CN111695288B - Transformer fault diagnosis method based on Apriori-BP algorithm - Google Patents
Transformer fault diagnosis method based on Apriori-BP algorithm Download PDFInfo
- Publication number
- CN111695288B CN111695288B CN202010374833.XA CN202010374833A CN111695288B CN 111695288 B CN111695288 B CN 111695288B CN 202010374833 A CN202010374833 A CN 202010374833A CN 111695288 B CN111695288 B CN 111695288B
- Authority
- CN
- China
- Prior art keywords
- transformer
- gas
- apriori
- neural network
- algorithm
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/50—Testing of electric apparatus, lines, cables or components for short-circuits, continuity, leakage current or incorrect line connections
- G01R31/62—Testing of transformers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Power Engineering (AREA)
- Investigating Or Analyzing Non-Biological Materials By The Use Of Chemical Means (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention discloses a transformer fault diagnosis method based on an Apriori-BP algorithm, which is characterized in that a BP neural network is trained by utilizing characteristic vector data of gas in transformer oil and corresponding transformer running states to obtain an Apriori-BP prediction model; and inputting the characteristic vector data of the gas in the transformer oil in real time into an Aprori-BP prediction model to obtain the prediction of the state of the transformer. According to the transformer fault diagnosis method based on the Apriori-BP algorithm, the potential relation among the eigenvectors is mined by using the Apriori algorithm to obtain the initial weight of the neural network, so that the defects of over-fitting and easy local optimization caused by unreasonable initialization values in the BP neural network are effectively overcome, the accuracy is effectively improved, and the method has higher practical use value.
Description
Technical Field
The invention relates to a transformer fault diagnosis method based on an Apriori-BP algorithm, and belongs to the technical field of transformer fault detection.
Background
Currently, transformers are the core equipment of an electrical power system. The safe operation of the transformer is not only related to the operation cost and the safety of the power system, but also related to the normal operation of all communities and the safety of lives and properties of people to a great extent. Through the fault diagnosis technology, faults can be found in advance in the running process of the transformer, measures can be taken in time, and accidents are reduced.
The existing transformer fault diagnosis method comprises a three-ratio method (IEC), a cluster analysis method, a support vector machine, an artificial neural network algorithm and the like. In the diagnosis and prediction of transformer faults, most of the algorithms adopt single or small fault feature quantities for analysis and calculation, and the common influence of various faults and possible links among the faults are not considered, so that the accuracy of prediction results is low.
Disclosure of Invention
The purpose is as follows: in order to overcome the defects in the prior art, the invention provides a transformer fault diagnosis method based on an Apriori-BP algorithm.
The technical scheme is as follows: in order to solve the technical problems, the invention adopts the following technical scheme:
a transformer fault diagnosis method based on an Apriori-BP algorithm comprises the following steps:
training the BP neural network by utilizing characteristic vector data of gas in transformer oil and a corresponding transformer running state to obtain an apreri-BP prediction model;
and inputting the characteristic vector data of the gas in the transformer oil in real time into an Aprori-BP prediction model to obtain the prediction of the state of the transformer.
Preferably, the step of acquiring the characteristic vector data of the gas in the transformer oil comprises the following steps:
obtaining the dissolved amount and concentration values of a plurality of gases in the transformer oil and corresponding different running states of the transformer;
taking one sample of the dissolved amount and concentration value of each gas under different operation states of the transformer, and carrying out normalization treatment according to the gas types to obtain the characteristic vector data of each gas under different operation states of the transformer.
As a preferable scheme, the formula for normalizing each gas dissolution amount and concentration value is as follows:
wherein:is the ith dimension characteristic value, x of the jth data i max And x i min The maximum and minimum values of the i-th dimension data in the dataset are respectively, i e {0,1}, and two dimensions 0 and 1 represent the dissolution and concentration values respectively, j e { 1..and l }, where l represents the number of training samples.
As a preferred scheme, the BP neural network comprises an input layer, a hidden layer and an output layer, the number of the collected gas types is set to be the number of input nodes of the BP neural network input layer, the dimension of each node is 2, and the dimension is respectively characteristic vector data corresponding to the dissolution quantity and concentration value; setting the number of kinds of running states of the transformer as the number of output nodes of the BP neural network output layer; calculating the node number of the hidden layer h by using an empirical formula:
or->
Where α is a constant between 1 and 10, input represents the number of input nodes and output represents the number of output nodes.
As a preferred scheme, the initial weight obtaining step of the apriori-BP prediction model is as follows:
the characteristic vector set of the dissolved amount and concentration value of various gases when the transformer operates is defined as X, and the characteristic vector set of various operating states of the transformer is defined as Y, namely:
X n = { n kinds of eigenvectors }
Y m = { m kinds of transformer operation states }
The confidence value may be expressed as an association rule x=>Y represents a conditional probability value of occurrence of Y on the premise of occurrence of X; conf XY The relation between the dissolved quantity and concentration value vector X of various gases in the transformer gasoline at a certain moment and the running state Y of the transformer is reflected; wherein the method comprises the steps of
Wherein, freq (·) is the frequency of occurrence of events in the dataset, i.e., freq (Y) represents the frequency of occurrence of events in the Y state, freq (X n Y) represents the frequency of occurrence of events that are satisfied simultaneously by X, Y;
a weight calculation formula derived from the corresponding operating condition type of each gas,
wherein the method comprises the steps of，ω ij For the weight of j class running states when the dissolution content of i class gas in transformer oil exceeds standard, conf ij Confidence coefficient of j-class running state when i-class gas dissolution content in transformer oil exceeds standard, conf nj And the confidence coefficient of j operating states when the dissolution content of n types of gases in the transformer oil exceeds the standard is obtained.
Preferably, the method further comprises the following steps:
and repeatedly carrying out forward calculation and error back propagation on the Aprori-BP prediction model, and adjusting the weight.
Preferably, the method further comprises the following steps:
and adjusting the Aprori-BP prediction model through prediction accuracy evaluation.
Preferably, the transformer oil comprises at least one of several gases: h 2 ，CH 4 ，C 2 H 6 ，C 2 H 4 And C 2 H 2 。
The beneficial effects are that: according to the transformer fault diagnosis method based on the Apriori-BP algorithm, the potential relation among the eigenvectors is mined by using the Apriori algorithm to obtain the initial weight of the neural network, so that the defects of over-fitting and easy local optimization caused by unreasonable initialization values in the BP neural network are effectively overcome, the accuracy is effectively improved, and the method has higher practical use value.
Drawings
FIG. 1 is a block diagram of a BP neural network model;
FIG. 2 is a general flow chart of the present invention;
FIG. 3 is a graph comparing predicted and measured values of the Apriori-BP model of the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings.
A transformer fault diagnosis method based on an Apriori-BP neural network algorithm comprises the following steps:
(1) And obtaining the dissolved amount and concentration values of a plurality of gases in the transformer oil and corresponding different running states of the transformer. The invention will collect H 2 ，CH 4 ，C 2 H 6 ，C 2 H 4 And C 2 H 2 The dissolution and concentration values of the five gases in the transformer oil under different operation states of the transformer.
(2) Taking one sample of the dissolved amount and concentration value of each gas under different operation states of the transformer, and carrying out normalization treatment according to the gas types to obtain the characteristic vector data of each gas under different operation states of the transformer. The normalization method is respectively standardized from two dimensions of solubility and concentration to eliminate the influence caused by different dimensions of the two dimensional characteristics;
further, in the step (2), the formula for normalizing the dissolved amount and concentration value of each gas is as follows:
wherein:is the ith dimension characteristic value, x of the jth data i max And x i min The maximum and minimum values of the i-th dimension data in the dataset are respectively, i e {0,1}, and two dimensions 0 and 1 represent the dissolution and concentration values respectively, j e { 1..and l }, where l represents the number of training samples.
(3) Determining a specific architecture of a BP neural network;
further, the step (3) specifically includes:
(31) The number of the collected gas types is set to be the number of input nodes of the BP neural network input layer, the dimension of each node is 2, and the dimension is respectively characteristic vector data corresponding to the dissolution quantity and the concentration value. The invention adopts five gas types in the step 1, and sets 5 input nodes.
(32) The number of kinds of the operation states of the transformer is set to be the number of output nodes of the BP neural network output layer. The operating states of a transformer include a fault state and a normal state, and the fault state can be generally classified into two types: the overheat fault and the discharge fault can be divided into two conditions of medium-low temperature overheat and high-temperature overheat according to the temperature; the discharge faults can be divided into two types of low-energy discharge and high-energy discharge according to different energy densities, and the total of the two types of the discharge faults is five running states in the normal state.
(33) In order to reasonably select the node number of the hidden layer h, the node number of the hidden layer h is generally determined by an empirical formula:
or->
Where α is a constant between 1 and 10, input represents the number of input nodes and output represents the number of output nodes. Here, we determine the number of hidden layer nodes h=5 according to an empirical formula, and the general BP network structure is shown in fig. 1, so that the neural network architecture is built up.
(4) Determining the initial weight of the BP neural network through an Apriori algorithm;
further, the step (4) includes:
(41) And according to the experimental sample data, analyzing the relation between the characteristic vector data and the running state of the transformer, and quantifying the relation to be used as an initial weight value of the neural network. The characteristic vector set of the dissolved amount and concentration value of various gases when the transformer operates is defined as X, and the characteristic vector set of various operating states of the transformer is defined as Y, namely:
X n = { n kinds of eigenvectors }
Y m = { m kinds of transformer operation states }
The confidence value may be expressed as an association rule x=>Y represents a conditional probability value of occurrence of Y on the premise of occurrence of X. Conf XY Reflects the relation between the dissolved amount and concentration value vector X of various gases in the transformer gasoline at a certain moment and the running state Y of the transformer. Wherein the method comprises the steps of
Where Freq (·) is the frequency of occurrence of events in the dataset, i.e., freq (Y) represents the frequency of occurrence of events in the Y state, freq (X n Y) represents the frequency of occurrence of events that are satisfied simultaneously by X, Y.
(42) According to the two formulas, the weight calculation formula obtained from the corresponding operation state type of each gas, namely the initial weight calculation formula of the BP neural network is as follows:
wherein omega ij For the weight of j class running states when the dissolution content of i class gas in transformer oil exceeds standard, conf ij Confidence coefficient of j-class running state when i-class gas dissolution content in transformer oil exceeds standard, conf nj And the confidence coefficient of j operating states when the dissolution content of n types of gases in the transformer oil exceeds the standard is obtained.
(5) Through training of BP neural network and model parameter tuning, an Aprori-BP prediction model is obtained, and the running state of the transformer is predicted after feature vector data of the solubility and the solubility value of various gases of the real-time transformer are input.
Further, the step (5) includes:
(51) And training the BP neural network to obtain a preliminary prediction model.
The initial weights of the BP neural network are obtained in step (4), in which weight adjustments are made by repeated forward calculations and error back propagation using the transformer fault-related training set to determine final model parameter values.
(52) And performing tuning of the BP neural network related super parameters to obtain an Apriori-BP prediction model.
And performing related adjustment of a subsequent model, such as selection of a BP optimizer, weight adjustment of gradient descent in back propagation, and activating optimization measures of functions to obtain an optimal Apriori-BP prediction model through prediction accuracy evaluation of the test set.
(53) And (5) inputting real-time transformer related characteristic data and then predicting the state of the transformer.
Relevant characteristic values of various gases of each transformer are collected in real time, and corresponding state prediction is carried out through the method.
Example 1:
in the example, 522 220Kv transformer oil chromatographic data in a certain southwest area are selected as research objects, 40 of the data are selected as test data sets, and the rest 482 data are selected as training data sets of the BP neural network. Information marking the training dataset: marking 5 kinds of fault characteristic gases respectively, X 1 Each represents H 2 Exceeding the standard, X 2 Represents CH 4 Exceeding the standard, X 3 Represents C exceeding the standard 2 H 6 ，X 4 Represents C exceeding the standard 2 H 4 ，X 5 Represents C exceeding the standard 2 H 2 The working conditions of 5 types of transformers are marked respectively, Y 1 And each represents normal operation, Y 2 Represents medium-low temperature overheat, Y 3 Represents high temperature overheat, Y 4 Indicating low energy discharge, Y 5 Representing a high energy discharge.
Referring to fig. 2, a flowchart of a transformer fault prediction method based on an Apriori-BP neural network is shown, and the method includes the following steps:
(1) Acquiring relevant data of actual operation of transformer
In the training set, statistics for various gases that exceed the standards and transformer operating conditions are shown in the table below. The frequency of the gas beyond standard or transformer operating conditions is shown in brackets:
transformer operation data statistics table
| X 1 (199) | X 2 (217) | X 3 (160) | X 4 (192) | X 5 (111) | |
| Y 1 (57) | 34 | 32 | 34 | 14 | 3 |
| Y 2 (79) | 22 | 53 | 46 | 29 | 1 |
| Y 3 (146) | 12 | 92 | 51 | 109 | 3 |
| Y 4 (82) | 55 | 14 | 16 | 11 | 30 |
| Y 5 (118) | 76 | 26 | 13 | 29 | 74 |
(2) Processing the data obtained in the step (1) through an Apriori algorithm to obtain initial weights of the BP neural network
The data in the above table are processed by the following two formulas to obtain X n And Y m Related confidence between the fault type and the weight coefficient of the corresponding fault type under the condition of excessive gas.
X n = { n-class superscalar }
Y m = { m kinds of transformer operation states }
According to Y 1 Calculating, namely calculating the normal running state of the transformer:
the same thing calculates omega 21 ＝0.233ω 31 ＝0.336ω 41 ＝0.115ω 51 ＝0.043
Likewise, the operating conditions of other types of transformers are calculated to obtain weight coefficients, and the weight coefficient matrix P is as follows:
(3) Initializing the weight coefficient matrix obtained in the step (2) to train and test the BP neural network
The training frequency of the network is set to 1000 times, the training target is 0.01, and the training learning rate is 0.1. After training the training dataset, testing the transformer fault diagnosis dataset. Among the 40 transformer data, there are 6 transformer normal operation data, 10 medium-low temperature overheat data, 7 high temperature overheat data, 10 low energy discharge data and 7 high energy discharge data. The fault diagnosis is carried out by adopting a BP neural network algorithm improved by association rules, and the result is shown in a figure 3, wherein the state of the transformer is marked by an integer of 1-5 in normal operation, low-temperature overheat, high-temperature overheat, low-energy discharge and high-energy discharge respectively.
(4) Comparing the performance of the BP neural network on the test set by combining the models before and after the step (2)
The test data set is subjected to 10 times of fault diagnosis by using a traditional BP neural network and a BP neural network algorithm improved by using association rules, and diagnosis results are compared, wherein the diagnosis results are shown in the following table:
results comparison table of two algorithms
| Method | Prediction accuracy/% | Run time/s |
| BP neural network | 77.5％ | 1.247 |
| Apriori-BP | 86.4％ | 0.7657 |
From the above table, it can be seen that the performance of the Apriori-BP model of the present invention is better than that of the BP model, regardless of the prediction accuracy or the average run time, indicating that the prediction result is better than that of the general single model. In actual transformer fault prediction, the intrinsic correlation between characteristic quantities is considered, and the information of the running state of the transformer is combined, so that potential data information can be fully mined.
The method is suitable for high-precision transformer fault prediction, can technically and gradually improve the transformer fault predictability, provides decision support for maintenance and updating of the transformer, and effectively ensures safe and stable operation of the transformer.
The foregoing is only a preferred embodiment of the invention, it being noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the present invention, and such modifications and adaptations are intended to be comprehended within the scope of the invention.
Claims (4)
1. A transformer fault diagnosis method based on an Apriori-BP algorithm is characterized by comprising the following steps of: the method comprises the following steps:
training the BP neural network by utilizing characteristic vector data of gas in transformer oil and a corresponding transformer running state to obtain an apreri-BP prediction model;
inputting the characteristic vector data of the gas in the transformer oil in real time into an Aprori-BP prediction model to obtain the prediction of the state of the transformer;
the characteristic vector data acquisition steps of the gas in the transformer oil are as follows:
obtaining the dissolved amount and concentration values of a plurality of gases in the transformer oil and corresponding different running states of the transformer;
taking one sample of the dissolved amount and concentration value of each gas under different running states of the transformer, and carrying out normalization treatment according to the gas types to obtain feature vector data of each gas under different running states of the transformer;
the formula for normalizing each gas dissolution and concentration value is as follows:
wherein:is the ith dimension characteristic value, x of the jth data i max And x i min Respectively representing the maximum value and the minimum value of the ith dimension data in the data set, wherein i epsilon {0,1}, and the two dimensions 0 and 1 respectively represent the dissolution amount and the concentration value, j epsilon { 1.. The first place, l }, wherein l represents the number of training samples;
the BP neural network comprises an input layer, a hidden layer and an output layer, the number of the collected gas types is set to be the number of input nodes of the BP neural network input layer, the dimension of each node is 2, and the dimension is respectively characteristic vector data corresponding to the dissolution quantity and concentration value; setting the number of kinds of running states of the transformer as the number of output nodes of the BP neural network output layer; calculating the node number of the hidden layer h by using an empirical formula:
or->Where α is a constant between 1 and 10, input represents the number of input nodes and output represents the number of output nodes;
the initial weight of the Aprori-BP prediction model is obtained as follows:
the characteristic vector set of the dissolved amount and concentration value of various gases when the transformer operates is defined as X, and the characteristic vector set of various operating states of the transformer is defined as Y, namely:
X n = { n kinds of eigenvectors }
Y m = { m kinds of transformer operation states }
The confidence value may be expressed as an association rule x=>Y represents a conditional probability value of occurrence of Y on the premise of occurrence of X; conf XY The relation between the dissolved quantity and concentration value vector X of various gases in the transformer gasoline at a certain moment and the running state Y of the transformer is reflected; wherein the method comprises the steps of
Wherein, freq (·) is the frequency of occurrence of events in the dataset, i.e., freq (Y) represents the frequency of occurrence of events in the Y state, freq (X n Y) represents the frequency of occurrence of events that are satisfied simultaneously by X, Y;
a weight calculation formula derived from the corresponding operating condition type of each gas,
wherein omega ij J types when the dissolution content of i types of gases in transformer oil exceeds standardWeights of running states, conf ij Confidence coefficient of j-class running state when i-class gas dissolution content in transformer oil exceeds standard, conf nj And the confidence coefficient of j operating states when the dissolution content of n types of gases in the transformer oil exceeds the standard is obtained.
2. The transformer fault diagnosis method based on Apriori-BP algorithm of claim 1, wherein the method comprises the following steps: the method also comprises the following steps:
and repeatedly carrying out forward calculation and error back propagation on the Aprori-BP prediction model, and adjusting the weight.
3. The transformer fault diagnosis method based on Apriori-BP algorithm of claim 2, wherein the method comprises the following steps: the method also comprises the following steps:
and adjusting the Aprori-BP prediction model through prediction accuracy evaluation.
4. A transformer fault diagnosis method based on Apriori-BP algorithm according to claim 3, wherein: the transformer oil comprises at least one of several gases: h 2 ，CH 4 ，C 2 H 6 ，C 2 H 4 And C 2 H 2 。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010374833.XA CN111695288B (en) | 2020-05-06 | 2020-05-06 | Transformer fault diagnosis method based on Apriori-BP algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010374833.XA CN111695288B (en) | 2020-05-06 | 2020-05-06 | Transformer fault diagnosis method based on Apriori-BP algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111695288A CN111695288A (en) | 2020-09-22 |
| CN111695288B true CN111695288B (en) | 2023-08-08 |
Family
ID=72476484
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010374833.XA Active CN111695288B (en) | 2020-05-06 | 2020-05-06 | Transformer fault diagnosis method based on Apriori-BP algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111695288B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111579978B (en) * | 2020-05-18 | 2024-01-02 | 珠海施诺电力科技有限公司 | System and method for realizing relay fault identification based on artificial intelligence technology |
| CN113009077B (en) * | 2021-02-18 | 2023-05-02 | 南方电网数字电网研究院有限公司 | Gas detection method, gas detection device, electronic equipment and storage medium |
| CN113640699B (en) * | 2021-10-14 | 2021-12-24 | 南京国铁电气有限责任公司 | Fault judgment method, system and equipment for microcomputer control type alternating current and direct current power supply system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012029154A1 (en) * | 2010-09-02 | 2012-03-08 | 株式会社かんでんエンジニアリング | Internal abnormality diagnosis method, internal abnormality diagnosis system, and decision tree generating method for internal abnormality diagnosis of oil-filled electric apparatus utilizing gas concentration in oil |
| WO2013100593A1 (en) * | 2011-12-26 | 2013-07-04 | 주식회사 효성 | Method for diagnosing internal fault of oil-immersed transformer through content ratios of dissolved gases |
| WO2016127598A1 (en) * | 2015-02-13 | 2016-08-18 | 国家电网公司 | Transformer internal composite defect fuzzy diagnosis method based on gas dissolved in oil |
| CN107862114A (en) * | 2017-10-25 | 2018-03-30 | 广西电网有限责任公司电力科学研究院 | Small echo PSO SVM Diagnosis Method of Transformer Faults based on three ratio feature amounts |
| CN109765333A (en) * | 2018-12-05 | 2019-05-17 | 国网辽宁省电力有限公司信息通信分公司 | A kind of Diagnosis Method of Transformer Faults based on GoogleNet model |
| CN109840548A (en) * | 2019-01-02 | 2019-06-04 | 广东工业大学 | One kind being based on BP neural network Diagnosis Method of Transformer Faults |
| CN110689069A (en) * | 2019-09-25 | 2020-01-14 | 贵州电网有限责任公司 | Transformer fault type diagnosis method based on semi-supervised BP network |
| CN110879373A (en) * | 2019-12-12 | 2020-03-13 | 国网电力科学研究院武汉南瑞有限责任公司 | Oil-immersed transformer fault diagnosis method with neural network and decision fusion |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103218662A (en) * | 2013-04-16 | 2013-07-24 | 郑州航空工业管理学院 | Transformer fault diagnosis method based on back propagation (BP) neural network |
| CN109034276A (en) * | 2018-09-04 | 2018-12-18 | 温州大学 | Valve body method for diagnosing faults based on BP neural network |
| CN110232240B (en) * | 2019-06-12 | 2020-03-13 | 贵州电网有限责任公司 | Improved transformer top layer oil temperature prediction method |
| CN110297141A (en) * | 2019-07-01 | 2019-10-01 | 武汉大学 | Fault Locating Method and system based on multilayer assessment models |
-
2020
- 2020-05-06 CN CN202010374833.XA patent/CN111695288B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012029154A1 (en) * | 2010-09-02 | 2012-03-08 | 株式会社かんでんエンジニアリング | Internal abnormality diagnosis method, internal abnormality diagnosis system, and decision tree generating method for internal abnormality diagnosis of oil-filled electric apparatus utilizing gas concentration in oil |
| WO2013100593A1 (en) * | 2011-12-26 | 2013-07-04 | 주식회사 효성 | Method for diagnosing internal fault of oil-immersed transformer through content ratios of dissolved gases |
| WO2016127598A1 (en) * | 2015-02-13 | 2016-08-18 | 国家电网公司 | Transformer internal composite defect fuzzy diagnosis method based on gas dissolved in oil |
| CN107862114A (en) * | 2017-10-25 | 2018-03-30 | 广西电网有限责任公司电力科学研究院 | Small echo PSO SVM Diagnosis Method of Transformer Faults based on three ratio feature amounts |
| CN109765333A (en) * | 2018-12-05 | 2019-05-17 | 国网辽宁省电力有限公司信息通信分公司 | A kind of Diagnosis Method of Transformer Faults based on GoogleNet model |
| CN109840548A (en) * | 2019-01-02 | 2019-06-04 | 广东工业大学 | One kind being based on BP neural network Diagnosis Method of Transformer Faults |
| CN110689069A (en) * | 2019-09-25 | 2020-01-14 | 贵州电网有限责任公司 | Transformer fault type diagnosis method based on semi-supervised BP network |
| CN110879373A (en) * | 2019-12-12 | 2020-03-13 | 国网电力科学研究院武汉南瑞有限责任公司 | Oil-immersed transformer fault diagnosis method with neural network and decision fusion |
Non-Patent Citations (1)
| Title |
|---|
| 电站主变压器匝间短路原因分析;王金根;科技资讯;第 16 卷(第 8 期);第37-38页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111695288A (en) | 2020-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111695288B (en) | Transformer fault diagnosis method based on Apriori-BP algorithm | |
| CN109164343B (en) | Transformer fault diagnosis method based on characteristic information quantization and weighted KNN | |
| CN110232203B (en) | Knowledge distillation optimization RNN short-term power failure prediction method, storage medium and equipment | |
| CN110929847A (en) | Converter transformer fault diagnosis method based on deep convolutional neural network | |
| CN110348615B (en) | Cable line fault probability prediction method based on ant colony optimization support vector machine | |
| CN115564114B (en) | Airspace carbon emission short-term prediction method and system based on graph neural network | |
| CN111340248A (en) | Transformer fault diagnosis method and system based on intelligent integration algorithm | |
| CN112396234A (en) | User side load probability prediction method based on time domain convolutional neural network | |
| CN112836876B (en) | Power distribution network line load prediction method based on deep learning | |
| CN114358116A (en) | Oil-immersed transformer fault diagnosis method and system and readable storage medium | |
| CN117493953B (en) | Lightning arrester state evaluation method based on defect data mining | |
| CN109344518A (en) | A kind of method for diagnosing faults of base station heat management system | |
| CN116010884A (en) | Fault diagnosis method of SSA-LightGBM oil-immersed transformer based on principal component analysis | |
| CN113205125A (en) | XGboost-based extra-high voltage converter valve operation state evaluation method | |
| CN114184861A (en) | Fault diagnosis method for oil-immersed transformer | |
| CN117473424A (en) | Transformer fault diagnosis method, system, equipment and medium based on random forest | |
| CN109886314B (en) | Kitchen waste oil detection method and device based on PNN neural network | |
| CN110059938B (en) | Power distribution network planning method based on association rule driving | |
| CN109630092B (en) | Data-based multi-model soft measurement method for pumping well pump efficiency | |
| CN110516792A (en) | Non-stable time series forecasting method based on wavelet decomposition and shallow-layer neural network | |
| CN113469252A (en) | Extra-high voltage converter valve operation state evaluation method considering unbalanced samples | |
| CN117131437A (en) | Isolation method for vibration faults of engine | |
| CN117037841A (en) | Acoustic signal hierarchical cavitation intensity identification method based on hierarchical transition network | |
| CN116992362A (en) | Transformer fault characterization feature quantity screening method and device based on Xia Puli value | |
| CN116384223A (en) | Nuclear equipment reliability assessment method and system based on intelligent degradation state identification |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |