CN111563348B - Transformer fault diagnosis method based on deep support vector machine - Google Patents
Transformer fault diagnosis method based on deep support vector machine Download PDFInfo
- Publication number
- CN111563348B CN111563348B CN202010280306.2A CN202010280306A CN111563348B CN 111563348 B CN111563348 B CN 111563348B CN 202010280306 A CN202010280306 A CN 202010280306A CN 111563348 B CN111563348 B CN 111563348B
- Authority
- CN
- China
- Prior art keywords
- wolf
- support vector
- vector machine
- model
- 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.)
- Active
Links
Images
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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Housings And Mounting Of Transformers (AREA)
Abstract
The invention discloses a transformer fault diagnosis method based on a depth support vector machine, which comprises the following steps: (1) Analyzing the collected dissolved gas in the power transformer oil to obtain original data of the volume fraction of the characteristic gas components, taking 70% of the original data as a training set, and taking the remaining 30% as a test set, wherein the original data is used for testing the accuracy of the obtained fault diagnosis model; (2) carrying out normalization processing on the original data; (3) Optimizing the support vector machine parameters by utilizing the gray wolf algorithm according to the normalized data obtained in the step (2) to obtain the optimal parameters for transformer fault diagnosis, and performing support vector machine model M through a limited Boltzmann machine in deep learning r And training to obtain an optimal model for transformer fault diagnosis. The method can further improve the efficiency of transformer fault diagnosis and effectively increase the accuracy of fault diagnosis.
Description
Technical Field
The invention belongs to the technical field of transformer fault on-line monitoring methods, and particularly relates to a transformer fault diagnosis method based on a deep support vector machine.
Background
In the world today, both developed and developing countries suffer from transformer safety to varying degrees. The transformer can cause faults due to various reasons in the operation process, if the faults cannot be found and processed in time, the machine stops running and the lighting equipment is extinguished, and if the faults cannot be found and processed in time, fire disasters cause casualties, so that great loss is caused to national economy. Therefore, how to quickly and accurately diagnose the transformer fault is a great concern all over the world.
At present, dissolved Gas Analysis (DGA) in oil is one of important methods for diagnosing faults of a transformer, and the DGA can effectively find latent faults of the transformer in advance and monitor the fault development condition on line at any time. The three-ratio method, the Rogers method, the Dornerburg method, the improved three-ratio method, and the like have appeared based on the DGA technique, but these methods have problems of missing codes and too absolute code boundaries in practical applications. Therefore, artificial intelligence methods such as a support vector machine, an artificial neural network, a Bayesian classifier and the like are combined on the basis of the traditional diagnosis method. The traditional support vector machine has the advantages of simple algorithm response, good robustness and the like in solving the problems of small samples, nonlinearity and high-dimensional identification, but the traditional support vector machine is a single-core learning model, has a diagnosis result sensitive to nuclear parameters and punishment parameters, is difficult to solve the problem of multi-classification, and is not suitable for large-scale training samples. Therefore, the invention provides a transformer fault diagnosis method based on a deep support vector machine, which can obtain more accurate transformer diagnosis results more quickly.
Disclosure of Invention
The invention aims to provide a transformer fault diagnosis method based on a deep support vector machine, which can not only further improve the efficiency of transformer fault diagnosis, but also effectively increase the accuracy of fault diagnosis, and provides a new thought for transformer fault diagnosis.
The invention adopts the technical scheme that a transformer fault diagnosis method based on a deep support vector machine comprises the following steps:
(1) Analyzing the collected dissolved gas in the power transformer oil to obtain original data of the volume fraction of the characteristic gas components, taking 70% of the original data as a training set, and taking the remaining 30% of the original data as a test set, wherein the original data is used for testing the accuracy of the obtained fault diagnosis model;
(2) Normalizing the original data;
(3) Optimizing the support vector machine parameters by utilizing the gray wolf algorithm according to the normalized data obtained in the step (2) to obtain the optimal parameters for transformer fault diagnosis, and performing support vector machine model M through a limited Boltzmann machine in deep learning r And training to obtain an optimal model for transformer fault diagnosis.
The present invention is also characterized in that,
in the step (1), the characteristic gas comprises H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 These 7 gases.
In the step (2), a formula used for normalizing the original data is shown as the following formula:
wherein x is q Is represented by H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 Volume fraction, x, of these 7 gases max And x min Respectively representing the maximum and minimum values, x, of the volume fractions of the 7 gases * The normalized data of the volume fractions of the 7 gases are shown.
In the step (3), the specific steps of optimizing the support vector machine parameters by using the gray wolf algorithm are as follows:
s1, support penalty parameter c of vector machine s And the kernel parameter σ in the radial basis kernel function u Performing an initialization operation and converting c s 、σ u As the location of each gray wolf; the entire population X of grey wolves includes N grey wolves, i.e., X = { X = { X 1 ,X 2 ,…,X N For each gray wolf X } L ,L∈[1,N]In other words, its position is X L =(c s ,σ u ) The distance between the position of the gray wolf and the position of the prey is expressed by fitness, and the fitness is higher when the distance is smaller; the distance D between the wolf pack and the prey in the process of completing the enclosure of the prey can be expressed as:
D=|C·X p (t)-X(t)| (2)
wherein C is a wobble factor, X p (t) is the position vector of the prey when iterating t times, and X (t) is the position vector of the wolf when iterating t times;
updating the self position of the wolf pack according to the position of the prey and the distance between the wolf pack and the prey can be expressed as:
X(t+1)=X p (t)-A·D (3)
wherein X (t + 1) is the position vector of the wolf when iterating t +1 times, A is a convergence factor, different positions of the wolf near a prey can be changed by adjusting the two factors of A and C, and the calculation formulas of A and C are as follows:
A=2ar 1 -a (4)
C=2·r 2 (5)
wherein r is 1 And r 2 Is [0,1 ]]A is linearly decreased from 2 to 0 in the iterative process;
s2, respectively defining the highest-grade head-lead alpha wolf as an optimal solution, the vice-head-lead beta wolf as a suboptimal solution, the secondary delta wolf as a suboptimal solution, the bottom layer omega wolf as 4 gray wolfs for searching individuals, wherein the alpha wolf has the highest status in the whole predation process, the beta wolf is the second best, the delta wolf is the second best, the lowest omega wolf is the omega wolf, wherein the alpha wolf, the beta wolf and the delta wolf are responsible for guiding, the omega wolf is responsible for surrounding and catching, and the distances between the alpha wolf, the beta wolf, the delta wolf and the prey are respectively:
and moving to the next step according to the current distance:
the omega wolf updates the position of the omega wolf according to the positions of the alpha wolf, the beta wolf and the delta wolf:
according to the method, c can be obtained by continuously iterating until the final condition is met s And σ u The optimal solution of (1).
In the step (3), a support vector machine model M is paired through a limited Boltzmann machine in deep learning r The specific steps of training to obtain the optimal model for transformer fault diagnosis are as follows:
firstly, a model M of a support vector machine is supported by a restricted Boltzmann machine in deep learning r Training is carried out, and for a certain limited Boltzmann machine, the energy of the machine is as follows:
wherein phi = { a = i ,b j ,ω ij Are parameters of a restricted Boltzmann machine and are all real numbers, v i Represents the visible cell state, h j Representing hidden unit states, ω ij Representing the weight between visible unit i and hidden unit j, a i To see the offset of cell i, b j For the bias of hidden unit j, the joint distribution probability of (v, h) is:
wherein Z (φ) is a normalization factor;
support vector machine model M r The estimated error calculation formula is as follows:
err d =y d -y' d (11)
wherein, err d Denotes the d-th estimated error, y' d Representing the volume fraction of dissolved gas in the transformer oil after normalization treatment;
therefore, the support vector machine model M r The average error of (a) is calculated as:
where err is the support vector machine model M r Average error of (2);
and finally, obtaining N models after optimizing parameters of the support vector machine by the gray wolf algorithm and training the support vector machine by the limited Boltzmann machine, wherein the optimal model is the model with the minimum err value, and the parameters of the model are also the optimal parameters.
The invention has the beneficial effects that: according to the method, the collected volume fraction data of the dissolved gas in the transformer oil is subjected to normalization processing, so that the data is limited in a certain range, adverse effects caused by singular sample data are eliminated, and the precision is improved; according to the method, the parameters of the support vector machine are optimized by using the wolf algorithm, so that the parameters are converged to a global optimum value more quickly, the stability of a model is improved, and the operation speed is increased; the method combines deep learning with the support vector machine, and because the deep learning has strong expression and is easy to reason, the fault diagnosis capability of the model is further improved, and the fault diagnosis accuracy is increased.
Drawings
FIG. 1 is a flow chart of a transformer fault diagnosis method based on a deep support vector machine according to the invention.
Detailed Description
The invention is described in detail below with reference to the drawings and the detailed description.
The invention relates to a transformer fault diagnosis method based on a depth support vector machine, which comprises the following steps as shown in figure 1:
(1) Analyzing the collected dissolved gas in the power transformer oil to obtain original data of the volume fraction of the characteristic gas components, taking 70% of the original data as a training set, and taking the remaining 30% as a test set, wherein the original data is used for testing the accuracy of the obtained fault diagnosis model;
in the step (1), the characteristic gas comprises H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 These 7 kinds of gases;
(2) Carrying out normalization processing on the original data;
in the step (2), a formula used for normalizing the original data is shown as the following formula:
wherein x is q Represents H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 Volume fraction, x, of these 7 gases max And x min Respectively representing the maximum and minimum values, x, of the volume fractions of the 7 gases * The volume fractions of the 7 gases are normalized;
(3) Optimizing the support vector machine parameters by utilizing the gray wolf algorithm according to the normalized data obtained in the step (2) to obtain the optimal parameters for transformer fault diagnosis, and performing support vector machine model M through a limited Boltzmann machine in deep learning r Training to obtain an optimal model for transformer fault diagnosis;
in the step (3), the specific steps of optimizing the support vector machine parameters by using the gray wolf algorithm are as follows:
s1, support penalty parameter c of vector machine s And the kernel parameter σ in the radial basis kernel function u Performing an initialization operation and converting c s 、σ u As the location of each gray wolf; the entire population of wolfs X comprises N wolf individuals, i.e. X = { X = { (X) 1 ,X 2 ,…,X N For each gray wolf X } L ,L∈[1,N]In other words, its position is X L =(c s ,σ u ) The distance between the position of the gray wolf and the position of the prey is expressed by fitness, and the fitness is higher when the distance is smaller; the distance D between the wolf pack and the prey in the process of completing the enclosure of the prey can be expressed as:
D=|C·X p (t)-X(t)| (2)
wherein C is a wobble factor, X p (t) is the position vector of the prey when iterating t times, and X (t) is the position vector of the wolf when iterating t times;
updating the self position of the wolf pack according to the position of the prey and the distance between the wolf pack and the prey can be expressed as:
X(t+1)=X p (t)-A·D (3)
wherein X (t + 1) is the position vector of the wolf when iterating for t +1 times, A is a convergence factor, different positions of the wolf near a prey can be changed by adjusting the two factors of A and C, and the calculation formulas of A and C are as follows:
A=2ar 1 -a (4)
C=2·r 2 (5)
wherein r is 1 And r 2 Is [0,1 ]]A is a random vector in between, and a is linearly reduced from 2 to 0 in the iteration process;
s2, respectively defining the highest-grade head-leading alpha wolf as an optimal solution, the vice-head-leading beta wolf as a suboptimal solution, the secondary delta wolf as a next-best solution, the bottom-layer omega wolf as 4 grey wolfs for searching individuals, wherein the alpha wolf has the highest position in the whole predation process, the beta wolf is the second-highest position, the delta wolf is the next-highest position, and the lowest omega wolf is the omega wolf, wherein the alpha wolf, the beta wolf and the delta wolf are used for guiding, the omega wolf is used for surrounding and catching, and the distances between the alpha wolf, the beta-delta wolf and a prey are respectively as follows:
and moving to the next step according to the current distance:
the omega wolf updates the position thereof according to the positions of the alpha wolf, the beta wolf and the delta wolf:
according to the method, c can be obtained by continuously iterating until the final condition is met s And σ u The optimal solution of (a).
Support vector machine model M in step (3) through limited Boltzmann machine in deep learning r The specific steps of training to obtain the optimal model for transformer fault diagnosis are as follows:
first, a model M of a support vector machine is modeled by a limited boltzmann machine in deep learning r Training is carried out, and for a certain limited Boltzmann machine, the energy of the machine is as follows:
wherein phi = { a = i ,b j ,ω ij Are parameters of a restricted Boltzmann machine and are all real numbers, v i Represents the visible cell state, h j Representing hidden unit states, ω ij Representing the weight between visible unit i and hidden unit j, a i Bias for visible cell i, b j For the bias of hidden unit j, the joint distribution probability of (v, h) is:
wherein Z (φ) is a normalization factor;
support vector machine model M r The estimated error calculation formula is as follows:
err d =y d -y' d (11)
wherein, err d Denotes the d-th estimated error, y' d Representing the volume fraction of dissolved gas in the transformer oil after normalization treatment;
therefore, the support vector machine model M r The average error of (a) is calculated as:
where err is the support vector machine model M r Average error of (2);
and finally, obtaining N models after optimizing parameters of the support vector machine by the gray wolf algorithm and training the support vector machine by the limited Boltzmann machine, wherein the optimal model is the model with the minimum err value, and the parameters of the model are also the optimal parameters.
Claims (3)
1. A transformer fault diagnosis method based on a deep support vector machine is characterized by comprising the following steps:
(1) Analyzing the collected dissolved gas in the power transformer oil to obtain original data of the volume fraction of the characteristic gas components, taking 70% of the original data as a training set, and taking the remaining 30% as a test set, wherein the original data is used for testing the accuracy of the obtained fault diagnosis model;
(2) Normalizing the original data;
(3) Optimizing the parameters of the support vector machine by utilizing a grey wolf algorithm according to the normalized data obtained in the step (2) to obtain the optimal parameters for fault diagnosis of the transformer, and carrying out deep learning on the support vector machine model M by using a limited Boltzmann machine r Training to obtain an optimal model for transformer fault diagnosis;
in step (1), the characteristic gas comprises H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 These 7 gases;
in the step (3), a support vector machine model M is paired through a limited Boltzmann machine in deep learning r The specific steps of training to obtain the optimal model for transformer fault diagnosis are as follows:
firstly, a model M of a support vector machine is supported by a restricted Boltzmann machine in deep learning r Training is carried out, and for a certain limited Boltzmann machine, the energy of the machine is as follows:
wherein phi = { a = i ,b j ,ω ij Are parameters of a restricted Boltzmann machine and are all real numbers, v i Represents the visible cell state, h j Representing hidden unit states, ω ij Representing the weight between visible unit i and hidden unit j, a i To see the offset of cell i, b j For the bias of hidden unit j, the joint distribution probability of (v, h) is:
wherein Z (φ) is a normalization factor;
support vector machine model M r The estimated error calculation formula is as follows:
err d =y d -y' d (11)
wherein, err d Denotes the d-th estimated error, y' d Expressing the volume fraction of dissolved gas in the transformer oil after normalization treatment;
therefore, the support vector machine model M r The average error of (a) is calculated as:
where err is the support vector machine model M r Average error of (2);
and finally, obtaining N models after optimizing parameters of the support vector machine by a grey wolf algorithm and training the support vector machine by a limited Boltzmann machine, wherein the optimal model is the model with the minimum err value, and the parameters of the model are also the optimal parameters.
2. The method for diagnosing the fault of the transformer based on the deep support vector machine according to claim 1, wherein in the step (2), the formula used for performing the normalization processing on the raw data is as follows:
wherein x is q Is represented by H 2 、CH 4 、C 2 H 2 、C 2 H 4 、C 2 H 6 、CO、CO 2 Volume fraction, x, of these 7 gases max And x min Respectively representing the maximum and minimum values, x, of the volume fractions of the 7 gases * The normalized volume fraction of these 7 gases is shown.
3. The method for diagnosing the fault of the transformer based on the deep support vector machine according to claim 1, wherein in the step (3), the specific steps of optimizing the parameters of the support vector machine by using the wolf algorithm are as follows:
s1, support penalty parameter c of vector machine s And the kernel parameter σ in the radial basis kernel function u Performing an initialization operation and converting c s 、σ u As the location of each gray wolf; the entire population X of grey wolves includes N grey wolves, i.e., X = { X = { X 1 ,X 2 ,…,X N For each gray wolf X } L ,L∈[1,N]In other words, its position is X L =(c s ,σ u ) The distance between the position of the gray wolf and the position of the prey is expressed by fitness, and the fitness is higher when the distance is smaller; the distance D between the wolf pack and the prey in the process of completing the enclosure of the prey can be expressed as:
D=|C·X p (t)-X(t)| (2)
wherein C is a wobble factor, X p (t) is the position vector of the prey when iterating t times, and X (t) is the position vector of the wolf when iterating t times;
updating the self position of the wolf pack according to the position of the prey and the distance between the wolf pack and the prey can be expressed as:
wherein X (t + 1) is the position vector of the wolf when iterating t +1 times, A is a convergence factor, different positions of the wolf near a prey can be changed by adjusting the two factors of A and C, and the calculation formulas of A and C are as follows:
A=2ar 1 -a (4)
C=2·r 2 (5)
wherein r is 1 And r 2 Is [0,1 ]]A is a random vector in between, and a is linearly reduced from 2 to 0 in the iteration process;
s2, respectively defining the highest-grade head-lead alpha wolf as an optimal solution, the vice-head-lead beta wolf as a suboptimal solution, the secondary delta wolf as a suboptimal solution, the bottom layer omega wolf as 4 gray wolfs for searching individuals, wherein the alpha wolf has the highest status in the whole predation process, the beta wolf is the second best, the delta wolf is the second best, the lowest omega wolf is the omega wolf, wherein the alpha wolf, the beta wolf and the delta wolf are responsible for guiding, the omega wolf is responsible for surrounding and catching, and the distances between the alpha wolf, the beta wolf, the delta wolf and the prey are respectively:
and moving to the next step according to the current distance:
the omega wolf updates the position of the omega wolf according to the positions of the alpha wolf, the beta wolf and the delta wolf:
according to the method, c can be obtained by continuously iterating until the final condition is met s And σ u The optimal solution of (1).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010280306.2A CN111563348B (en) | 2020-04-10 | 2020-04-10 | Transformer fault diagnosis method based on deep support vector machine |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010280306.2A CN111563348B (en) | 2020-04-10 | 2020-04-10 | Transformer fault diagnosis method based on deep support vector machine |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111563348A CN111563348A (en) | 2020-08-21 |
CN111563348B true CN111563348B (en) | 2023-04-18 |
Family
ID=72074224
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010280306.2A Active CN111563348B (en) | 2020-04-10 | 2020-04-10 | Transformer fault diagnosis method based on deep support vector machine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111563348B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112285541B (en) * | 2020-09-21 | 2023-05-30 | 南京理工大学 | Fault diagnosis method for current frequency conversion circuit |
CN113553755B (en) * | 2021-06-07 | 2022-12-13 | 国网河北省电力有限公司馆陶县供电分公司 | Power system state estimation method, device and equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017128455A1 (en) * | 2016-01-25 | 2017-08-03 | 合肥工业大学 | Analogue circuit fault diagnosis method based on generalized multiple kernel learning-support vector machine |
CN108414923A (en) * | 2018-02-05 | 2018-08-17 | 武汉大学 | A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization |
CN110132596A (en) * | 2019-04-24 | 2019-08-16 | 昆明理工大学 | A method of the rolling bearing fault diagnosis based on wavelet packet and GWO-SVM |
CN110376458A (en) * | 2019-07-03 | 2019-10-25 | 东华大学 | Optimize the diagnosing fault of power transformer system of twin support vector machines |
-
2020
- 2020-04-10 CN CN202010280306.2A patent/CN111563348B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017128455A1 (en) * | 2016-01-25 | 2017-08-03 | 合肥工业大学 | Analogue circuit fault diagnosis method based on generalized multiple kernel learning-support vector machine |
CN108414923A (en) * | 2018-02-05 | 2018-08-17 | 武汉大学 | A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization |
CN110132596A (en) * | 2019-04-24 | 2019-08-16 | 昆明理工大学 | A method of the rolling bearing fault diagnosis based on wavelet packet and GWO-SVM |
CN110376458A (en) * | 2019-07-03 | 2019-10-25 | 东华大学 | Optimize the diagnosing fault of power transformer system of twin support vector machines |
Non-Patent Citations (2)
Title |
---|
张艳 ; 吴玲 ; .基于支持向量机和交叉验证的变压器故障诊断.中国电力.2012,(11),全文. * |
许静 ; 朱永利 ; .基于受限玻尔兹曼机的变压器故障分类.电力科学与工程.2016,(07),全文. * |
Also Published As
Publication number | Publication date |
---|---|
CN111563348A (en) | 2020-08-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11840998B2 (en) | Hydraulic turbine cavitation acoustic signal identification method based on big data machine learning | |
CN111563348B (en) | Transformer fault diagnosis method based on deep support vector machine | |
CN110929847A (en) | Converter transformer fault diagnosis method based on deep convolutional neural network | |
CN111523778A (en) | Power grid operation safety assessment method based on particle swarm algorithm and gradient lifting tree | |
CN116204794B (en) | Method and system for predicting dissolved gas in transformer oil by considering multidimensional data | |
CN114360703A (en) | Medical equipment fault early warning method | |
CN111680875A (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
CN111027629A (en) | Power distribution network fault outage rate prediction method and system based on improved random forest | |
CN114528949A (en) | Parameter optimization-based electric energy metering abnormal data identification and compensation method | |
CN111210089A (en) | Stock price prediction method of gated cyclic unit neural network based on Kalman filtering | |
CN109711435A (en) | A kind of support vector machines on-Line Voltage stability monitoring method based on genetic algorithm | |
CN112504682A (en) | Chassis engine fault diagnosis method and system based on particle swarm optimization algorithm | |
CN109856494A (en) | A kind of Diagnosis Method of Transformer Faults based on support vector machines | |
Jingming et al. | Research on credit risk measurement of small and micro enterprises based on the integrated algorithm of improved GSO and ELM | |
CN116451117A (en) | Power data anomaly detection method based on federal learning | |
CN114553671A (en) | Diagnosis method for power communication network fault alarm | |
CN117370766A (en) | Satellite mission planning scheme evaluation method based on deep learning | |
Liang et al. | Dissolved gas analysis of transformer oil based on Deep Belief Networks | |
CN117276600A (en) | PSO-GWO-DELM-based proton exchange membrane fuel cell system fault diagnosis method | |
CN115659258B (en) | Power distribution network fault detection method based on multi-scale graph roll-up twin network | |
CN113255223B (en) | Short-term prediction method and system for air conditioner load | |
Hu et al. | Neural network-assisted expensive optimisation algorithm for pollution source rapid positioning of drinking water | |
CN116010884A (en) | Fault diagnosis method of SSA-LightGBM oil-immersed transformer based on principal component analysis | |
Liu et al. | Swarm intelligence for classification of remote sensing data | |
Liu et al. | PM2. 5 Concentration Prediction Based on Pollutant Pattern Recognition Using PCA-clustering Method and CS Algorithm Optimized SVR |
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 |