CN109102012A - A kind of defect identification method and system of local discharge signal - Google Patents
A kind of defect identification method and system of local discharge signal Download PDFInfo
- Publication number
- CN109102012A CN109102012A CN201810852601.3A CN201810852601A CN109102012A CN 109102012 A CN109102012 A CN 109102012A CN 201810852601 A CN201810852601 A CN 201810852601A CN 109102012 A CN109102012 A CN 109102012A
- Authority
- CN
- China
- Prior art keywords
- discharge signal
- local discharge
- defect
- trained
- encoding encoder
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Abstract
The invention discloses a kind of defect identification methods of local discharge signal, it is comprising steps of (1) obtains the training sample for characterizing the local discharge signal of several shelf depreciation insulation defect type, its phase-resolved pulse train data is extracted based on training sample, phase-resolved pulse train data are normalized;(2) it is trained by sparse self-encoding encoder model of the phase-resolved pulse train data after normalized to building, the training feature vector of dimension is reduced based on trained sparse self-encoding encoder model;(3) extreme learning machine model is trained using the training feature vector;(4) local discharge signal to be identified is inputted into trained sparse self-encoding encoder model, to extract testing feature vector, testing feature vector is inputted into trained extreme learning machine model, to obtain the defect type of local discharge signal characterization to be identified.The invention also discloses a kind of defect recognition systems of local discharge signal.
Description
Technical field
The present invention relates to a kind of recognition methods more particularly to a kind of recognition methods for defect recognition.
Background technique
Stable operation of the high voltage installation since internal flaw causes shelf depreciation, so as to cause failure, to electric system
It threatens.Therefore, the research of shelf depreciation knows meaning with great to the safety for ensureing electric system.
And with the development of artificial intelligence, neural network, support vector machine (SupportVectorMachine, SVM), shellfish
Leaf this network, decision tree etc. are widely used in the fault diagnosis of high-tension apparatus, the especially identification to local discharge mode.
However, there are still more deficiencies for current mode identification method, such as when sample is less, certain pattern-recognition sides
Method is unstable, and discrimination is not high, in another example when in face of imbalanced data sets, (for example fault sample number is far smaller than normal sample
Number) when, certain mode identification methods will appear the problem of classification interface deviates.In addition, current mode identification method identifies speed
Degree is slower, actual needs is unable to satisfy, so that existing mode identification method is very restricted in actual application.
Based on this, it is expected that obtaining a kind of defect identification method, shelf depreciation can be believed with overcome the deficiencies in the prior art
Number effectively and timely defect recognition is carried out, to effectively obtain the insulation status of equipment, removes a hidden danger in time, avoid major accident
Generation, to equipment safety maintenance have great directive significance.
Summary of the invention
One of the objects of the present invention is to provide a kind of defect identification method of local discharge signal, the local discharge signals
Defect identification method timely and effectively defect recognition can be carried out to local discharge signal, and its recognition effect and performance are equal
It is promoted compared with the prior art.
Based on above-mentioned purpose, the invention proposes a kind of defect identification methods of local discharge signal comprising step:
(1) training sample for characterizing the local discharge signal of several shelf depreciation insulation defect type is obtained, based on instruction
Practice its phase-resolved pulse train data of sample extraction, phase-resolved pulse train data are normalized;
(2) it is carried out by sparse self-encoding encoder model of the phase-resolved pulse train data after normalized to building
Training, the training feature vector of dimension is reduced based on trained sparse self-encoding encoder model;
(3) extreme learning machine model is trained using the training feature vector;
(4) local discharge signal to be identified is inputted into trained sparse self-encoding encoder model, it is special to extract test
Vector is levied, testing feature vector is inputted into trained extreme learning machine model, to obtain local discharge signal table to be identified
The defect type of sign.
In the defect identification method of local discharge signal of the present invention, by sparse self-encoding encoder model to training
Sample applies sparsity restrictive condition, to carry out dimensionality reduction to training sample, obtains the training feature vector for reducing dimension.It will drop
The training feature vector of low dimensional is trained extreme learning machine model, is then used for trained extreme learning machine model
To local discharge signal carry out defect recognition, process be by local discharge signal to be identified input it is trained it is sparse from
Testing feature vector is inputted trained extreme learning machine model to extract testing feature vector by encoder model, with
The defect type characterized to local discharge signal to be identified.Since the prior art has iteration repeatedly, to increase algorithm
Time, so that it occurs calculating the problem that the time is longer, recognition speed is slower in the identification process for carrying out defect type, and it is right
For defect identification method of the present invention, which employs extreme learning machine model to local discharge signal to be identified into
The advantages of row defect recognition, there is learning process once to complete, be not necessarily to iteration for extreme learning machine model, thus, it is of the present invention
Defect identification method can achieve faster defect recognition speed.
It can thus be seen that the defect identification method of portion's discharge signal of this case, can be directed to the office of different defect types
Portion's discharge signal carries out effective defect recognition, and its discrimination is higher, fast speed, has more preferably recognition performance, gram
Deficiency of the prior art is taken.
Further, in the defect identification method of local discharge signal of the present invention, the sparse self-encoding encoder
Model has 4 layers of hidden layer.
Further, in the defect identification method of local discharge signal of the present invention, the nerve of the hidden layer
First number is 25.
Further, in the defect identification method of local discharge signal of the present invention, the sparse self-encoding encoder
The activation primitive of model uses Sigmod function.
Further, it in the defect identification method of local discharge signal of the present invention, in step (2), uses
Stochastic gradient descent method is trained sparse self-encoding encoder model, to be iterated update to its parameter, obtains optimizing ginseng
Number.
Further, in the defect identification method of local discharge signal of the present invention, in training for step (3)
Cheng Zhong seeks the optimal solution for minimizing loss function using least square method.
Correspondingly, another object of the present invention is to provide a kind of defect recognition system of local discharge signal, the parts
The defect recognition system of discharge signal can carry out timely and effectively defect recognition to local discharge signal.
Based on above-mentioned purpose, the invention also provides a kind of defect recognition systems of local discharge signal comprising:
Preprocessing module locates the local discharge signal for characterizing several shelf depreciation insulation defect type in advance
Reason, to extract its phase-resolved pulse train data, and is normalized phase-resolved pulse train data;
Sparse self-encoding encoder module, based on the phase-resolved pulse train data after normalized to building it is sparse from
Encoder model is trained, to be reduced the training feature vector of dimension;
Extreme learning machine module is trained extreme learning machine model using the training feature vector;
Wherein, when needing to identify local discharge signal, local discharge signal to be identified is inputted by instruction
Testing feature vector is inputted the trained limit and learnt by experienced sparse self-encoding encoder model to extract testing feature vector
Machine model, to obtain the defect type of local discharge signal characterization to be identified.
Further, in the defect recognition system of local discharge signal of the present invention, the sparse self-encoding encoder
Model has 4 layers of hidden layer.
Further, in the defect recognition system of local discharge signal of the present invention, the nerve of the hidden layer
First number is 25.
Further, in the defect recognition system of local discharge signal of the present invention, the sparse self-encoding encoder
The activation primitive of model uses Sigmod function.
The defect identification method and system of local discharge signal of the present invention have the following advantages and beneficial effects:
The defect identification method of the local discharge signal is by sparse self-encoding encoder model to the phase got point
It distinguishes that pulse train data are trained, the training feature vector for reducing dimension is obtained, by training feature vector to extreme learning machine
Model is trained, and local discharge signal to be identified is inputted trained sparse self-encoding encoder model, to extract test
Testing feature vector is inputted trained extreme learning machine model, to obtain local discharge signal to be identified by feature vector
The defect type of characterization, to finally realize to local discharge signal effectively and timely defect recognition.
The defect identification method recognition result is accurate, and recognition performance is good, can effectively obtain the insulation shape of equipment
Condition is removed a hidden danger in time, avoids the generation of major accident, has great directive significance to equipment safety maintenance.
In addition, the defect recognition system of local discharge signal of the present invention also has above advantages and intentionally effect
Fruit.
Detailed description of the invention
Fig. 1 is the frame structure of the defect recognition system of local discharge signal of the present invention in one embodiment
Schematic diagram.
Fig. 2 is the process signal of the defect identification method of local discharge signal of the present invention in one embodiment
Figure.
Specific embodiment
It below will according to specific embodiment and Figure of description is to the defect recognition of local discharge signal of the present invention
Method and system are described further, but the explanation does not constitute the improper restriction to technical solution of the present invention.
As shown in Figure 1, in the present embodiment, the defect recognition system of local discharge signal includes preprocessing module, dilute
Dredge self-encoding encoder module and extreme learning machine module.
Wherein, preprocessing module locates the local discharge signal for characterizing several shelf depreciation insulation defect type in advance
Reason, to extract its phase-resolved pulse train data, and is normalized phase-resolved pulse train data.
Then, sparse self-encoding encoder module by based on the phase-resolved pulse train data after normalized to building
Sparse self-encoding encoder model is trained, to be reduced the training feature vector of dimension.Wherein, in the present embodiment, dilute
Dredge self-encoding encoder model include 4 layers of hidden layer, hidden layer neuron number be 25.The activation letter of sparse self-encoding encoder model
Number uses Sigmod function.The last layer hidden layer of sparse self-encoding encoder module exports training feature vector to be learnt to the limit
Machine module.
In addition, extreme learning machine module is trained extreme learning machine model using above-mentioned training feature vector.
When needing to identify local discharge signal, local discharge signal to be identified is inputted trained dilute
It dredges self-encoding encoder model and testing feature vector is inputted into trained extreme learning machine model to extract testing feature vector,
To obtain the defect type of local discharge signal characterization to be identified.
Fig. 2 shows the process of the defect identification method of local discharge signal of the present invention in one embodiment
Schematic diagram.
As shown in Fig. 2, the recognition methods of the defects of present embodiment the following steps are included:
Step 100: obtaining the training sample for characterizing the local discharge signal of several shelf depreciation insulation defect type, base
Its phase-resolved pulse train data is extracted in training sample, phase-resolved pulse train data are normalized.
Step 200: by the phase-resolved pulse train data after normalized to the sparse self-encoding encoder mould of building
Type is trained, and the training feature vector of dimension is reduced based on trained sparse self-encoding encoder model.
Step 300: extreme learning machine model being trained using the training feature vector.
Step 400: local discharge signal to be identified being inputted into trained sparse self-encoding encoder model, is surveyed with extracting
Feature vector is tried, testing feature vector is inputted into trained extreme learning machine model, to obtain shelf depreciation letter to be identified
Number characterization defect type.
It should be pointed out that in step 100, the part of the several shelf depreciation insulation defect type of characterization of acquisition is put
The training sample of electric signal can acquire local discharge signal sample data by partial discharge simulation experiment and obtain, for example,
In present embodiment, such as tip corona is obtained using digital PD meter, oscillograph and Portable partial discharge detector and is lacked
It falls into, the local discharge signal of the shelf depreciation insulation defect of floating potential defect, bubble-discharge defect and creeping discharge defect
Sample data.
Certainly, those skilled in the art can also acquire acquisition training sample by other means, however it is not limited to logical
It crosses partial discharge simulation experiment acquisition local discharge signal sample data to obtain, and shelf depreciation insulation defect type is not yet
It is confined to tip corona defect, floating potential defect, bubble-discharge defect and creeping discharge defect, those skilled in that art
It can be configured according to the concrete condition of embodiment, therefore, here, repeating no more.
The training sample that will acquire is according to phase-resolved pulse train ((Phase Resolved Pulse
Sequence, abbreviation PRPS) map extracts to obtain phase-resolved pulse train data.PRPS map expresses a shelf depreciation
The distribution characteristics of partial discharge pulse's amplitude and pulse number that data are counted according to phase, can be by a two-dimensional matrix table
Show, two dimensions of matrix respectively represent phase and period, and the numerical value of matrix represents the amplitude of partial discharge pulse, and difference is come
The data of the local discharge signal in source can different from phase resolution and amplitude resolution.And in the present embodiment,
PRPS map periods dimension is 50, and phase-resolved degree is 5, and phase is having a size of 72, that is to say, that phase-resolved pulse train data
Size be 50 × 72=3600 dimension, acquired phase-resolved pulse train data are normalized using following formula:
In formula, yRFor the sample amplitude after normalization, RdFor dynamic range lower limit, RuFor dynamic range headroom, yminFor sample
The minimum value of this amplitude, ymaxFor the maximum value of sample amplitude, y indicates sample amplitude, and size is between yminTo ymaxBetween.
In step 200, the phase-resolved pulse train data after normalized are divided into training set and test set, example
Such as 8568 groups of phase-resolved pulse train data in total, 6000 groups of data therein are divided into training set, remaining data at random
Then it is test set, the sparse self-encoding encoder model of building is trained using stochastic gradient descent method using training set, with right
Its parameter is iterated update, obtains the most optimized parameter, and finally, trained sparse self-encoding encoder model is reduced dimension
To 25 training feature vectors.Wherein, the sparsity parameter of sparse self-encoding encoder model can be chosen for 0.05, and penalty factor can
To use relative entropy.In addition, sparse self-encoding encoder model include 4 layers of hidden layer, hidden layer neuron number be 25.And
The activation primitive of sparse self-encoding encoder model uses Sigmod function.
In step 300, in the training process using the training feature vector training extreme learning machine model of acquisition, at certain
In a little embodiments, the input layer of extreme learning machine model is with the connection weight of implicit interlayer and the threshold value of hidden layer neuron
It is randomly generated, the optimal solution for minimizing loss function then can be sought using least square method.
In order to verify this case defect identification method recognition effect, will using this case defect identification method embodiment
1 carries out identification comparison with comparative example 1-2, and comparing result is listed in table 1.Wherein, comparative example 1 is using the branch based on statistical nature
Hold the recognition methods of vector machine, comparative example 2 using BP neural network recognition methods.
Table 1.
As can be seen from Table 1, although embodiment of this case 1 part shelf depreciation insulation defect type discrimination lower than pair
Ratio 1, but 1 overall performance of embodiment of this case is stablized, for different shelf depreciation insulation defect types, embodiment 1 is identified
Rate is above 91%, significantly larger than comparative example 2, illustrates to have using the embodiment 1 of the defect identification method of this case better
Discrimination and more preferably recognition performance are highly suitable in actual application knowing shelf depreciation insulation defect type
Not.
In conclusion in conjunction with Fig. 1 to Fig. 2 and table 1 as can be seen that the defect of local discharge signal of the present invention is known
Other method and system are trained the phase-resolved pulse train data got by sparse self-encoding encoder model, are dropped
The training feature vector of low dimensional is trained extreme learning machine model by training feature vector, part to be identified is put
Electric signal inputs trained sparse self-encoding encoder model, to extract testing feature vector, testing feature vector is inputted and is passed through
Trained extreme learning machine model is crossed, to obtain the defect type of local discharge signal characterization to be identified, thus finally realization pair
Local discharge signal effectively and timely defect recognition.
In addition, the defect identification method and system identification result is accurate, recognition performance is good, can effectively obtain and set
Standby insulation status, removes a hidden danger in time, avoids the generation of major accident, has great directive significance to equipment safety maintenance.
It should be noted that prior art part is not limited to given by present specification in protection scope of the present invention
Embodiment, all prior arts not contradicted with the solution of the present invention, including but not limited to first patent document, formerly
Public publication, formerly openly use etc., it can all be included in protection scope of the present invention.
In addition, it should also be noted that, institute in the combination of each technical characteristic and unlimited this case claim in this case
Combination documented by the combination or specific embodiment of record, all technical characteristics documented by this case can be to appoint
Where formula is freely combined or is combined, unless generating contradiction between each other.
The above embodiment is a preferred embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, other any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present invention,
It should be equivalent substitute mode, be included within the scope of the present invention.
Claims (10)
1. a kind of defect identification method of local discharge signal, which is characterized in that comprising steps of
(1) training sample for characterizing the local discharge signal of several shelf depreciation insulation defect type is obtained, based on training sample
This extracts its phase-resolved pulse train data, and phase-resolved pulse train data are normalized;
(2) it is instructed by sparse self-encoding encoder model of the phase-resolved pulse train data after normalized to building
Practice, the training feature vector of dimension is reduced based on trained sparse self-encoding encoder model;
(3) extreme learning machine model is trained using the training feature vector;
(4) local discharge signal to be identified is inputted into trained sparse self-encoding encoder model, with extract test feature to
Amount, inputs trained extreme learning machine model for testing feature vector, to obtain local discharge signal characterization to be identified
Defect type.
2. the defect identification method of local discharge signal as described in claim 1, which is characterized in that the sparse self-encoding encoder
Model has 4 layers of hidden layer.
3. the defect identification method of local discharge signal as claimed in claim 2, which is characterized in that the nerve of the hidden layer
First number is 25.
4. the defect identification method of local discharge signal as described in claim 1, the activation of the sparse self-encoding encoder model
Function uses Sigmod function.
5. the defect identification method of local discharge signal as described in claim 1, which is characterized in that in step (2), use
Stochastic gradient descent method is trained sparse self-encoding encoder model, to be iterated update to its parameter, obtains optimizing ginseng
Number.
6. the defect identification method of the local discharge signal as described in any one of claim 1-5, which is characterized in that in step
Suddenly in the training process of (3), the optimal solution for minimizing loss function is sought using least square method.
7. a kind of defect recognition system of local discharge signal characterized by comprising
Preprocessing module pre-processes the local discharge signal for characterizing several shelf depreciation insulation defect type, with
Its phase-resolved pulse train data is extracted, and phase-resolved pulse train data are normalized;
Sparse self-encoding encoder module, based on the phase-resolved pulse train data after normalized to the sparse from coding of building
Device model is trained, to be reduced the training feature vector of dimension;
Extreme learning machine module is trained extreme learning machine model using the training feature vector;
Wherein, when needing to identify local discharge signal, local discharge signal to be identified is inputted trained
Testing feature vector is inputted trained extreme learning machine mould to extract testing feature vector by sparse self-encoding encoder model
Type, to obtain the defect type of local discharge signal characterization to be identified.
8. the defect recognition system of local discharge signal as claimed in claim 7, which is characterized in that the sparse self-encoding encoder
Model has 4 layers of hidden layer.
9. the defect recognition system of local discharge signal as claimed in claim 8, which is characterized in that the nerve of the hidden layer
First number is 25.
10. the defect recognition system of local discharge signal as claimed in claim 7, the activation of the sparse self-encoding encoder model
Function uses Sigmod function.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810852601.3A CN109102012A (en) | 2018-07-30 | 2018-07-30 | A kind of defect identification method and system of local discharge signal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810852601.3A CN109102012A (en) | 2018-07-30 | 2018-07-30 | A kind of defect identification method and system of local discharge signal |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109102012A true CN109102012A (en) | 2018-12-28 |
Family
ID=64848119
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810852601.3A Pending CN109102012A (en) | 2018-07-30 | 2018-07-30 | A kind of defect identification method and system of local discharge signal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109102012A (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109596955A (en) * | 2018-12-30 | 2019-04-09 | 国网北京市电力公司 | Shelf depreciation state determines method and device |
CN110108754A (en) * | 2019-04-25 | 2019-08-09 | 四川沐迪圣科技有限公司 | The light stimulus infrared thermal imaging defect inspection method decomposed based on structural sparse |
CN110161389A (en) * | 2019-06-12 | 2019-08-23 | 国网河北省电力有限公司电力科学研究院 | A kind of Electric Power Equipment Insulation defect identification method and AEVB self-encoding encoder |
CN110286279A (en) * | 2019-06-05 | 2019-09-27 | 武汉大学 | Based on extreme random forest and the sparse Fault Diagnosis of Power Electronic Circuits method from encryption algorithm of stacking-type |
CN110334948A (en) * | 2019-07-05 | 2019-10-15 | 上海交通大学 | Power equipment shelf depreciation Severity method and system based on characteristic quantity prediction |
CN110348489A (en) * | 2019-06-19 | 2019-10-18 | 西安理工大学 | A kind of partial discharge of transformer mode identification method based on autoencoder network |
CN110632484A (en) * | 2019-11-06 | 2019-12-31 | 广东电网有限责任公司 | ELM-based GIS partial discharge defect diagnosis and classification system and method |
CN111028210A (en) * | 2019-11-25 | 2020-04-17 | 北京航天控制仪器研究所 | Deep neural network glass tube end surface defect detection method based on sparse automatic encoder |
CN111275109A (en) * | 2020-01-20 | 2020-06-12 | 国网山东省电力公司枣庄供电公司 | Power equipment state data characteristic optimization method and system based on self-encoder |
CN112036320A (en) * | 2020-09-07 | 2020-12-04 | 华北电力大学(保定) | Partial discharge state identification method and device, discharge simulator and identification equipment |
CN112287953A (en) * | 2019-07-24 | 2021-01-29 | 国网山东省电力公司济南供电公司 | Method and system for GIS insulation defect category identification |
CN112305379A (en) * | 2019-07-24 | 2021-02-02 | 国网山东省电力公司济南供电公司 | Mode identification method and system for GIS insulation defect |
CN112485394A (en) * | 2020-11-10 | 2021-03-12 | 浙江大学 | Water quality soft measurement method based on sparse self-coding and extreme learning machine |
CN112686182A (en) * | 2021-01-04 | 2021-04-20 | 华北电力大学(保定) | Partial discharge mode identification method and terminal equipment |
CN113506252A (en) * | 2021-06-29 | 2021-10-15 | 国家电网有限公司 | Transformer bushing typical defect type identification method based on t-SNE and nuclear extreme learning machine |
CN113780308A (en) * | 2021-08-27 | 2021-12-10 | 吉林省电力科学研究院有限公司 | GIS partial discharge mode identification method and system based on kernel principal component analysis and neural network |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105548832A (en) * | 2015-12-10 | 2016-05-04 | 国网四川省电力公司电力科学研究院 | High-voltage power cable fault recognition method |
CN106022465A (en) * | 2016-05-19 | 2016-10-12 | 江南大学 | Extreme learning machine method for improving artificial bee colony optimization |
CN107704883A (en) * | 2016-11-21 | 2018-02-16 | 东北大学 | A kind of sorting technique and system of the grade of magnesite ore |
-
2018
- 2018-07-30 CN CN201810852601.3A patent/CN109102012A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105548832A (en) * | 2015-12-10 | 2016-05-04 | 国网四川省电力公司电力科学研究院 | High-voltage power cable fault recognition method |
CN106022465A (en) * | 2016-05-19 | 2016-10-12 | 江南大学 | Extreme learning machine method for improving artificial bee colony optimization |
CN107704883A (en) * | 2016-11-21 | 2018-02-16 | 东北大学 | A kind of sorting technique and system of the grade of magnesite ore |
Non-Patent Citations (2)
Title |
---|
JU TANG ET.AL: "Assessment of PD severity in gas-insulated switchgear with an SSAE", 《IET SCIENCE, MEASUREMENT & TECHNOLOGY》 * |
张秦梫 等: "基于OS-ELM的变压器局部放电模式识别", 《高压电技术》 * |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109596955B (en) * | 2018-12-30 | 2021-06-22 | 国网北京市电力公司 | Partial discharge state determination method and device |
CN109596955A (en) * | 2018-12-30 | 2019-04-09 | 国网北京市电力公司 | Shelf depreciation state determines method and device |
CN110108754A (en) * | 2019-04-25 | 2019-08-09 | 四川沐迪圣科技有限公司 | The light stimulus infrared thermal imaging defect inspection method decomposed based on structural sparse |
CN110108754B (en) * | 2019-04-25 | 2021-10-22 | 四川沐迪圣科技有限公司 | Structured sparse decomposition-based light-excitation infrared thermal imaging defect detection method |
CN110286279B (en) * | 2019-06-05 | 2021-03-16 | 武汉大学 | Power electronic circuit fault diagnosis method based on extreme tree and stack type sparse self-coding algorithm |
CN110286279A (en) * | 2019-06-05 | 2019-09-27 | 武汉大学 | Based on extreme random forest and the sparse Fault Diagnosis of Power Electronic Circuits method from encryption algorithm of stacking-type |
CN110161389A (en) * | 2019-06-12 | 2019-08-23 | 国网河北省电力有限公司电力科学研究院 | A kind of Electric Power Equipment Insulation defect identification method and AEVB self-encoding encoder |
CN110348489A (en) * | 2019-06-19 | 2019-10-18 | 西安理工大学 | A kind of partial discharge of transformer mode identification method based on autoencoder network |
CN110348489B (en) * | 2019-06-19 | 2021-04-06 | 西安理工大学 | Transformer partial discharge mode identification method based on self-coding network |
CN110334948A (en) * | 2019-07-05 | 2019-10-15 | 上海交通大学 | Power equipment shelf depreciation Severity method and system based on characteristic quantity prediction |
CN110334948B (en) * | 2019-07-05 | 2023-04-07 | 上海交通大学 | Power equipment partial discharge severity evaluation method and system based on characteristic quantity prediction |
CN112287953A (en) * | 2019-07-24 | 2021-01-29 | 国网山东省电力公司济南供电公司 | Method and system for GIS insulation defect category identification |
CN112305379A (en) * | 2019-07-24 | 2021-02-02 | 国网山东省电力公司济南供电公司 | Mode identification method and system for GIS insulation defect |
CN110632484A (en) * | 2019-11-06 | 2019-12-31 | 广东电网有限责任公司 | ELM-based GIS partial discharge defect diagnosis and classification system and method |
CN111028210A (en) * | 2019-11-25 | 2020-04-17 | 北京航天控制仪器研究所 | Deep neural network glass tube end surface defect detection method based on sparse automatic encoder |
CN111028210B (en) * | 2019-11-25 | 2023-07-18 | 北京航天控制仪器研究所 | Glass tube end face defect detection method based on deep neural network |
CN111275109A (en) * | 2020-01-20 | 2020-06-12 | 国网山东省电力公司枣庄供电公司 | Power equipment state data characteristic optimization method and system based on self-encoder |
CN112036320A (en) * | 2020-09-07 | 2020-12-04 | 华北电力大学(保定) | Partial discharge state identification method and device, discharge simulator and identification equipment |
CN112485394A (en) * | 2020-11-10 | 2021-03-12 | 浙江大学 | Water quality soft measurement method based on sparse self-coding and extreme learning machine |
CN112686182A (en) * | 2021-01-04 | 2021-04-20 | 华北电力大学(保定) | Partial discharge mode identification method and terminal equipment |
CN112686182B (en) * | 2021-01-04 | 2023-12-26 | 华北电力大学(保定) | Partial discharge mode identification method and terminal equipment |
CN113506252A (en) * | 2021-06-29 | 2021-10-15 | 国家电网有限公司 | Transformer bushing typical defect type identification method based on t-SNE and nuclear extreme learning machine |
CN113780308A (en) * | 2021-08-27 | 2021-12-10 | 吉林省电力科学研究院有限公司 | GIS partial discharge mode identification method and system based on kernel principal component analysis and neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109102012A (en) | A kind of defect identification method and system of local discharge signal | |
CN109660206B (en) | Wasserstein GAN-based photovoltaic array fault diagnosis method | |
Manohar et al. | Spatio-temporal information based protection scheme for PV integrated microgrid under solar irradiance intermittency using deep convolutional neural network | |
CN110609477B (en) | Electric power system transient stability discrimination system and method based on deep learning | |
CN103076547A (en) | Method for identifying GIS (Gas Insulated Switchgear) local discharge fault type mode based on support vector machines | |
CN110726898B (en) | Power distribution network fault type identification method | |
CN110161388A (en) | A kind of the fault type recognition method and its system of high-tension apparatus | |
CN109164362A (en) | A kind of recognition methods and system of direct current cables shelf depreciation defect failure | |
CN110689069A (en) | Transformer fault type diagnosis method based on semi-supervised BP network | |
CN109002926A (en) | The photovoltaic power generation quantity prediction model and its construction method of a kind of high accuracy and application | |
CN103761537A (en) | Image classification method based on low-rank optimization feature dictionary model | |
CN110738232A (en) | grid voltage out-of-limit cause diagnosis method based on data mining technology | |
CN111313403A (en) | Low-voltage power distribution system network topology identification method based on Markov random field | |
CN112486137A (en) | Method and system for constructing fault feature library of active power distribution network and fault diagnosis method | |
CN111090747A (en) | Power communication fault emergency disposal method based on neural network classification | |
CN114169249A (en) | Artificial intelligence identification method for high-resistance grounding fault of power distribution network | |
CN112686404B (en) | Power distribution network fault first-aid repair-based collaborative optimization method | |
CN109344146A (en) | A kind of extensive overvoltage data intelligence method for self-cleaning | |
CN111898446A (en) | Single-phase earth fault studying and judging method based on multi-algorithm normalization analysis | |
CN111814834A (en) | High-voltage cable partial discharge mode identification method, computer equipment and storage medium | |
CN116400168A (en) | Power grid fault diagnosis method and system based on depth feature clustering | |
CN115128402A (en) | Power distribution network fault type identification method and system based on data driving | |
Liu et al. | A deep neural network pruning method based on gradient L1-norm | |
Wei et al. | Deep belief network based faulty feeder detection of single-phase ground fault | |
Fahim et al. | An agreement based dynamic routing method for fault diagnosis in power network with enhanced noise immunity |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181228 |