of current transformer fault diagnosis.But the information that oil chromatography is contained is limited, only can carry out preliminary and judgement roughly to the malfunction of transformer, in order to obtain detailed failure mode information, must carry out resultant fault diagnosis to transformer.Resultant fault diagnosis is to merge the process that fault mode that all status flag amounts may exist transformer carries out complex reasoning diagnosis.Due to transformer fault process complexity, status flag measurer has the feature such as ambiguity, incompleteness, and it is not corresponding one by one with fault mode.Therefore, Intelligent Diagnosis Technology is used in existing research more, as the methods such as neural network, Bayesian network, expert system and evidential reasoning are processed problems, so that diagnostic result is more accurate.
In the diagnostic procedure based on resultant fault diagnostic model, the evidence of input model is more, and diagnostic result more approaches truth, and the estimation of probability of malfunction is also just more accurate, if can obtain entire evidence information, favourable to the estimation of risk.But limit because of on-the-spot physical condition, the evidence detecting is all incomplete, the classification of these evidences and quantity directly affect the accuracy that probability of malfunction is estimated.General resultant fault diagnostic model does not relate to the problem of evidence validity, but the evidence detecting is directly inputted to diagnostic model, the fault mode of transformer is judged, this diagnosis mechanism is called to static failure diagnosis mechanism herein, existing resultant fault diagnostic model is nearly all static failure diagnosis mechanism.Static failure diagnosis mechanism is not optimized screening to evidence acquisition process, and the evidence of inputting in diagnostic procedure may be omitted the status flag amount that can reflect transformer fault situation, and this will directly affect the accuracy of probability of malfunction estimation.
Summary of the invention
The object of the invention is to propose a kind of transformer dynamic fault diagnosis method based on Bayesian network, to overcome the deficiency of the static failure diagnostic mode that has method for diagnosing faults, result of calculation is conformed to actual transformer fault situation.
A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network of the embodiment of the present invention, comprises the following steps:
(1) set up the Bayesian network model of transformer fault according to the fault case of transformer and expertise;
(2) collect the partial status characteristic quantity that Transformer monitors value, and be regarded as the evidence of transformer fault diagnosis, be input in the Bayesian network model of setting up in step (1) and carry out dynamic comprehensive fault diagnosis.
In one embodiment of the invention, described step (1) comprises the following steps:
(1-1), according to fault case and the expertise of transformer, compiles common failure pattern F and the status flag amount S of transformer;
Cause-effect relationship between (1-2) failure data analyzing status flag amount S and fault mode F based on expertise and transformer, and set up corresponding Bayesian network model.
In one embodiment of the invention, described step (2) comprises the following steps:
(2-1) collects the partial status characteristic quantity value e that Transformer detects, is regarded as evidence E=e, sets the probability threshold value Pth in dynamic fault diagnosis process;
(2-2), by evidence E input Bayesian network model, ask for the posterior probability of all fault modes, by the fault mode f of posterior probability maximum ibe considered as contingent fault mode, its posteriority probability tables is shown P max;
(2-3) are to the posterior probability P obtaining in step (2-2) maxjudge, if P maxbe more than or equal to P thturn to step (2-6); If P maxbe less than P th, implement in order following steps;
(2-4) assumed fault pattern f ioccur, by evidence E=[e, f i] input Bayesian network model, obtain status flag amount the status flag amount s of middle posterior probability maximum j;
(2-5) implement status flag amount s jverification experimental verification, judge s jwhether abnormal, if s jnot having extremely more fresh evidence is E=[e, s j=0], otherwise E=[e, s j=1], repeating step (2-3) is to (2-5);
(2-6) dynamic fault diagnosis process finishes, and finally obtaining the current fault mode of transformer is f.
The present invention proposes a kind of transformer dynamic fault diagnosis method based on Bayesian network, its advantage is that the computation process of the inventive method has adopted dynamic fault diagnosis mechanism, has reduced the blindness of diagnostic procedure, can coincide better with actual transformer fault situation, applicability is higher.
Embodiment
Describe embodiments of the invention below in detail, the example of described embodiment illustrates, wherein same or similar label represents same or similar element or has the element of identical or similar functions from start to finish.Be exemplary below by the embodiment describing, be intended to for explaining the present invention, and can not be interpreted as limitation of the present invention.
Obtaining evidence limited in the situation that, the present invention extends to forward evidence by resultant fault diagnostic model and obtains the stage, proposition dynamic fault diagnosis mechanism, optimize evidence acquisition process according to certain principle, preferentially select the status flag amount of transformer fault situation support maximum as evidence information.Dynamic fault diagnosis mechanism is intended to preferential status flag amount that transformer operation troubles diagnostic procedure the is had the greatest impact input parameter as model of selecting, detect and omit other unnecessary tests, the in the situation that of resource-constrained, reduce checkup item and can obtain again Risk parameter comparatively accurately.
A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network of the embodiment of the present invention, comprises the following steps:
(1) set up the Bayesian network model of transformer fault according to the fault case of transformer and expertise, specifically comprise the following steps:
(1-1), according to fault case and the expertise of transformer, compiles common failure pattern F and the status flag amount S of transformer.
F=[f 1,f 2,f 3,f 4,f 5,f 6,f 7,f 8,f 9,f 10];S=[s 1,s 2,s 3,s 4,s 5,s 6,s 7,s 8,s 9];
Wherein, f 1for multipoint earthing of iron core, f 2for insulation ag(e)ing, f 3for leakage field heating, f 4for short circuit in winding, f 5for humidified insulation, f 6for shunting switch fault, f 7for suspended discharge, f 8for screen electric discharge, f 9for winding deformation, f 10for discharging in oil; s 1for iron core grounding current, s 2for code of direct ratio was hot fault signature, s 3for the three-phase imbalance coefficient of winding D.C. resistance, s 4for transformer body Water in oil amount, s 5for code of direct ratio is discharging fault feature, s 6for winding no-load voltage ratio deviation, s 7for shelf depreciation, s 8for , s 9for absorptance or the polarization index of winding.
Cause-effect relationship between (1-2) failure data analyzing status flag amount S and fault mode F based on expertise and transformer, and set up corresponding Bayesian network model.
Known by analyzing, F is S reason, the result that S is F, and between F and S, cause and effect intensity matrix R is
Wherein, line display F, S is shown in list, R ijvalue represent f iplant fault mode and cause s jplant status flag amount abnormal probability occurs, special, "-" represents noncausal relationship between the two.
R is in Bayesian network the annexation of the two, can directly set up Bayesian network according to R.Another important parameter of Bayesian network, the prior probability P of F initial, the fault statistics data based on transformer can obtain:
P initial=[0.45,0.11,0.13,0.12,0.10,0.26,0.16,0.28,0.24,0.14]。
(2) value of the partial status characteristic quantity that collection Transformer monitors , and be regarded as the evidence of transformer fault diagnosis, be input in the Bayesian network model of setting up in step (1) and carry out dynamic comprehensive fault diagnosis.Wherein S partwith complementation, the two set of adding up equals S.Step (2) specifically comprises the following steps:
(2-1) collects the partial status characteristic quantity value e=[s that Transformer detects 2=1, s 5=0], be regarded as evidence E=e, set the probability threshold value P in dynamic fault diagnosis process th=0.8.
Wherein the status flag value of measuring is that 0 expression does not occur extremely, and 1 represents to occur extremely.
(2-2), by evidence E input Bayesian network model, ask for the posterior probability of all fault modes, and the fault mode of posterior probability maximum is f 1, be regarded as contingent fault mode, its posteriority probability P max=0.1965.
(2-3) are to the posterior probability P obtaining in step (2-2) maxjudge P maxbe less than P th.
(2-4) assumed fault pattern f1 occurs, by evidence E=[e, f] input Bayesian network model, obtain status flag amount the status flag amount of middle posterior probability maximum is s 1.
(2-5) implement status flag amount s 1verification experimental verification, s 1do not occur extremely, more fresh evidence is E=[e, s 1=0];
Repeating step (2-2) is to (2-5), until P maxbe more than or equal to P thtill.
(2-6) dynamic fault diagnosis process finishes, and finally obtaining the current fault mode of transformer is f 6.
In the description of this instructions, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means to be contained at least one embodiment of the present invention or example in conjunction with specific features, structure, material or the feature of this embodiment or example description.In this manual, to the schematic statement of above-mentioned term not must for be identical embodiment or example.And, specific features, structure, material or the feature of description can one or more embodiment in office or example in suitable mode combination.In addition,, not conflicting in the situation that, those skilled in the art can carry out combination and combination by the feature of the different embodiment that describe in this instructions or example and different embodiment or example.
Although illustrated and described embodiments of the invention above, be understandable that, above-described embodiment is exemplary, can not be interpreted as limitation of the present invention, and those of ordinary skill in the art can change above-described embodiment within the scope of the invention, amendment, replacement and modification.

Claims (3)

1. the transformer dynamic comprehensive method for diagnosing faults based on Bayesian network, is characterized in that, comprises the following steps:
(1) set up the Bayesian network model of transformer fault according to the fault case of transformer and expertise;
(2) collect the partial status characteristic quantity that Transformer monitors value, and be regarded as the evidence of transformer fault diagnosis, be input in the Bayesian network model of setting up in step (1) and carry out dynamic comprehensive fault diagnosis.
2. the transformer dynamic comprehensive method for diagnosing faults based on Bayesian network according to claim 1, is characterized in that, described step (1) comprises the following steps:
(1-1), according to fault case and the expertise of transformer, compiles common failure pattern F and the status flag amount S of transformer;
Cause-effect relationship between (1-2) failure data analyzing status flag amount S and fault mode F based on expertise and transformer, and set up corresponding Bayesian network model.
3. the transformer dynamic comprehensive method for diagnosing faults based on Bayesian network according to claim 1, is characterized in that, described step (2) comprises the following steps:
(2-1) collects the partial status characteristic quantity value e that Transformer detects, is regarded as evidence E=e, sets the probability threshold value P in dynamic fault diagnosis process th;
(2-2), by evidence E input Bayesian network model, ask for the posterior probability of all fault modes, by the fault mode f of posterior probability maximum ibe considered as contingent fault mode, its posteriority probability tables is shown P max;
(2-3) are to the posterior probability P obtaining in step (2-2) maxjudge, if P maxbe more than or equal to P thturn to step (2-6); If P maxbe less than P th, implement in order following steps;
(2-4) assumed fault pattern f ioccur, by evidence E=[e, f i] input Bayesian network model, obtain status flag amount the status flag amount s of middle posterior probability maximum j;
(2-5) implement status flag amount s jverification experimental verification, judge s jwhether abnormal, if s jnot having extremely more fresh evidence is E=[e, s j=0], otherwise E=[e, s j=1], repeating step (2-3) is to (2-5);
(2-6) dynamic fault diagnosis process finishes, and finally obtaining the current fault mode of transformer is f.
CN201410222904.9A 2014-05-23 2014-05-23 A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network Active CN104007343B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410222904.9A CN104007343B (en) 2014-05-23 2014-05-23 A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410222904.9A CN104007343B (en) 2014-05-23 2014-05-23 A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network

Publications (2)

Publication Number Publication Date
CN104007343A true CN104007343A (en) 2014-08-27
CN104007343B CN104007343B (en) 2018-04-20

Family

ID=51368095

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410222904.9A Active CN104007343B (en) 2014-05-23 2014-05-23 A kind of transformer dynamic comprehensive method for diagnosing faults based on Bayesian network

Country Status (1)

Country Link
CN (1) CN104007343B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105242129A (en) * 2015-08-28 2016-01-13 广西电网有限责任公司电力科学研究院 Fault probability determination method for transformer winding
CN105372528A (en) * 2015-11-24 2016-03-02 湖南大学 Power transformer internal fault condition maintenance method
CN107846016A (en) * 2017-11-16 2018-03-27 中国南方电网有限责任公司 A kind of Distribution Network Failure localization method and equipment based on Bayes and Complex event processing
CN107907778A (en) * 2017-10-31 2018-04-13 华北电力大学(保定) A kind of Synthesized Diagnosis On Transformer Faults method based on multiple features parameter
CN111272222A (en) * 2020-02-28 2020-06-12 西南交通大学 Transformer fault diagnosis method based on characteristic quantity set

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102411106A (en) * 2011-11-18 2012-04-11 广东电网公司广州供电局 Fault monitoring method and device for power transformer
CN102779230A (en) * 2012-06-14 2012-11-14 华南理工大学 State analysis and maintenance decision judging method of power transformer system
CN103197177A (en) * 2013-03-20 2013-07-10 山东电力集团公司济宁供电公司 Transformer fault diagnosis analysis method based on bayesian network
CN103245861A (en) * 2013-05-03 2013-08-14 云南电力试验研究院(集团)有限公司电力研究院 Transformer fault diagnosis method based on Bayesian network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102411106A (en) * 2011-11-18 2012-04-11 广东电网公司广州供电局 Fault monitoring method and device for power transformer
CN102779230A (en) * 2012-06-14 2012-11-14 华南理工大学 State analysis and maintenance decision judging method of power transformer system
CN103197177A (en) * 2013-03-20 2013-07-10 山东电力集团公司济宁供电公司 Transformer fault diagnosis analysis method based on bayesian network
CN103245861A (en) * 2013-05-03 2013-08-14 云南电力试验研究院(集团)有限公司电力研究院 Transformer fault diagnosis method based on Bayesian network

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
吕启深 等: "基于贝叶斯网络和粗糙集约简的变压器故障诊断", 《中国电力》 *
吴立增 等: "基于贝叶斯网络分类器的变压器综合故障诊断方法", 《电工技术学报》 *
白翠粉 等: "基于3层贝叶斯网络的变压器综合故障诊断", 《高电压技术》 *
赵文清: "基于选择性贝叶斯分类器的变压器故障诊断", 《电力自动化设备》 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105242129A (en) * 2015-08-28 2016-01-13 广西电网有限责任公司电力科学研究院 Fault probability determination method for transformer winding
CN105242129B (en) * 2015-08-28 2018-03-13 广西电网有限责任公司电力科学研究院 A kind of transformer winding fault probability determination method
CN105372528A (en) * 2015-11-24 2016-03-02 湖南大学 Power transformer internal fault condition maintenance method
CN105372528B (en) * 2015-11-24 2018-10-09 湖南大学 A kind of state maintenance method of Power Transformer Internal Faults and New Transformer
CN107907778A (en) * 2017-10-31 2018-04-13 华北电力大学(保定) A kind of Synthesized Diagnosis On Transformer Faults method based on multiple features parameter
CN107907778B (en) * 2017-10-31 2020-06-19 华北电力大学(保定) Transformer comprehensive fault diagnosis method based on multiple characteristic parameters
CN107846016A (en) * 2017-11-16 2018-03-27 中国南方电网有限责任公司 A kind of Distribution Network Failure localization method and equipment based on Bayes and Complex event processing
CN111272222A (en) * 2020-02-28 2020-06-12 西南交通大学 Transformer fault diagnosis method based on characteristic quantity set

Also Published As

Publication number Publication date
CN104007343B (en) 2018-04-20

Similar Documents

Publication Publication Date Title
Seem Using intelligent data analysis to detect abnormal energy consumption in buildings
EP2466322B1 (en) Method and apparatus for transformer diagnosis
Corino et al. Radiomic analysis of soft tissues sarcomas can distinguish intermediate from high‐grade lesions
Tate et al. Line outage detection using phasor angle measurements
CN103139011B (en) The resultant fault diagnosis of communication in car and Forecasting Methodology
CN104142466B (en) A kind of AUTOMOTIVE RELAY contact closure condition detecting system and its detection method
CN100554980C (en) Be used for supporting the method for the fault functional unit of recognology device
US20150377982A1 (en) Electrical Power Diagnostic System and Methods
Do Coutto Filho et al. Forecasting-aided state estimation—Part II: Implementation
Coble et al. Identifying optimal prognostic parameters from data: a genetic algorithms approach
Kim et al. General model for the risk priority number in failure mode and effects analysis
US20080183406A1 (en) Online IED Fault Diagnosis Device and Method for Substation Automation System Based on IEC61850
Abbas et al. An intelligent diagnostic/prognostic framework for automotive electrical systems
Strachan et al. Providing decision support for the condition-based maintenance of circuit breakers through data mining of trip coil current signatures
CN108544925B (en) Battery management system
CN107229827A (en) The data check method for early warning and device tested and analyzed for autoimmune disease
DE102011010605A1 (en) Functional prediction for a complex system using error modeling
CN106199276B (en) The intelligent diagnosis system and method for exception information in a kind of power information acquisition system
WO2016144496A1 (en) Battery test system with camera
Coble et al. Applying the general path model to estimation of remaining useful life
KR101539896B1 (en) Method for diagnosis of induction motor fault
CN106384210B (en) A kind of power transmission and transforming equipment maintenance prioritization method based on maintenance risk income
KR102142745B1 (en) Method for determining a control observer for the soc
TWI386652B (en) Method and system for recognizing status of electric appliances, and computer program product thereof
CN102947718B (en) String failure monitoring method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant