CN110133410A - Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm - Google Patents
Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm Download PDFInfo
- Publication number
- CN110133410A CN110133410A CN201910468729.4A CN201910468729A CN110133410A CN 110133410 A CN110133410 A CN 110133410A CN 201910468729 A CN201910468729 A CN 201910468729A CN 110133410 A CN110133410 A CN 110133410A
- Authority
- CN
- China
- Prior art keywords
- characteristic gas
- fault
- fuzzy
- transformer
- sample data
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
Abstract
The present invention provides the Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm, including the following steps: obtains fault sample data, carries out characteristic gas normalized to it, obtains the data set of the constituent content comprising the characteristic gas；Construct the Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm；Above-mentioned data set is input to above-mentioned Fault Diagnosis Model for Power Transformer, fault type is exported, obtains fault diagnosis result.Meanwhile the present invention provides the transformer fault diagnosis systems based on Fuzzy C-Means Cluster Algorithm.Using technical solution of the present invention, by selecting characteristic gas normalization method handling failure sample data, normalization processing method is reduced to the influence degree of cluster result, and then improve the fault diagnosis accuracy rate of Fuzzy C-Means Cluster Algorithm, final realize improves transformer fault diagnosis accuracy, really to realize that power transmission and transformation equipment state overhauling provides strong technological means.
Description
Technical field
It is the present invention relates to transformer fault diagnosis technical field, in particular to a kind of based on Fuzzy C-Means Cluster Algorithm
Diagnosis Method of Transformer Faults.Meanwhile the invention further relates to a kind of transformer fault diagnosis based on Fuzzy C-Means Cluster Algorithm
System.
Background technique
The rapid development of State Grid's industry, electric system are being growing.Electrical equipment malfunction is always to threaten power grid
The principal element of safety, to guarantee Operation of Electric Systems reliability, it is necessary to which real-time monitoring electric equipment operation state checks in time
Initial failure out.In various electrical equipments, power transformer is one of most important electrical equipment in electric system, due to becoming
Depressor long-term work is run, inevitably incipient fault problem.Prevention and treatment and the hair for reducing transformer fault and accident
It is raw, guarantee the normal operation of transformer, is electric system problem in the urgent need to address.The tradition of many detection transformer faults
Method cannot obtain the real time information of transformer state, the development of mismatch state maintenance modernization.Therefore, to transformer state
Carrying out real-time monitoring is particularly important.To prevent accident from damaging to transformer, it is necessary to inside transformer truth
It is grasped in time, this just needs to be realized by transformer online monitoring and fault diagnosis.Accurately event is carried out to transformer
Barrier diagnosis not only increases the reliability of electric system, and has apparent economic benefit, therefore, studies transformer fault
Diagnostic techniques has great importance.
Dissolved gas analysis technology (DGA) is the main means of power transformer interior fault diagnosis, it is understanding transformation indirectly
General incipient faults in device provide foundation.By the measurement and analysis to gas is decomposed in transformer, transformation can be tentatively judged
The internal state of device.Routine diagnostic method based on technology has the methods of characteristic gas method, gas component ratio method.Due in oil
Dissolved gas is that many factors are coefficient as a result, so electric power transformer insulated fault type and oil dissolved gas component
There are certain ambiguity, the features such as uncertain and non-linear for relationship between content.In recent years, many scholars are to based on skill
The transformer fault diagnosis of art conducts in-depth research, and a variety of data minings and artificial intelligence theory and technology are applied to should
Field, clustering are one of core contents of data mining, and it is to further increase electricity that DGA technology is combined with clustering
A kind of effective way of power transformer fault diagnosis accuracy, improve to a certain extent fault diagnosis application accuracy rate and
Practicability has important theoretical research and practical application value.
CN107656154A discloses a kind of Diagnosis Method of Transformer Faults based on improvement Fuzzy C-Means Cluster Algorithm,
Including obtaining Gases Dissolved in Transformer Oil data and fault type data as sample, the sample being divided into training sample
And test sample；Gases Dissolved in Transformer Oil data in the sample are handled, and determine the training sample
Class number and corresponding initial cluster center of all categories；It is further determined that using improvement Fuzzy C-Means Cluster Algorithm described
The corresponding new cluster centre of each classification of training sample calculates test sample and belongs to probability of all categories；According to test sample category
In probability of all categories and it is of all categories in ratio shared by each fault type, calculate the generation that test sample corresponds to each fault type
Probability, and determine according to the probability of happening fault type of the test sample.Grinding about transformer fault diagnosis at present
Study carefully the building for being mostly focused on algorithm and model, not yet recognizes that method for normalizing is also associated in based on fuzzy C-means clustering
The accuracy of the Diagnosis Method of Transformer Faults of algorithm.
Summary of the invention
In view of this, the present invention is directed to propose a kind of Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm,
To improve the fault diagnosis accuracy rate of Fuzzy C-Means Cluster Algorithm as far as possible, it is finally reached and improves transformer fault diagnosis standard
The purpose of true property.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm, includes the following steps
(1) fault sample data are obtained, characteristic gas normalized is carried out to the fault sample data, is included
The data set of the constituent content of the characteristic gas；
(2) Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm is constructed；
(3) data set described in step (1) is input to the change based on Fuzzy C-Means Cluster Algorithm described in step (2)
Depressor fault diagnosis model exports fault type, obtains fault diagnosis result.
Further, the formula of the characteristic gas normalized are as follows:
In formula, xiFor the content before characteristic gas normalization；xmaxFor feature gas described in the fault sample data
Body content maximum value；xminFor characteristic gas content minimum value described in the fault sample data；xi' it is the characteristic gas
Content after normalization.
Further, the method for the characteristic gas normalized is characterized gas concentration normalization method.
Further, the characteristic gas includes H2、CH4、C2H2、C2H4And C2H6；The fault type includes high temperature mistake
Heat, high-energy discharge, low energy electric discharge, shelf depreciation and middle cryogenic overheating.
Further, the formula of the characteristic gas concentration normalization method is
In formula, XijContent before indicating the characteristic gas normalization, X 'ijAfter representing the characteristic gas normalization
Content, each characteristic gas total amount in fault sample data where denominator represents the characteristic gas；
Or
In formula, H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、
CH4、C2H2、C2H4And C2H6Content before each characteristic gas normalization.
Further, the transformer fault diagnosis system based on Fuzzy C-Means Cluster Algorithm includes
Fault sample data acquisition module, for obtaining the fault sample data；
Fault sample data processing module is obtained for fault sample data described in normalized comprising the feature
The data set of the constituent content of gas；
Fault diagnosis result generation module, for obtaining transformer fault diagnosis as a result, will be described comprising the data set
It is input in the Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm, exports fault type, obtain transformer fault
Diagnostic result.
Further, the fault sample data processing module is used for normalized fault sample data, is included
The data set of the constituent content of the characteristic gas, specifically includes:
The formula of the characteristic gas normalized are as follows:
In formula, xiFor the content before characteristic gas normalization；xmaxFor feature gas described in the fault sample data
Body content maximum value；xminFor characteristic gas content minimum value described in the fault sample data；xi' it is the characteristic gas
Content after normalization.
Further, the fault sample data processing module is used for normalized fault sample data, is included
The data set of the constituent content of the characteristic gas, specifically includes: the method for the characteristic gas normalized is characterized gas
Bulk concentration normalization method.
Further, the fault sample data processing module is used for normalized fault sample data, is included
The data set of the constituent content of the characteristic gas, specifically includes:
The characteristic gas includes H2、CH4、C2H2、C2H4And C2H6；The fault type includes that hyperthermia and superheating, high energy are put
Electricity, low energy electric discharge, shelf depreciation and middle cryogenic overheating.
Further, fault sample data processing module is stated, normalized fault sample data are used for, is obtained comprising institute
The data set for stating the constituent content of characteristic gas, specifically includes:
The formula of the characteristic gas concentration normalization method is
In formula, XijContent before indicating the characteristic gas normalization, X 'ijAfter representing the characteristic gas normalization
Content, each characteristic gas total amount in fault sample data where denominator represents the characteristic gas；
Or
In formula, H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、
CH4、C2H2、C2H4And C2H6Content before each characteristic gas normalization.Compared with the existing technology, the present invention has following excellent
Gesture: using technical solution of the present invention, by optimizing the normalization processing method of fault sample data, selects characteristic gas normalizing
Change method handling failure sample data reduces normalization processing method to the influence degree of cluster result, optimizes cluster result, in turn
The fault diagnosis accuracy rate for improving Fuzzy C-Means Cluster Algorithm is finally reached the purpose for improving transformer fault diagnosis accuracy,
Really to realize that power transmission and transformation equipment state overhauling provides strong technological means.Meanwhile the present invention can mention for overhaul of the equipments
Judge for initial stage, the work for not allowing substitution is played for discovery inside transformer defect that may be present or performance deterioration ahead of time
With.In addition, the present invention can reduce the incidence of major accident, the maintenance quantity and maintenance cost of equipment are reduced.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.The present invention relates to a kind of Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm, main design ideas
It is: in view of the sensitivity level of gas with various reaction failure is different, therefore passes through the normalized of optimization fault sample data
Method reduces normalization processing method to the influence degree of cluster result, optimizes cluster result.Pass through the whole design thought
Setting, can be improved the fault diagnosis accuracy rate of Fuzzy C-Means Cluster Algorithm, and it is accurate to be finally reached raising transformer fault diagnosis
The purpose of property.
Based on design philosophy as above, in the specific limits scheme of one of which of the invention, it is based on Fuzzy C-Means Cluster Algorithm
Diagnosis Method of Transformer Faults, including the following steps:
(1) fault sample data are obtained, characteristic gas normalized is carried out to the fault sample data, is included
The data set of the constituent content of the characteristic gas；
(2) Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm is constructed；
(3) data set in step (1) is input to the transformer based on Fuzzy C-Means Cluster Algorithm described in step (2)
Fault diagnosis model exports fault type, obtains fault diagnosis result.
In Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm, due to fault sample data in dimension and
Difference on the order of magnitude, therefore before it is input to the Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm, it must be right
Fault sample data make normalized.Since the sensitivity of gas with various reaction failure is different, so its method for normalizing meeting
Cluster result is directly affected, and then influences the fault diagnosis accuracy rate of Fuzzy C-Means Cluster Algorithm.The present embodiment passes through optimization event
Hinder sample data normalization processing method, select characteristic gas normalization method handling failure sample data, then as
The input of Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm can reduce normalization processing method to cluster
As a result influence degree, and then the fault diagnosis accuracy rate of Fuzzy C-Means Cluster Algorithm is improved, it is finally reached and improves transformer event
Hinder the purpose of diagnostic accuracy, really to realize that power transmission and transformation equipment state overhauling provides strong technological means.
Meanwhile the present embodiment can also provide initial stage judgement for overhaul of the equipments, for discovery inside transformer may ahead of time
Existing defect or performance deterioration play the role of not allowing substitution.In addition, the present invention can reduce the incidence of major accident, drop
The low maintenance quantity and maintenance cost of equipment.
In order to further improve the accuracy of the Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm,
In the specific limits scheme of another kind of the invention, the formula of the characteristic gas normalized are as follows:
In formula, xiFor the content before characteristic gas normalization；xmaxFor feature gas described in the fault sample data
Body content maximum value；xminFor characteristic gas content minimum value described in the fault sample data；xi' it is the characteristic gas
Content after normalization.The method for normalizing can be improved the classification accuracy of transformer, and then improves fuzzy C-means clustering and calculate
The fault diagnosis accuracy of method.
In order to further increase the classification accuracy of the Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm,
In a kind of specific embodiment of the invention, the method for the characteristic gas normalized is characterized gas concentration normalization
Method.
H is chosen in the present embodiment2、CH4、C2H2、C2H4And C2H6This five kinds of gases are as characteristic gas, they are in certain journey
The hyperthermia and superheating of power transformer, high-energy discharge, low energy electric discharge, shelf depreciation and middle cryogenic overheating this five kinds of failures are carried on degree
Information, this corresponding five kinds of fault types.
As an embodiment, the formula of the characteristic gas concentration normalization method is
In formula, XijContent before indicating the characteristic gas normalization, X 'ijAfter representing the characteristic gas normalization
Content, each characteristic gas total amount in fault sample data where denominator represents the characteristic gas；
However, the sensitivity level that gas with various reacts failure is different in practical dissolved gas analysis, although such as H2Content
It is more, but it is lower to the sensitivity of breakdown judge, and C2H2Content is less, but contributes fault identification larger.It is as a result,
Reduction H2Judgement to Determination of Alkane Content reduces normalization processing method to the influence degree of cluster result, and then improves Fuzzy C
The fault diagnosis accuracy rate of means clustering algorithm, as another embodiment, the characteristic gas concentration normalization method
Formula is
H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、CH4、
C2H2、C2H4And C2H6Content before each characteristic gas normalization.
The present invention also provides a kind of transformer fault diagnosis systems based on Fuzzy C-Means Cluster Algorithm, comprising:
Fault sample data acquisition module, for obtaining the fault sample data；
Fault sample data processing module is obtained for fault sample data described in normalized comprising the feature
The data set of the constituent content of gas；
Fault diagnosis result generation module, for obtaining transformer fault diagnosis as a result, will be described comprising the data set
It is input in the Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm, exports fault type, obtain transformer fault
Diagnostic result.
It should be noted that in the Diagnosis Method of Transformer Faults provided by the invention based on Fuzzy C-Means Cluster Algorithm
The step of, it can use corresponding module, device etc. in the transformer fault diagnosis system based on Fuzzy C-Means Cluster Algorithm and give
To realize, the technical solution that those skilled in the art are referred to the system realizes the step process of the method, i.e., described
Embodiment in system can be regarded as realizing the preference of the method, and it will not be described here.
Based on whole design as above, following embodiments carry out specifically a portion concrete application under the design
It is bright.
Embodiment 1
The present embodiment is related to a kind of Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm, including following step
It is rapid:
(1) fault sample data are obtained, characteristic gas normalized, the normalizing are carried out to the fault sample data
Change any one of processing method in following formula (a), (b), (c):In formula (a) formula, xiIt is described
Content before characteristic gas normalization；xmaxFor characteristic gas content maximum value described in the fault sample data；xminFor institute
State characteristic gas content minimum value described in fault sample data；xi' for the characteristic gas normalization after content.In formula (b) formula, XijContent before indicating the characteristic gas normalization,
X′ijContent after representing the characteristic gas normalization, each institute in fault sample data where denominator represents the characteristic gas
State characteristic gas total amount.
In formula, H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、
CH4、C2H2、C2H4And C2H6Content before each characteristic gas normalization.
Fault sample data obtain the data set comprising characteristic gas constituent content after the above method normalizes:
In formula, the normalized data set of X ' expression, xijFor the characteristic value of cluster sample j, index i, wherein i=1,
2 ..., m, j=1,2 ..., n.
(2) Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm is constructed, including the following steps:
Firstly, setting cluster classification number c, stopping iteration threshold ε and initialization cluster centre V (0)；
Secondly, calculating and updating degree of membership U (t)；
Again, each cluster centre V (t) is updated.
Finally, the subordinating degree function u acquired according to above-mentioned stepsijWith cluster centre Vi, calculating target function, if its value is big
In ε, then recalculates and update degree of membership U (t), updates each cluster centre V (t), calculating target function, until its value is less than ε,
Stop iteration, obtains the cluster centre, fuzzy membership function and the number of iterations of each classification.
Specifically, firstly, setting cluster classification number c (2≤c≤n), the cluster centre vector of c classification are as follows:
Setting stops iteration threshold ε (> 0), ε is smaller, and computational accuracy is higher；Weighted Index q, q (1 ∈, ∞)；
Setting initialization subordinated-degree matrix:
In formula, uhjThe relative defects of classification h, h=1,2 ..., c are belonged to for j.
For the optimal fuzzy classification of acquisition, target objective function must be made to reach minimum:
dij 2=| | Xj-Vi||2
In formula: U is initial subordinated-degree matrix；V is cluster centre, i.e. V={ V1,V2…,Vc}；M is Smoothness Index；uhjFor
The Euclidean distance of the j sample and ith cluster center,；dijIt is j-th of sample to the degree of membership at ith cluster center.
Secondly, according to formula(h≤c, j≤n) is calculated and is updated degree of membership U (t).
Again, according to formula(i≤m, h≤c) updates each cluster centre V (t).
Finally, the subordinating degree function U and cluster centre V that are asked according to above-mentioned steps, calculating target function, until max (| aij-
aij|) < ε, stop iteration, obtains the cluster centre V (t) of each classification, fuzzy membership function uiAnd the number of iterations.
(3) data set in step (1) is input to the Fault Diagnosis Model for Power Transformer in step (2), obtains each class
Other cluster centre V (t), fuzzy membership function uiAnd the number of iterations, fault sample point is realized according to Subject Matrix U (t)
Class, so that the affiliated fault type of judgement sample, obtains fault diagnosis result.
Embodiment 2
The present embodiment is related to the verifying of the Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm.
Firstly, the present embodiment has determined that 100 groups of fault type from having delivered to collect to sort out in document and laboratory report
Sample (is shown in Table 1), as fault sample data.Transformer fault type is divided into 5 kinds, respectively middle cryogenic overheating,
Hyperthermia and superheating, low energy electric discharge, high-energy discharge and shelf depreciation.
Table 1: fault sample data composition
Above-mentioned 100 groups of fault sample data are used into three kinds of normalization processing methods described in step (1) in embodiment 1 respectively
Formula a), formula b) and formula c) are handled, and the data set comprising characteristic gas constituent content is obtained.
Secondly, according to the transformer fault diagnosis based on Fuzzy C-Means Cluster Algorithm described in step (2) in embodiment 1
The method of model constructs Fault Diagnosis Model for Power Transformer.
Finally, the data set for including characteristic gas constituent content is inputted the change based on Fuzzy C-Means Cluster Algorithm
Depressor fault diagnosis model determines the size of the degree of membership of cluster centre the affiliated failure classes of sample according to fault sample data
Type obtains fault diagnosis result (being shown in Table 2).
2 fault sample test result of table
Table 2 shows the fault sample test result using 3 kinds of method for normalizing.Data in analytical table are it is found that this feature
Gas normalization method can improve the accuracy of the transformer fault diagnosis based on Fuzzy C-Means Cluster Algorithm, it was demonstrated that suitably return
One change method can optimize cluster result, and then improve the accurate of the transformer fault diagnosis based on Fuzzy C-Means Cluster Algorithm
Degree.In addition, from the point of view of integration test result, this method for hyperthermia and superheating and shelf depreciation judgement relative to other fault types more
It is accurate.
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 replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of Diagnosis Method of Transformer Faults based on Fuzzy C-Means Cluster Algorithm, it is characterised in that: include the following steps
(1) fault sample data are obtained, characteristic gas normalized is carried out to the fault sample data, is obtained comprising described
The data set of the constituent content of characteristic gas；
(2) Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm is constructed；
(3) data set described in step (1) is input to the transformer based on Fuzzy C-Means Cluster Algorithm described in step (2)
Fault diagnosis model exports fault type, obtains fault diagnosis result.
2. the Diagnosis Method of Transformer Faults according to claim 1 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the formula of the characteristic gas normalized are as follows:
In formula, xiFor the content before characteristic gas normalization；xmaxContain for characteristic gas described in the fault sample data
Measure maximum value；xminFor characteristic gas content minimum value described in the fault sample data；xi' it is the characteristic gas normalizing
Content after change.
3. the Diagnosis Method of Transformer Faults according to claim 1 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the method for the characteristic gas normalized is characterized gas concentration normalization method.
4. the Diagnosis Method of Transformer Faults according to claim 3 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the characteristic gas includes H2、CH4、C2H2、C2H4And C2H6；The fault type includes hyperthermia and superheating, high-energy discharge, low energy
Electric discharge, shelf depreciation and middle cryogenic overheating.
5. the Diagnosis Method of Transformer Faults according to claim 4 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the formula of the characteristic gas concentration normalization method is
In formula, XijContent before indicating the characteristic gas normalization, X 'ijContent after representing the characteristic gas normalization,
Each characteristic gas total amount in fault sample data where denominator represents the characteristic gas；
Or
In formula, H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、CH4、
C2H2、C2H4And C2H6Content before each characteristic gas normalization.
6. a kind of transformer fault diagnosis system based on Fuzzy C-Means Cluster Algorithm, it is characterised in that: described to be based on Fuzzy C
The transformer fault diagnosis system of means clustering algorithm includes
Fault sample data acquisition module, for obtaining the fault sample data；
Fault sample data processing module is obtained for fault sample data described in normalized comprising the characteristic gas
Constituent content data set；
Fault diagnosis result generation module, for obtaining transformer fault diagnosis as a result, described will input comprising the data set
Into the Fault Diagnosis Model for Power Transformer based on Fuzzy C-Means Cluster Algorithm, fault type is exported, transformer fault diagnosis is obtained
As a result.
7. the transformer fault diagnosis system according to claim 6 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the fault sample data processing module is used for normalized fault sample data, obtains comprising the characteristic gas
The data set of constituent content, specifically includes:
The formula of the characteristic gas normalized are as follows:
In formula, xiFor the content before characteristic gas normalization；xmaxContain for characteristic gas described in the fault sample data
Measure maximum value；xminFor characteristic gas content minimum value described in the fault sample data；xi' it is the characteristic gas normalizing
Content after change.
8. the transformer fault diagnosis system according to claim 6 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the fault sample data processing module is used for normalized fault sample data, obtains comprising the characteristic gas
The data set of constituent content, specifically includes: the method for the characteristic gas normalized is characterized gas concentration normalization method.
9. the transformer fault diagnosis system according to claim 8 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the fault sample data processing module is used for normalized fault sample data, obtains comprising the characteristic gas
The data set of constituent content, specifically includes:
The characteristic gas includes H2、CH4、C2H2、C2H4And C2H6；The fault type includes hyperthermia and superheating, high-energy discharge, low
It can electric discharge, shelf depreciation and middle cryogenic overheating.
10. the transformer fault diagnosis system according to claim 9 based on Fuzzy C-Means Cluster Algorithm, feature exist
In: the fault sample data processing module is used for normalized fault sample data, obtains comprising the characteristic gas
The data set of constituent content, specifically includes:
The formula of the characteristic gas concentration normalization method is
In formula, XijContent before indicating the characteristic gas normalization, X 'ijContent after representing the characteristic gas normalization,
Each characteristic gas total amount in fault sample data where denominator represents the characteristic gas；
Or
In formula, H2′、CH4′、C2H6′、C2H4' and C2H2' respectively represent the content after each characteristic gas normalization, H2、CH4、
C2H2、C2H4And C2H6Content before each characteristic gas normalization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910468729.4A CN110133410A (en) | 2019-05-31 | 2019-05-31 | Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910468729.4A CN110133410A (en) | 2019-05-31 | 2019-05-31 | Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110133410A true CN110133410A (en) | 2019-08-16 |
Family
ID=67583271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910468729.4A Pending CN110133410A (en) | 2019-05-31 | 2019-05-31 | Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110133410A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111931423A (en) * | 2020-08-12 | 2020-11-13 | 国网吉林省电力有限公司吉林供电公司 | Power transformer fault diagnosis method based on fuzzy proximity of state data |
| CN112149569A (en) * | 2020-09-24 | 2020-12-29 | 福州大学 | Voiceprint fault diagnosis method of transformer based on fuzzy C-means clustering algorithm |
| CN112733878A (en) * | 2020-12-08 | 2021-04-30 | 国网辽宁省电力有限公司锦州供电公司 | Transformer fault diagnosis method based on kmeans-SVM algorithm |
| CN113190728A (en) * | 2021-04-06 | 2021-07-30 | 国网浙江省电力有限公司湖州供电公司 | Oil-immersed transformer fault diagnosis method based on cluster optimization |
| CN114545294A (en) * | 2022-01-14 | 2022-05-27 | 国电南瑞科技股份有限公司 | Transformer fault diagnosis method and system, storage medium and computing device |
| WO2024007580A1 (en) * | 2022-07-07 | 2024-01-11 | 南京国电南自电网自动化有限公司 | Power equipment parallel fault diagnosis method and apparatus based on hybrid clustering |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101587155A (en) * | 2009-06-08 | 2009-11-25 | 浙江大学 | Oil soaked transformer fault diagnosis method |
| CN101701940A (en) * | 2009-10-26 | 2010-05-05 | 南京航空航天大学 | On-line transformer fault diagnosis method based on SVM and DGA |
| CN106596900A (en) * | 2016-12-13 | 2017-04-26 | 贵州电网有限责任公司电力科学研究院 | Transformer fault diagnosis method based on improved semi-supervised classification of graph |
| CN107656154A (en) * | 2017-09-18 | 2018-02-02 | 杭州安脉盛智能技术有限公司 | Based on the Diagnosis Method of Transformer Faults for improving Fuzzy C-Means Cluster Algorithm |
| US20180302011A1 (en) * | 2017-04-18 | 2018-10-18 | General Electric Company | Method for Protecting an Electrical Power System |
| CN109061426A (en) * | 2018-11-02 | 2018-12-21 | 国网河北省电力有限公司电力科学研究院 | Partial discharge of transformer method for diagnosing faults and on-Line Monitor Device |
| CN109596684A (en) * | 2019-01-16 | 2019-04-09 | 苏州天地衡遥感科技有限公司 | A kind of detection of gas with multiple constituents device and method based on RS485 network |
| CN109816031A (en) * | 2019-01-30 | 2019-05-28 | 南京邮电大学 | A kind of Transformer State Assessment clustering method based on the unbalanced measurement of data |
-
2019
- 2019-05-31 CN CN201910468729.4A patent/CN110133410A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101587155A (en) * | 2009-06-08 | 2009-11-25 | 浙江大学 | Oil soaked transformer fault diagnosis method |
| CN101701940A (en) * | 2009-10-26 | 2010-05-05 | 南京航空航天大学 | On-line transformer fault diagnosis method based on SVM and DGA |
| CN106596900A (en) * | 2016-12-13 | 2017-04-26 | 贵州电网有限责任公司电力科学研究院 | Transformer fault diagnosis method based on improved semi-supervised classification of graph |
| US20180302011A1 (en) * | 2017-04-18 | 2018-10-18 | General Electric Company | Method for Protecting an Electrical Power System |
| CN107656154A (en) * | 2017-09-18 | 2018-02-02 | 杭州安脉盛智能技术有限公司 | Based on the Diagnosis Method of Transformer Faults for improving Fuzzy C-Means Cluster Algorithm |
| CN109061426A (en) * | 2018-11-02 | 2018-12-21 | 国网河北省电力有限公司电力科学研究院 | Partial discharge of transformer method for diagnosing faults and on-Line Monitor Device |
| CN109596684A (en) * | 2019-01-16 | 2019-04-09 | 苏州天地衡遥感科技有限公司 | A kind of detection of gas with multiple constituents device and method based on RS485 network |
| CN109816031A (en) * | 2019-01-30 | 2019-05-28 | 南京邮电大学 | A kind of Transformer State Assessment clustering method based on the unbalanced measurement of data |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111931423A (en) * | 2020-08-12 | 2020-11-13 | 国网吉林省电力有限公司吉林供电公司 | Power transformer fault diagnosis method based on fuzzy proximity of state data |
| CN112149569A (en) * | 2020-09-24 | 2020-12-29 | 福州大学 | Voiceprint fault diagnosis method of transformer based on fuzzy C-means clustering algorithm |
| CN112149569B (en) * | 2020-09-24 | 2022-06-14 | 福州大学 | Voiceprint fault diagnosis method of transformer based on fuzzy C-means clustering algorithm |
| CN112733878A (en) * | 2020-12-08 | 2021-04-30 | 国网辽宁省电力有限公司锦州供电公司 | Transformer fault diagnosis method based on kmeans-SVM algorithm |
| CN113190728A (en) * | 2021-04-06 | 2021-07-30 | 国网浙江省电力有限公司湖州供电公司 | Oil-immersed transformer fault diagnosis method based on cluster optimization |
| CN114545294A (en) * | 2022-01-14 | 2022-05-27 | 国电南瑞科技股份有限公司 | Transformer fault diagnosis method and system, storage medium and computing device |
| CN114545294B (en) * | 2022-01-14 | 2023-06-16 | 国电南瑞科技股份有限公司 | Transformer fault diagnosis method, system, storage medium and computing device |
| WO2024007580A1 (en) * | 2022-07-07 | 2024-01-11 | 南京国电南自电网自动化有限公司 | Power equipment parallel fault diagnosis method and apparatus based on hybrid clustering |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110133410A (en) | Diagnosis Method of Transformer Faults and system based on Fuzzy C-Means Cluster Algorithm | |
| CN107656154B (en) | Based on the Diagnosis Method of Transformer Faults for improving Fuzzy C-Means Cluster Algorithm | |
| CN107132310B (en) | Transformer equipment health status method of discrimination based on gauss hybrid models | |
| CN103617568B (en) | Setting method for abnormal data determination threshold in steady-state power quality early-warning mechanism | |
| CN111598150B (en) | Transformer fault diagnosis method considering operation state grade | |
| CN106156485A (en) | Method for diagnosing fault of power transformer and device | |
| CN108268905A (en) | A kind of Diagnosis Method of Transformer Faults and system based on support vector machines | |
| CN108241901A (en) | A kind of transformer Forewarn evaluation method and device based on prediction data | |
| CN103745119A (en) | Oil-immersed transformer fault diagnosis method based on fault probability distribution model | |
| Jiang et al. | Dynamic fault prediction of power transformers based on lasso regression and change point detection by dissolved gas analysis | |
| CN107037306A (en) | Transformer fault dynamic early-warning method based on HMM | |
| CN105242155A (en) | Transformer fault diagnosis method based on entropy weight method and grey correlation analysis | |
| CN110333414A (en) | The multi-level state evaluating method of power transformer | |
| CN109298225B (en) | Automatic identification model system and method for abnormal state of voltage measurement data | |
| CN110223193A (en) | The method of discrimination and system of operation of power networks state are used for based on fuzzy clustering and RS-KNN model | |
| CN111680875A (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
| CN104123678A (en) | Electricity relay protection status overhaul method based on status grade evaluation model | |
| CN109918720A (en) | Diagnosis Method of Transformer Faults based on krill group's Support Vector Machines Optimized | |
| CN113592359A (en) | Health degree evaluation method and device for power transformer | |
| CN108874984A (en) | A kind of increased quality method to second-rate grid equipment defect text | |
| CN107300907B (en) | In conjunction with the flight control system Reliable Evaluating Methods of Their Performance of comprehensive assessment and hypothesis testing | |
| CN109214500A (en) | A kind of transformer fault recognition methods based on integrated intelligent algorithm | |
| CN105137238A (en) | Fault diagnosis system for gas insulation combination electric appliance | |
| Min et al. | Fault prediction for distribution network based on CNN and LightGBM algorithm | |
| CN116755000A (en) | Transformer fault rapid identification device and method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190816 |