CN102944769A - Fault diagnosis method of power transformer based on extreme learning machine - Google Patents
Fault diagnosis method of power transformer based on extreme learning machine Download PDFInfo
- Publication number
- CN102944769A CN102944769A CN2012103968617A CN201210396861A CN102944769A CN 102944769 A CN102944769 A CN 102944769A CN 2012103968617 A CN2012103968617 A CN 2012103968617A CN 201210396861 A CN201210396861 A CN 201210396861A CN 102944769 A CN102944769 A CN 102944769A
- Authority
- CN
- China
- Prior art keywords
- transformer
- learning machine
- extreme learning
- data
- fault 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
Images
Landscapes
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to a fault diagnosis method of a power transformer based on an extreme learning machine, which can be applied to a transformer monitoring/detection device or system. Fault characteristics are extracted based on data collected by the monitoring/detection device or system, and fault diagnosis module leaning of the extreme leaning machine of the transformer is carried out by selecting state samples of the transformer. The method comprises the steps of dividing the operating state of the transformer; selecting monitoring/detection data comprising the operating state of the transformer as a data source; extracting characteristics of the data source of the transformer, and determining characteristic variables; determining target vector expression manner of the extreme learning machine of the transformer in various operating states; selecting sample data of the transformer in various operating states; determining training sample data and testing the sample data; determining an input layer, a hidden layer, node number of an output layer and an excitation function of the fault diagnosis module of the extreme learning machine of the transformer; and learning and verifying the fault diagnosis module of the extreme learning machine of the transformer.
Description
Technical field
The present invention relates to a kind of method of diagnosing fault of power transformer, belong to the technical field of Fault Diagnosis for Electrical Equipment.
Background technology
Transformer is the visual plant of electric system, and its running status directly affects the security of system level.Multiple prison/pick-up unit or the systems such as broadband on-line monitoring of oil chromatography prison/detection (gas content is analyzed DGA in the oil), shelf depreciation prison/detection (pulse current, ultrahigh frequency UHF and ultrasonic method etc.), ground current have appearred comprising at present.A large amount of prisons/detection data are difficult to finish by manual analysis, must seek fast automatically that method for diagnosing faults is embedded in prison/pick-up unit or the system, in order in time find incipient fault and fault type, for the State Maintenance of transformer provides foundation.
In the existing Diagnosis Method of Transformer Faults, the Bayesian network diagnostic method needs the great amount of samples data, and characteristic variable is discrete variable, however discrete threshold values choose and do not have theoretical foundation, and departure process can cause losing of transformer state information; Support vector machine diagnostic method regularization parameter and kernel functional parameter are determined difficulty, transformer fault diagnosis is in the nature many classification problems in addition, and support vector machine is two sorting algorithms, need to be translated into many classification by " one-to-many ", " one to one " or methods such as " binary trees ", have the overlapping and unclassified of classification, need structure than problems such as multi-categorizer, deviation accumulations; The network parameter that Neural Network Diagnosis Method need to be found the solution is many, and need in training process, iteration determine that search volume and calculated amount are very large, must choose suitable learning rate and input initial weight, just desirable result can be obtained, local optimum may be absorbed in addition.
Summary of the invention
The object of the present invention is to provide a kind of Power Transformer Faults intelligent diagnosing method, can be applicable to transformer prison/pick-up unit or system.Extract fault signature on the data basis that prison/pick-up unit or system gather, and choose the extreme learning machine fault diagnosis model study that the transformer state sample carries out transformer, and then realize the fault diagnosis to transformer.
In order to achieve the above object, technical scheme of the present invention provides a kind of method for diagnosing fault of power transformer based on extreme learning machine, and it comprises the following steps:
Step 1: the running status of dividing transformer according to the characteristics that study a question.If the overall operation state of research transformer can be divided into the running status of transformer normal condition and low energy discharge, high-energy discharge, middle cryogenic overheating, hyperthermia and superheating, five kinds of malfunctions of shelf depreciation.If will further study the electric discharge type of the shelf depreciation of transformer, the running status of transformer can be divided into needle point discharge, bubble electric discharge, suspended discharge, four kinds of electric discharge types of creeping discharge.
Step 2: choose the prison of the online or off-line that contains running state of transformer/detection data as data source.The prison of the prison/pick-up unit such as oil chromatography prison/detect (gas content is analyzed DGA in the oil), shelf depreciation (pulse current, ultrahigh frequency UHF and ultrasonic method etc.) or system/inspection data all can be used as data source.
Step 3: the transformer data source is carried out feature extraction, determine characteristic variable, characteristic variable both can be that discrete variable also can be continuous variable.If select oil chromatography to supervise/detect data as data source, characteristic variable can be selected hydrogen (H
2), methane (CH
4), ethane (C
2H
6), ethene (C
2H
4), acetylene (C
2H
2) five kinds of characteristic gas content, or hydrogen (H
2) account for the ratio of total gas content, methane (CH
4), ethane (C
2H
6), ethene (C
2H
4), acetylene (C
2H
2) account for the ratio of total hydrocarbon amount.Supervise/detect data as data source if select ultrahigh frequency method shelf depreciation, characteristic variable can be selected the macroscopic statistics amounts such as measure of skewness, standout, local peaks number, discharge degree of asymmetry, phase place degree of asymmetry, the simple crosscorrelation factor and phase place intermediate value of being extracted by the discharge collection of illustrative plates, and the microscopic feature amounts such as pulse rise time, pulse fall time, pulse width, duration of pulse, 10% amplitude pulse duration.
Step 4: determine the object vector expression way of the extreme learning machine under the various running statuses of transformer, usually adopt 0-1 vector expression way.
Step 5: choose the sample data of transformer under various running statuses.Principle according to representativeness, popularity and compactedness is chosen sample data, in addition for avoiding the data skew problem, the shared ratio of the sample data of choosing under the various running statuses should be close, if the eigenwert diversity ratio of sample data characteristic variable is larger, needs sample data is carried out standardization.
Step 6: sample data is divided into training sample data and test sample book data according to a certain percentage.The training sample data are used for the extreme learning machine fault diagnosis model of study transformer, and the test sample book data are used for the extreme learning machine fault diagnosis model of checking transformer.
Step 7: determine input layer number, the number of hidden nodes, the output layer nodes of the extreme learning machine fault diagnosis model of transformer, choose the excitation function of the extreme learning machine fault diagnosis model of transformer
Step 8: with the input of training sample data as extreme learning machine, carry out the study of the extreme learning machine fault diagnosis model of transformer.
Step 9: adopt the test sample book data that the extreme learning machine Fault Diagnosis Model for Power Transformer that step 8 obtains is verified.
Wherein, described step 8 further comprises the following steps:
Step 82: the hidden layer output matrix that calculates corresponding training sample data
, wherein,
Step 83: adopt least square method to find the solution the hidden layer output matrix
The Moore-Penrose generalized inverse
Described step 9 comprises the following steps:
Step 91: the hidden layer output matrix that calculates corresponding test sample book data
Wherein,
Step 93: with fault type corresponding to greatest member value in the output row vector of extreme learning machine as diagnostic result;
Step 94: the diagnostic result of extreme learning machine and the physical fault type of transformer are made comparisons, calculate rate of correct diagnosis, the extreme learning machine diagnostic model of transformer is verified.
The invention discloses a kind of Power Transformer Faults intelligent diagnosing method based on extreme learning machine in the Fault Diagnosis for Electrical Equipment technical field, can be applicable to transformer prison/pick-up unit or system.Extract fault signature on the data basis that prison/pick-up unit or system gather, and choose the extreme learning machine fault diagnosis model study that the transformer state sample carries out transformer.In this diagnostic method, characteristic variable can be continuous variable, has overcome the discrete threshold values of discrete variable and has determined the problem of difficulty and transformer state information dropout, and only required sample representative; Can directly realize many classification; The study of diagnostic model need not iteration, can obtain global optimum; Diagnosis speed is fast, and accuracy is high.
Description of drawings
Fig. 1 is the process flow diagram of the Diagnosis Method of Transformer Faults based on extreme learning machine of the present invention;
Fig. 2 is the learning process figure of the extreme learning machine fault diagnosis model of transformer among the present invention;
Fig. 3 is the checking process flow diagram of the extreme learning machine fault diagnosis model of transformer among the present invention.
Embodiment
Below in conjunction with accompanying drawing, preferred embodiment is elaborated.Should be emphasized that following explanation only is exemplary, rather than in order to limit the scope of the invention and to use.
Fig. 1 is based on the Diagnosis Method of Transformer Faults process flow diagram of extreme learning machine.Among Fig. 1, the Diagnosis Method of Transformer Faults based on extreme learning machine that the present invention proposes comprises the following steps:
Step 1: the running status of dividing transformer.With the running status of transformer be divided into normally, low energy discharge, high-energy discharge, middle cryogenic overheating, hyperthermia and superheating, six kinds of running statuses of shelf depreciation, wherein rear five kinds of states are dissimilar failure operation states.
Step 2: choose the prison of the online or off-line that contains running state of transformer/detection data as data source.Select hydrogen (H
2), methane (CH
4), ethane (C
2H
6), ethene (C
2H
4), acetylene (C
2H
2) prison/detection data of five kinds of characteristic gas content is as data source, this data source can be supervised by oil chromatography/and pick-up unit obtains by online or off-line prison/detect.
Step 3: the transformer data source is carried out feature extraction, determine characteristic variable.Choose H
2Account for the ratio of total gas content, be designated as
x 1, CH
4, C
2H
6, C
2H
4, C
2H
2Account for the ratio of total hydrocarbon amount, be designated as respectively
x 2 , x 3 , x 4 , x 5, as characteristic variable, the pattern of characteristic variable is: [
x 1 x 2 x 3 x 4 x 5].Characteristic variable is continuous variable, does not therefore exist discrete threshold values to choose difficulty and the problem of losing transformer state information.
Step 4: the object vector expression way of determining the extreme learning machine under the various running statuses of transformer.The object vector of the extreme learning machine fault diagnosis model of transformer adopts the vector form of 6 dimension 0-1 to express, that is, adopt respectively [0,0,0,0,0,1]
T, [0,0,0,0,1,0]
T, [0,0,0,1,0,0]
T, [0,0,1,0,0,0]
T, [0,1,0,0,0,0]
T, [1,0,0,0,0,0]
TNormal, the low energy discharge of indication transformer, high-energy discharge, middle cryogenic overheating, hyperthermia and superheating, six kinds of running statuses of shelf depreciation.Because the object vector of extreme learning machine fault diagnosis model adopts aforesaid way to express the running status of transformer, so the diagnosis of the extreme learning machine fault diagnosis model of transformer is output as the form of 6 dimensional vectors.If in the sextuple element value maximum of sextuple element then diagnostic result be normal, if the value maximum of the 5th dimension element then diagnostic result is the low energy discharge, if the value maximum of fourth dimension element then diagnostic result is high-energy discharge, if the value maximum of third dimension element then diagnostic result is middle cryogenic overheating, if second the dimension element the value maximum then diagnostic result be hyperthermia and superheating, if first the dimension element the value maximum then diagnostic result be shelf depreciation.The form of row vector is adopted in diagnosis output.
Step 5: raw data is screened, classified, carry out characteristic and extract, choose the transformer sample data.Choose the sample data of transformer under normal, low energy discharge, high-energy discharge, middle cryogenic overheating, hyperthermia and superheating, six kinds of states of shelf depreciation.
Step 6: respectively the sample data of transformer under six kinds of running statuses is divided into training sample data collection and test sample book data set in the ratio of about 3:2.Described training sample data collection
, wherein
Be the dimension of proper vector,
Be the status number of the transformer state divided,
Sample number for training sample; Described test sample book data set
, wherein
Sample number for test sample book.
Step 7: determine input layer number, the number of hidden nodes, the output layer nodes of the fault diagnosis extreme learning machine model of transformer, choose the excitation function of extreme learning machine
The input layer number is the dimension of proper vector, the proper vector pattern be [
x 1 x 2 x 3 x 4 x 5], be 5 dimensions, so the input layer number is 5.The output layer nodes is the status number of the running state of transformer of division, and the running status of transformer is divided into 6 kinds, so the output layer nodes is 6.The hidden node number is not well-determined, for example can elect 2 times of dimension of proper vector namely 10 as.Excitation function is selected the sigmoidal function:
Step 8: with the input of training sample data as extreme learning machine, carry out the study of the extreme learning machine fault diagnosis model of transformer.
Process flow diagram referring to the study of the extreme learning machine fault diagnosis model of transformer shown in Figure 2 further comprises the following steps: in the step 8
Step 81: random assignment input weight vectors
And biasing
,
,
Be the number of hidden nodes.
And biasing
Any value.
Step 82: the hidden layer output matrix that calculates corresponding training sample data
,
Wherein,
Step 83: adopt least square method to find the solution the hidden layer output matrix
The Moore-Penrose generalized inverse
Be well-determined, directly tried to achieve by least square method that need not iteration, it is fast to find the solution speed, can not be absorbed in local optimum.
Step 9: adopt the test sample book data that the extreme learning machine Fault Diagnosis Model for Power Transformer that step 8 obtains is verified.
Checking process flow diagram referring to the extreme learning machine fault diagnosis model of transformer shown in Figure 3 further comprises the following steps: in the step 9
Step 91: the hidden layer output matrix that calculates corresponding transformer testing sample data
,
Wherein,
Step 92: the output of the extreme learning machine fault diagnosis model of the transformer of calculating test sample book
Step 93: as diagnostic result, directly realization is diagnosed, and need not the transformer fault diagnosis problem is converted into a plurality of two classification problems with fault type corresponding to greatest member value in the capable vector of extreme learning machine output.
Step 94: the diagnostic result of the extreme learning machine of test sample book and the physical fault type of transformer are made comparisons, calculate rate of correct diagnosis, the extreme learning machine diagnostic model of transformer is verified.
The above; only for the better embodiment of the present invention, but protection scope of the present invention is not limited to this, anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
For example, in a further embodiment, if will further study the electric discharge type of the shelf depreciation of transformer, then can in the correlation step of the method for the invention, make to adapt to and revise, be summarized as follows:
Step 1: the running status of transformer is divided into needle point discharge, bubble electric discharge, suspended discharge, four kinds of electric discharge types of creeping discharge.
Step 2: choose the prison of shelf depreciation (pulse current, ultrahigh frequency UHF and ultrasonic method etc.) prison/pick-up unit or system/detection data as data source.
Step 3: if select ultrahigh frequency method shelf depreciation to supervise/detect data as data source, characteristic variable can be selected the macroscopic statistics amounts such as measure of skewness, standout, local peaks number, discharge degree of asymmetry, phase place degree of asymmetry, the simple crosscorrelation factor and phase place intermediate value of being extracted by the discharge collection of illustrative plates, and the microscopic feature amounts such as pulse rise time, pulse fall time, pulse width, duration of pulse, 10% amplitude pulse duration.Characteristic variable both can be that discrete variable also can be continuous variable.
Step 4: determine the object vector expression way of the extreme learning machine under the various running statuses of transformer, then adopted the 0-1 vector expression way of 4 dimensions according to four kinds of electric discharge types.
Step 5: choose the sample data of transformer under various running statuses.Specifically choose sample data according to the principle of representativeness, popularity and compactedness, in addition for avoiding the data skew problem, the shared ratio of the sample data of choosing under the various running statuses should be close, if the eigenwert diversity ratio of sample data characteristic variable is larger, need sample data is carried out standardization.
Step 6: sample data is divided into training sample data and test sample book data according to a certain percentage.The training sample data are used for the extreme learning machine fault diagnosis model of study transformer, and the test sample book data are used for the extreme learning machine fault diagnosis model of checking transformer.
Step 7: the input layer number of determining the extreme learning machine fault diagnosis model of transformer is the dimension of proper vector in the step 3, determines that the output layer nodes is the electric discharge type several 4 in the step 1; Also need to choose the number of hidden nodes, and excitation function
, specifically can be with reference to the description among the embodiment above.
Step 8: with reference to the description among the embodiment above, with the input of training sample data as extreme learning machine, carry out the study of the extreme learning machine fault diagnosis model of transformer.
Step 9: with reference to the description among the embodiment above, adopt the test sample book data that the extreme learning machine Fault Diagnosis Model for Power Transformer that step 8 obtains is verified.
In sum, the present invention utilizes the collected prison of mounted prison/pick-up unit or system/detection data, foundation is based on the intelligent transformer diagnostic system of extreme learning machine, can in time find transformer fault and fault type, for Repair of Transformer plan and localization of fault provide foundation, in time to fix a breakdown.This transformer diagnostic method characteristic variable both can be that discrete variable also can be continuous variable; Can directly realize many classification, need not transformer fault diagnosis is converted into a plurality of two classification problems; The study of diagnostic model only need arrange suitable node numbers of hidden layers, carries out random assignment for inputting weight and deviation, then exports weight and obtains by least square method, whole process is once finished, and need not iteration, does not have the problem of local optimum, diagnosis speed is fast, and accuracy is high.
Claims (6)
1. the method for diagnosing fault of power transformer based on extreme learning machine is characterized in that, described method comprises following steps:
Step 1: according to the characteristics that transformer studies a question, A the running status of dividing transformer;
Step 2: choose the prison of the online or off-line that contains running state of transformer/detection data as data source;
Step 3: the transformer data source is carried out feature extraction, determine several characteristic variables, characteristic variable both can be that discrete variable also can be continuous variable;
Step 4: adopt the 0-1 vector expression way of A dimension, determine the object vector of the extreme learning machine under the various running statuses of transformer;
Step 5: choose the sample data of transformer under various running statuses;
Step 6: sample data is divided into training sample data and test sample book data in the ratio of setting; Wherein, the training sample data are used for the extreme learning machine fault diagnosis model of study transformer, and the test sample book data are used for the extreme learning machine fault diagnosis model of checking transformer;
Step 7: determine input layer number, the number of hidden nodes, the output layer nodes of the extreme learning machine fault diagnosis model of transformer, choose the excitation function of the extreme learning machine fault diagnosis model of transformer
Step 8: with the input of training sample data as extreme learning machine, carry out the study of the extreme learning machine fault diagnosis model of transformer;
Step 9: adopt the test sample book data that the extreme learning machine Fault Diagnosis Model for Power Transformer that step 8 obtains is verified.
?
2. the method for claim 1 is characterized in that, described method is in order to study the overall operation state of transformer, then
In the step 1, the running status of transformer is divided into: normal condition, and low energy discharge, high-energy discharge, middle cryogenic overheating, hyperthermia and superheating, five kinds of malfunctions of shelf depreciation;
In the step 2, will supervise via oil chromatography/oil that pick-up unit or system obtain in gas content analyze data as data source;
In the step 3, characteristic variable is characteristic gas content or characteristic gas content ratio.
?
3. the method for claim 1 is characterized in that, described method is in order to the electric discharge type of the shelf depreciation of studying transformer, then
In the step 1, the running status of transformer is divided into: needle point discharge, bubble electric discharge, suspended discharge, four kinds of electric discharge types of creeping discharge;
In the step 2, the prison that will obtain based on shelf depreciation prison/pick-up unit or the system of pulse current, ultrahigh frequency UHF or ultrasonic method/detection data are as data source;
In the step 3, characteristic variable is macroscopic statistics amount or microscopic feature amount, wherein, described macroscopic statistics amount comprises one or more in the following variable: measure of skewness, standout, local peaks number, discharge degree of asymmetry, phase place degree of asymmetry, the simple crosscorrelation factor and the phase place intermediate value extracted by the discharge collection of illustrative plates;
Described microscopic feature amount comprises one or more in the following variable: pulse rise time, pulse fall time, pulse width, duration of pulse, 10% amplitude pulse duration.
?
4. such as claim 1 or 2 or 3 described methods, it is characterized in that,
In the step 7, described input layer number is the dimension of proper vector in the step 3, and described output layer nodes is that the running status in the step 1 is counted A; The hidden node number is 2 times of dimension of proper vector; Excitation function is selected the sigmoidal function:
?
5. method as claimed in claim 4 is characterized in that, described step 8 further comprises the following steps:
Wherein,
Step 83: adopt least square method to find the solution the hidden layer output matrix
The Moore-Penrose generalized inverse
?
6. method as claimed in claim 4 is characterized in that, described step 9 further comprises the following steps:
Step 93: with running status corresponding to greatest member value in the output row vector of extreme learning machine as diagnostic result;
Step 94: the diagnostic result of extreme learning machine and the actual motion state of transformer are made comparisons, calculate rate of correct diagnosis, the extreme learning machine diagnostic model of transformer is verified.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012103968617A CN102944769A (en) | 2012-10-18 | 2012-10-18 | Fault diagnosis method of power transformer based on extreme learning machine |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012103968617A CN102944769A (en) | 2012-10-18 | 2012-10-18 | Fault diagnosis method of power transformer based on extreme learning machine |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102944769A true CN102944769A (en) | 2013-02-27 |
Family
ID=47727729
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012103968617A Pending CN102944769A (en) | 2012-10-18 | 2012-10-18 | Fault diagnosis method of power transformer based on extreme learning machine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102944769A (en) |
Cited By (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103336231A (en) * | 2013-07-01 | 2013-10-02 | 杭州电子科技大学 | Neural network method for detecting AE position of partial discharge |
CN103364669A (en) * | 2013-07-31 | 2013-10-23 | 广州供电局有限公司 | Online detecting method and system for GIS (Gas Insulated Switchgear) device operating state |
CN103472343A (en) * | 2013-09-29 | 2013-12-25 | 国家电网公司 | High voltage equipment state predicting method |
CN103941156A (en) * | 2014-04-16 | 2014-07-23 | 华北电力大学 | Multi-message fusion section locating method based on extreme learning machine |
CN104034794A (en) * | 2014-06-12 | 2014-09-10 | 东北大学 | Extreme learning machine-based pipeline magnetic flux leakage defect detection method |
CN104616030A (en) * | 2015-01-21 | 2015-05-13 | 北京工业大学 | Extreme learning machine algorithm-based recognition method |
CN104777410A (en) * | 2015-04-22 | 2015-07-15 | 东北电力大学 | Partial discharge pattern identification method for crosslinked polyethylene cable |
CN104809328A (en) * | 2014-10-09 | 2015-07-29 | 许继电气股份有限公司 | Transformer fault diagnosis method based on information bottleneck |
CN105335759A (en) * | 2015-11-12 | 2016-02-17 | 南方电网科学研究院有限责任公司 | Generating-probability-model-based transformer fault detection method |
CN105606914A (en) * | 2015-09-06 | 2016-05-25 | 南京航空航天大学 | IWO-ELM-based Aviation power converter fault diagnosis method |
CN106199305A (en) * | 2016-07-01 | 2016-12-07 | 太原理工大学 | Underground coal mine electric power system dry-type transformer insulation health state evaluation method |
CN106228184A (en) * | 2016-07-19 | 2016-12-14 | 东北大学 | A kind of based on the blast furnace fault detection system and the method that optimize extreme learning machine |
CN106295044A (en) * | 2016-08-18 | 2017-01-04 | 湖南工业大学 | A kind of heavy loading locomotive tacky state recognition methods based on extreme learning machine |
CN106707221A (en) * | 2017-01-05 | 2017-05-24 | 云南电网有限责任公司电力科学研究院 | Fault diagnosis method and system for sensor in electric energy metering device |
CN107271829A (en) * | 2017-05-09 | 2017-10-20 | 安徽继远软件有限公司 | A kind of controller switching equipment running state analysis method and device |
CN107563251A (en) * | 2016-07-01 | 2018-01-09 | 华北电力大学(保定) | Fault Diagnosis of Fan method based on extreme learning machine |
CN107884663A (en) * | 2017-10-27 | 2018-04-06 | 国网天津市电力公司电力科学研究院 | A kind of Diagnosis Method of Transformer Faults based on combination core Method Using Relevance Vector Machine |
CN108895195A (en) * | 2018-07-23 | 2018-11-27 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN109001602A (en) * | 2018-08-20 | 2018-12-14 | 广东工业大学 | Shelf depreciation Severity method based on extreme learning machine algorithm |
CN109144039A (en) * | 2018-11-04 | 2019-01-04 | 兰州理工大学 | A kind of batch process fault detection method keeping extreme learning machine based on timing extension and neighborhood |
CN109298258A (en) * | 2018-09-18 | 2019-02-01 | 四川大学 | In conjunction with the Diagnosis Method of Transformer Faults and system of RVM and DBN |
CN109683066A (en) * | 2018-11-08 | 2019-04-26 | 山东康威通信技术股份有限公司 | Power cable typical defect local discharge signal recognition methods |
CN109685138A (en) * | 2018-12-25 | 2019-04-26 | 东南大学 | A kind of XLPE power cable shelf depreciation kind identification method |
CN110824384A (en) * | 2019-11-20 | 2020-02-21 | 国家电网有限公司 | Transformer fault online diagnosis system based on artificial intelligence extreme learning machine |
CN111025041A (en) * | 2019-11-07 | 2020-04-17 | 深圳供电局有限公司 | Electric vehicle charging pile monitoring method and system, computer equipment and medium |
CN111783531A (en) * | 2020-05-27 | 2020-10-16 | 福建亿华源能源管理有限公司 | Water turbine set fault diagnosis method based on SDAE-IELM |
CN112748372A (en) * | 2020-12-21 | 2021-05-04 | 湘潭大学 | Transformer fault diagnosis method of artificial bee colony optimization extreme learning machine |
CN113341347A (en) * | 2021-06-02 | 2021-09-03 | 云南大学 | Dynamic fault detection method for distribution transformer based on AOELM |
CN116821591A (en) * | 2023-04-04 | 2023-09-29 | 浙江万能弹簧机械有限公司 | Real-time monitoring method and system for discharge condition of high-frequency power supply |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN102735760A (en) * | 2012-06-26 | 2012-10-17 | 河海大学 | Method for predicting transformer oil chromatographic data based on extreme learning machine |
-
2012
- 2012-10-18 CN CN2012103968617A patent/CN102944769A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
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 |
CN102735760A (en) * | 2012-06-26 | 2012-10-17 | 河海大学 | Method for predicting transformer oil chromatographic data based on extreme learning machine |
Non-Patent Citations (3)
Title |
---|
HUIXIN TIAN: "Multiple ANNs combined scheme for fault diagnosis of power transformers", 《CONTROL AND DECISION CONFERENCE (CCDC), 2011 CHINESE》 * |
张哲: "用SVM和LS-SVM分析变压器故障诊断", 《信息化纵横》 * |
王彩雄: "局部放电特高频检测抗干扰与诊断技术的研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (46)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103336231A (en) * | 2013-07-01 | 2013-10-02 | 杭州电子科技大学 | Neural network method for detecting AE position of partial discharge |
CN103364669B (en) * | 2013-07-31 | 2016-04-20 | 广州供电局有限公司 | GIS equipment operational condition online test method and system |
CN103364669A (en) * | 2013-07-31 | 2013-10-23 | 广州供电局有限公司 | Online detecting method and system for GIS (Gas Insulated Switchgear) device operating state |
CN103472343A (en) * | 2013-09-29 | 2013-12-25 | 国家电网公司 | High voltage equipment state predicting method |
CN103941156A (en) * | 2014-04-16 | 2014-07-23 | 华北电力大学 | Multi-message fusion section locating method based on extreme learning machine |
CN103941156B (en) * | 2014-04-16 | 2016-06-22 | 华北电力大学 | Multi-information acquisition Section Location based on extreme learning machine |
CN104034794A (en) * | 2014-06-12 | 2014-09-10 | 东北大学 | Extreme learning machine-based pipeline magnetic flux leakage defect detection method |
CN104034794B (en) * | 2014-06-12 | 2017-01-04 | 东北大学 | A kind of pipe leakage defect inspection method based on extreme learning machine |
CN104809328A (en) * | 2014-10-09 | 2015-07-29 | 许继电气股份有限公司 | Transformer fault diagnosis method based on information bottleneck |
CN104616030A (en) * | 2015-01-21 | 2015-05-13 | 北京工业大学 | Extreme learning machine algorithm-based recognition method |
CN104616030B (en) * | 2015-01-21 | 2019-03-29 | 北京工业大学 | A kind of recognition methods based on extreme learning machine algorithm |
CN104777410A (en) * | 2015-04-22 | 2015-07-15 | 东北电力大学 | Partial discharge pattern identification method for crosslinked polyethylene cable |
CN105606914A (en) * | 2015-09-06 | 2016-05-25 | 南京航空航天大学 | IWO-ELM-based Aviation power converter fault diagnosis method |
CN105335759A (en) * | 2015-11-12 | 2016-02-17 | 南方电网科学研究院有限责任公司 | Generating-probability-model-based transformer fault detection method |
CN105335759B (en) * | 2015-11-12 | 2019-07-16 | 南方电网科学研究院有限责任公司 | A kind of transformer fault detection method based on generating probability model |
CN106199305B (en) * | 2016-07-01 | 2018-12-28 | 太原理工大学 | Underground coal mine power supply system dry-type transformer insulation health state evaluation method |
CN106199305A (en) * | 2016-07-01 | 2016-12-07 | 太原理工大学 | Underground coal mine electric power system dry-type transformer insulation health state evaluation method |
CN107563251A (en) * | 2016-07-01 | 2018-01-09 | 华北电力大学(保定) | Fault Diagnosis of Fan method based on extreme learning machine |
CN107563251B (en) * | 2016-07-01 | 2021-11-09 | 华北电力大学(保定) | Fan fault diagnosis method based on extreme learning machine |
CN106228184B (en) * | 2016-07-19 | 2019-08-06 | 东北大学 | A kind of blast furnace fault detection method based on optimization extreme learning machine |
CN106228184A (en) * | 2016-07-19 | 2016-12-14 | 东北大学 | A kind of based on the blast furnace fault detection system and the method that optimize extreme learning machine |
CN106295044B (en) * | 2016-08-18 | 2019-08-13 | 湖南工业大学 | A kind of heavy loading locomotive tacky state recognition methods based on extreme learning machine |
CN106295044A (en) * | 2016-08-18 | 2017-01-04 | 湖南工业大学 | A kind of heavy loading locomotive tacky state recognition methods based on extreme learning machine |
CN106707221B (en) * | 2017-01-05 | 2019-09-06 | 云南电网有限责任公司电力科学研究院 | Sensor fault diagnosis method and system in a kind of electric energy metering device |
CN106707221A (en) * | 2017-01-05 | 2017-05-24 | 云南电网有限责任公司电力科学研究院 | Fault diagnosis method and system for sensor in electric energy metering device |
CN107271829A (en) * | 2017-05-09 | 2017-10-20 | 安徽继远软件有限公司 | A kind of controller switching equipment running state analysis method and device |
CN107884663A (en) * | 2017-10-27 | 2018-04-06 | 国网天津市电力公司电力科学研究院 | A kind of Diagnosis Method of Transformer Faults based on combination core Method Using Relevance Vector Machine |
CN108895195A (en) * | 2018-07-23 | 2018-11-27 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN108895195B (en) * | 2018-07-23 | 2019-11-26 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN109001602B (en) * | 2018-08-20 | 2020-07-28 | 广东工业大学 | Partial discharge severity evaluation method based on extreme learning machine algorithm |
CN109001602A (en) * | 2018-08-20 | 2018-12-14 | 广东工业大学 | Shelf depreciation Severity method based on extreme learning machine algorithm |
CN109298258A (en) * | 2018-09-18 | 2019-02-01 | 四川大学 | In conjunction with the Diagnosis Method of Transformer Faults and system of RVM and DBN |
CN109144039A (en) * | 2018-11-04 | 2019-01-04 | 兰州理工大学 | A kind of batch process fault detection method keeping extreme learning machine based on timing extension and neighborhood |
CN109144039B (en) * | 2018-11-04 | 2021-06-22 | 兰州理工大学 | Intermittent process fault detection method based on time sequence expansion and neighborhood preserving extreme learning machine |
CN109683066A (en) * | 2018-11-08 | 2019-04-26 | 山东康威通信技术股份有限公司 | Power cable typical defect local discharge signal recognition methods |
CN109685138A (en) * | 2018-12-25 | 2019-04-26 | 东南大学 | A kind of XLPE power cable shelf depreciation kind identification method |
CN109685138B (en) * | 2018-12-25 | 2023-04-07 | 东南大学 | XLPE power cable partial discharge type identification method |
CN111025041A (en) * | 2019-11-07 | 2020-04-17 | 深圳供电局有限公司 | Electric vehicle charging pile monitoring method and system, computer equipment and medium |
CN110824384A (en) * | 2019-11-20 | 2020-02-21 | 国家电网有限公司 | Transformer fault online diagnosis system based on artificial intelligence extreme learning machine |
CN111783531A (en) * | 2020-05-27 | 2020-10-16 | 福建亿华源能源管理有限公司 | Water turbine set fault diagnosis method based on SDAE-IELM |
CN111783531B (en) * | 2020-05-27 | 2024-03-19 | 福建亿华源能源管理有限公司 | Water turbine set fault diagnosis method based on SDAE-IELM |
CN112748372A (en) * | 2020-12-21 | 2021-05-04 | 湘潭大学 | Transformer fault diagnosis method of artificial bee colony optimization extreme learning machine |
CN113341347A (en) * | 2021-06-02 | 2021-09-03 | 云南大学 | Dynamic fault detection method for distribution transformer based on AOELM |
CN113341347B (en) * | 2021-06-02 | 2022-05-03 | 云南大学 | Dynamic fault detection method for distribution transformer based on AOELM |
CN116821591A (en) * | 2023-04-04 | 2023-09-29 | 浙江万能弹簧机械有限公司 | Real-time monitoring method and system for discharge condition of high-frequency power supply |
CN116821591B (en) * | 2023-04-04 | 2024-03-08 | 浙江万能弹簧机械有限公司 | Real-time monitoring method and system for discharge condition of high-frequency power supply |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102944769A (en) | Fault diagnosis method of power transformer based on extreme learning machine | |
CN109738776A (en) | Fan converter open-circuit fault recognition methods based on LSTM | |
CN101701940B (en) | On-line transformer fault diagnosis method based on SVM and DGA | |
EP3460611B1 (en) | System and method for aircraft fault detection | |
CN102435910B (en) | Power electronic circuit health monitoring method based on support vector classification | |
Smarsly et al. | Machine learning techniques for structural health monitoring | |
CN104753461B (en) | Method for diagnosing and classifying faults of photovoltaic power generation arrays on basis of particle swarm optimization support vector machines | |
Danilczyk et al. | Smart grid anomaly detection using a deep learning digital twin | |
CN109460618A (en) | A kind of rolling bearing remaining life on-line prediction method and system | |
CN108268905A (en) | A kind of Diagnosis Method of Transformer Faults and system based on support vector machines | |
CN110414155A (en) | A kind of detection of fan part temperature anomaly and alarm method with single measuring point | |
Zhao et al. | Hierarchical anomaly detection and multimodal classification in large-scale photovoltaic systems | |
CN105846780A (en) | Decision tree model-based photovoltaic assembly fault diagnosis method | |
GB2476246A (en) | Diagnosing an operation mode of a machine | |
CN102222151B (en) | Analog circuit fault prediction method based on ARMA (Autoregressive Moving Average) | |
CN109165604A (en) | The recognition methods of non-intrusion type load and its test macro based on coorinated training | |
CN103103570B (en) | Based on the aluminium cell condition diagnostic method of pivot similarity measure | |
CN109617526A (en) | A method of photovoltaic power generation array fault diagnosis and classification based on wavelet multiresolution analysis and SVM | |
CN110580492A (en) | Track circuit fault precursor discovery method based on small fluctuation detection | |
Ndjakomo Essiane et al. | Faults detection and identification in PV array using kernel principal components analysis | |
Kim et al. | Anomaly detection using clustered deep one-class classification | |
CN207992717U (en) | A kind of gate of hydropower station on-line condition monitoring system | |
CN103675356A (en) | Anemometer fault detection method and system on the basis of particle swarm optimization | |
Høiem et al. | Comparative study of event prediction in power grids using supervised machine learning methods | |
CN109116833A (en) | Based on improvement drosophila-bat algorithm mechanical failure diagnostic 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 | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130227 |