CN110428113A - A kind of predicting model for dissolved gas in transformer oil method based on random forest - Google Patents
A kind of predicting model for dissolved gas in transformer oil method based on random forest Download PDFInfo
- Publication number
- CN110428113A CN110428113A CN201910735129.XA CN201910735129A CN110428113A CN 110428113 A CN110428113 A CN 110428113A CN 201910735129 A CN201910735129 A CN 201910735129A CN 110428113 A CN110428113 A CN 110428113A
- Authority
- CN
- China
- Prior art keywords
- model
- random forest
- transformer oil
- gas
- dissolved
- 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
- 238000007637 random forest analysis Methods 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title claims abstract description 31
- 239000007789 gas Substances 0.000 claims abstract description 51
- 238000012360 testing method Methods 0.000 claims description 33
- 238000012549 training Methods 0.000 claims description 28
- 238000003066 decision tree Methods 0.000 claims description 12
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 11
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 claims description 9
- HSFWRNGVRCDJHI-UHFFFAOYSA-N alpha-acetylene Natural products C#C HSFWRNGVRCDJHI-UHFFFAOYSA-N 0.000 claims description 9
- OTMSDBZUPAUEDD-UHFFFAOYSA-N Ethane Chemical compound CC OTMSDBZUPAUEDD-UHFFFAOYSA-N 0.000 claims description 6
- 229910002092 carbon dioxide Inorganic materials 0.000 claims description 6
- 239000001569 carbon dioxide Substances 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 125000002534 ethynyl group Chemical group [H]C#C* 0.000 claims description 6
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 5
- VGGSQFUCUMXWEO-UHFFFAOYSA-N Ethene Chemical compound C=C VGGSQFUCUMXWEO-UHFFFAOYSA-N 0.000 claims description 4
- 239000005977 Ethylene Substances 0.000 claims description 4
- 229910052739 hydrogen Inorganic materials 0.000 claims description 4
- 239000001257 hydrogen Substances 0.000 claims description 4
- 238000004090 dissolution Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000013138 pruning Methods 0.000 claims description 3
- 238000011160 research Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 229960004424 carbon dioxide Drugs 0.000 claims 1
- 125000004435 hydrogen atom Chemical class [H]* 0.000 claims 1
- 238000012423 maintenance Methods 0.000 abstract description 6
- 238000012544 monitoring process Methods 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 5
- 229910002091 carbon monoxide Inorganic materials 0.000 description 4
- 238000012706 support-vector machine Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004587 chromatography analysis Methods 0.000 description 2
- 230000000052 comparative effect Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 150000002431 hydrogen Chemical class 0.000 description 2
- 230000010076 replication Effects 0.000 description 2
- 239000004215 Carbon black (E152) Substances 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- UFHFLCQGNIYNRP-UHFFFAOYSA-N Hydrogen Chemical compound [H][H] UFHFLCQGNIYNRP-UHFFFAOYSA-N 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 238000010292 electrical insulation Methods 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Resources & Organizations (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The predicting model for dissolved gas in transformer oil method based on random forest that this application discloses a kind of, for the content for accurately analyzing Gases Dissolved in Transformer Oil, transformer oil chromatographic online monitoring data is arranged first, is determined the associated arguments of concentration of dissolved gas and is normalized;Then historical events dimensional characteristics information architecture random forest network model is obtained to be trained;It finally is predicted as exporting by the input of model, characteristic gas concentration of various characteristic gas concentration, realizes the prediction to gas dissolved in oil of power trans-formers.Prediction result is assessed using average percentage error and maximum percentage error two indices.By analyzing gas prediction result, it can judge to provide foundation for the operation conditions of power transformer, offer reference for operation maintenance personnel maintenance.
Description
Technical field
This application involves electric apparatus monitoring field more particularly to a kind of predicting model for dissolved gas in transformer oil sides
Method.
Background technique
Currently, electric power networks have been developed as trans-regional interconnected power grid, power transformer is entire electric system
The hub device of core the most is the important equipment that electric energy is transmitted, distributed in electric system, is the critical asset of grid company,
Its safely and steadily run be power grid reliable power supply guarantee, will seriously affect the stable operation of power grid once breaking down.It is right
Oil dissolved gas concentration carry out analysis and prediction of the development trend, can for the operating status of transformer assessment provide it is important according to
According to.
Hydrogen H when power transformer generates high-energy discharge in gas2With acetylene C2H2Content increase, methane CH4And ethylene
C2H4Increase be due to the increase of built-in electrical insulation oil, in the case that system encounters strong electric field, the content of hydrocarbon gas all can
Increase, and can all show relevance, therefore can be the event of transformer by the analysis of Gases Dissolved in Transformer Oil (DGA)
Barrier, which differentiates, important help.Since traditional BP neural network and support vector machines (SVM) have, convergence rate is slow, network structure
The defects of different, the more classification problems of solution have difficulties, prediction effect is not scientific enough is selected, a kind of transformer solution gas is proposed
The prediction model of body.
Summary of the invention
For solve traditional BP neural network and support vector machines (SVM) method have the period it is long, it is complicated for operation, by people
Member's experience, the features such as error is big easily lead to delay and judge that the operating status of transformer in turn results in certain economic loss,
Not the problem of not being suitable for oil dissolved gas concentration prediction and analysis.The application provides technical solution below: one kind based on
The predicting model for dissolved gas in transformer oil method of machine forest, the parameter which needs to adjust is few, training effectiveness is high,
Predict that the accuracy rate of dissolved gas is higher, method includes:
Step 1: the associated arguments of gas dissolved in oil of power trans-formers are determined;
Step 2: the historical sample data of associated arguments is obtained;
Step 3: building random forest network model;
Step 4: the sample data training Random Forest model based on the historical time dimension determines the correlation of model
Parameter, and extract the characteristic information of the historical time dimension of corresponding sample data;
Step 5: the characteristic information of the characteristic information prediction future time dimension based on the historical time dimension;
Step 6: using housebroken Random Forest model in the step 4, to realize to being dissolved in transformer oil
The prediction of gas concentration.
Predicting model for dissolved gas in transformer oil method described herein based on random forest, the associated arguments
Refer to the parameter from each other with prediction interdependence effects, includes concentration of dissolved gas to be predicted.Due to the sample number
According on time dimension, distribution is with regularity, conducive to forecast analysis.
Predicting model for dissolved gas in transformer oil method described herein based on random forest is based on random gloomy
Forest network network predicts gas dissolved in oil of power trans-formers.
Further, in the predicting model for dissolved gas in transformer oil method described herein based on random forest,
The associated arguments include hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), carbon monoxide
(CO), carbon dioxide (CO2) concentration, can be indicated with a column vector, i.e.,WhereinIt is expressed as
N gas is in the concentration of ordinary dissolution of t moment, and in our current research, n value is 7, indicates that this model prediction is related to 7 kinds of gases.
In above scheme, offline oil chromatography sample data is normalized, is mapped the data between [0,1],
Transfer function are as follows:
Wherein xminFor sample data minimum value, xmaxFor sample data maximum value, x is the sample data before conversion, x*For
Sample data after conversion.
Random forest is a kind of combination integrated study class algorithm, by more CART (Classification and
Regression Tree) it constitutes, it can be indicated with the set of CART, it may be assumed that { h (X, θk) | k=1,2 ..., N }, X indicates input
Vector, θkIndicate generate k stalk tree, it be utilize Bootstrap repeat replication k sample is extracted from original sample, and
The capacity of each sample is as original training set;K sample establishes k decision-tree model respectively, and subtree all has phase
Same distribution, obtains k kind classification results;Finally, obtaining final prediction result according to k kind classification results statistics.
Random Forest model obtains the difference between subtree in terms of two:
(1) in data, the Bootstrap methods of sampling is utilized, sampling from original training set with putting back to, it is N number of only to generate
The vertical training dataset with distribution;
(2) in structure, when generating subtree, a subset is randomly selected out from feature set, the node of subtree is divided
It splits.Therefore, division random character subset capacity becomes the key parameter that RF model must determine in use.
The building process of random forest disaggregated model is as follows:
It using the Bootstrap methods of sampling, is concentrated from training data and extracts N number of training sample subset, form training set Si
I=1,2 ..., N };
For above-mentioned each training set, corresponding subtree CART is generated1、CART2、…、CARTN, include:
If training set has M dimension, Split Attribute collection of the F feature vector as present node is taken out at random from M attribute;
Using this F attribute as feature vector, which is divided, subtree complete growth is without beta pruning;
Using the performance of test data set test model, the predicted value CART of subtree output is obtained1(Test)、CART2
(Test)、…、CARTN(Test);
It takes in a manner of average value, the predicted value of statistics N decision tree output, and is averaged what all subtrees exported in advance
Measured value renormalization is as final predicted value.
In predicting model for dissolved gas in transformer oil method described herein based on random forest, subtree tree
(ntree) taking default value is 500, uses Forest-RI form, if training set has M dimension, randomly chooses F (F≤M) a feature
Vector carries out, if F acquirement is sufficiently small, the correlation between base decision tree tends to weaken therewith;Meanwhile the effect that base decision tree is integrated
Fruit is improved with the increase of F, is comprehensively considered, it usually needs empirically formula (2) determines disruptive features subset capacity
(mtry) F is 4.
F=1+log2d (2)
Wherein, d is 7 to be originally inputted characteristic.
The utility model has the advantages that in the application, the predicting model for dissolved gas in transformer oil method based on random forest, with various
Characteristic gas concentration is the input of model, is predicted as exporting with characteristic gas concentration, finally using average percentage error
The assessment of prediction result is carried out with maximum percentage error two indices.It can be more acurrate relative to traditional machine learning method
Oil dissolved gas concentration is predicted on ground, and the error of generation is minimum, can be judged to provide foundation for the operation conditions of power transformer, is
Operation maintenance personnel maintenance is offered reference.
Detailed description of the invention
In order to illustrate more clearly of the technical solution of the application, letter will be made to attached drawing needed in the embodiment below
Singly introduce, it should be apparent that, for those of ordinary skills, without any creative labor,
It is also possible to obtain other drawings based on these drawings.
Fig. 1 is the predicting model for dissolved gas in transformer oil method workflow described herein based on random forest
Schematic diagram;
Fig. 2 is to predict C in the embodiment of the present application2H2The result of concentration.
Specific embodiment
It is a kind of stream based on random forest predicting model for dissolved gas in transformer oil provided by the present application referring to Fig. 1
Cheng Tu, method include:
S01: power transformer decomposites a small amount of gas, predominantly hydrogen because of insulating oil and solid insulation ageing cracking etc.
(H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), carbon monoxide (CO), carbon dioxide (CO2) concentration
It is indicated with a column vector, i.e.,WhereinIt is dense in the dissolution of t moment to be expressed as n gas
Degree, in our current research, n value are 7, indicate that this model prediction is related to 7 kinds of gases.I.e. sample data has 171, and every has 7
Feature, using 80% (137) of sample data as training set, each column 20% (34) construct at random as test set
Forest prediction model.
Offline oil chromatography sample data is normalized, is mapped the data between [0,1], transfer function are as follows:
Wherein xminFor sample data minimum value, xmaxFor sample data maximum value, x is the sample data before conversion, x*For
Sample data after conversion.
S02: building random forest network model determines Parameters in Mathematical Model.
S021 detailed process is as follows:
It using the Bootstrap methods of sampling, is concentrated from training data and extracts N number of training sample subset, form training set Si
I=1,2 ..., N };
For above-mentioned each training set, corresponding subtree CART is generated1、CART2、…、CARTN, include:
(1) if training set has M dimension, Split Attribute of the F feature vector as present node is taken out at random from M attribute
Collection;
(2) using this F attribute as feature vector, which is divided, subtree complete growth is without beta pruning;It utilizes
The performance of test data set test model obtains the predicted value CART of subtree output1(Test)、CART2(Test)、…、CARTN
(Test);
S022: random forest is a kind of combination integrated study class algorithm, by more CART (Classification and
Regression Tree) it constitutes, it can be indicated with the set of CART, it may be assumed that { h (X, θk) | k=1,2 ..., N }, X indicates input
Vector, θkIndicate generate k stalk tree, it be utilize Bootstrap repeat replication k sample is extracted from original sample, and
The capacity of each sample is as original training set;K sample establishes k decision-tree model respectively, and subtree all has phase
Same distribution, obtains k kind classification results;Finally, obtaining final prediction result according to k kind classification results statistics.
S03: in Random Forest model, it is 500 that subtree tree (ntree), which takes default value, uses Forest-RI form,
If training set has M dimension, a feature vector of random selection F (F≤M) is carried out, if F acquirement is sufficiently small, therewith between base decision tree
Correlation tends to weaken;Meanwhile the effect that base decision tree integrates is improved with the increase of F, is comprehensively considered, it usually needs according to
Empirical equation (2) determines that disruptive features subset capacity (mtry) F is 4.
F=1+log2d (2)
Wherein, d is 7 to be originally inputted characteristic
S031: testing the performance of step model using test data set, obtains the predicted value CART of subtree output1
(Test)、CART2(Test)、…、CARTN(Test), it takes in a manner of average value, the prediction of statistics N decision tree output
Value, specific testing scheme are as follows: to the random forest prediction model that previous step constructs, input of 6 column therein as model is taken,
Output of the other column as model prediction result, this model is with C2H2Concentration for, to complete the transformation based on random forest
The prediction of device oil dissolved gas.
S032: using housebroken random forest network model in step S02, the feature letter based on future time dimension
Breath, rebuilds the associated arguments of future time dimension prediction data, to realize to gas dissolved in oil of power trans-formers
Prediction.It takes in a manner of average value, the predicted value of statistics N decision tree output, and the consensus forecast that all subtrees are exported
It is worth renormalization as final predicted value.
S04: by the final predicted value in step S032, using average percentage error and maximum percentage error two
A index, the test result of valuation prediction models, expression formula are as follows:
WhereinIndicate gas content value, xtIndicate that content true value, average percentage error and maximum percentage miss
The smaller expression forecast result of model of difference is better.
An embodiment is enumerated below.
With the oil colours in certain 220kV transformer oil chromatographic on-Line Monitor Device on December 29th, 11 days 1 July in 2018
For modal data, wherein monitoring cycle is 1 day, totally 171 groups of data.As shown in table 1 below, random forest regression model application and
In comparative test, it is used as training set by 137 groups in whole sample datas, remaining 34 groups are used as test set, and open up on this basis
Comparative test is opened, to verify the validity of this paper regression model.
1 training set of table and test set capacity
With dissolved acetylene (C in oil2H2) concentration prediction for, utilize RF method carry out model prediction result such as 2 institute of table
Show:
2 RF forecast of regression model C of table2H2The result of concentration
Known to analysis: attached drawing 2 intuitively shows the fitting degree of 34 test sample prediction results and true value.Test
The result shows that the average test relative error of RF regression model is 3.66%, full test relative error is 4.99%.Above-mentioned knot
Fruit shows that, for dissolved acetylene concentration prediction in oil, it can be power transformer that RF regression model, which has excellent and stable performance,
Operation conditions judgement provide foundation, for operation maintenance personnel maintenance offer reference.
Claims (7)
1. a kind of predicting model for dissolved gas in transformer oil method based on random forest, which is characterized in that the method packet
It includes:
Step 1: the associated arguments of gas dissolved in oil of power trans-formers are determined;
Step 2: the historical sample data of associated arguments is obtained;
Step 3: building random forest network model;
Step 4: the sample data training Random Forest model based on the historical time dimension determines the relevant parameter of model,
And extract the characteristic information of the historical time dimension of corresponding sample data;
Step 5: the characteristic information of the characteristic information prediction future time dimension based on the historical time dimension;
Step 6: using housebroken Random Forest model in the step 4, to realize to Gases Dissolved in Transformer Oil
The prediction of concentration.
2. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It is characterized in that, the associated arguments include hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), one
Carbonoxide (CO), carbon dioxide (CO2) concentration, can be indicated with a column vector, i.e., WhereinN gas is expressed as in the concentration of ordinary dissolution of t moment, in our current research, n value is 7, indicates that this model prediction is related to 7 kinds
Gas.
3. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It is characterized in that, the step 2 further includes that the sample data of the history dimension of the associated arguments is normalized, will
Data are mapped between [0,1], transfer function are as follows:
Wherein xminFor sample data minimum value, xmaxFor sample data maximum value, x is the sample data before conversion, x*After conversion
Sample data.
4. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It is characterized in that, the building process of the step 3 random forest network model is as follows:
It using the Bootstrap methods of sampling, is concentrated from training data and extracts N number of training sample subset, form training set SiI=1,
2,…,N};
For above-mentioned each training set, corresponding subtree CART is generated1、CART2、…、CARTN, include:
If training set has M dimension, Split Attribute collection of the F feature vector as present node is taken out at random from M attribute;With this
F attribute divides the node as feature vector, and subtree complete growth is without beta pruning.
5. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It being characterized in that, in the relevant parameter of model, it is 500 that subtree tree (ntree), which takes default value, Forest-RI form is used, if
Training set has a M dimension, and a feature vector of random selection F (F≤M) carries out, if F obtains sufficiently small, the phase between base decision tree therewith
Closing property tends to weaken;Meanwhile the effect that base decision tree integrates is improved with the increase of F, is comprehensively considered, it usually needs according to warp
It tests formula (2) and determines that disruptive features subset capacity (mtry) F is 4;
F=1+log2 d (2)
Wherein, d is 7 to be originally inputted characteristic.
6. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It is characterized in that, using the performance of test data set test model, obtains the predicted value CART of subtree output1(Test)、CART2
(Test)、…、CARTN(Test);
It takes in a manner of average value, the predicted value of statistics N decision tree output, and the mean predicted value that all subtrees are exported
Renormalization is as final predicted value.
7. a kind of predicting model for dissolved gas in transformer oil method based on random forest according to claim 1,
It is characterized in that, test prediction result is commented using average percentage error and maximum percentage error two indices
Valence, expression formula are as follows:
WhereinIndicate gas content value, xtIndicate content true value, average percentage error and maximum percentage error are got over
Small expression forecast result of model is better.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910735129.XA CN110428113A (en) | 2019-08-09 | 2019-08-09 | A kind of predicting model for dissolved gas in transformer oil method based on random forest |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910735129.XA CN110428113A (en) | 2019-08-09 | 2019-08-09 | A kind of predicting model for dissolved gas in transformer oil method based on random forest |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110428113A true CN110428113A (en) | 2019-11-08 |
Family
ID=68415341
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910735129.XA Pending CN110428113A (en) | 2019-08-09 | 2019-08-09 | A kind of predicting model for dissolved gas in transformer oil method based on random forest |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110428113A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111461263A (en) * | 2020-05-19 | 2020-07-28 | 昆明理工大学 | Method for predicting concentration of dissolved gas in transformer oil based on EMD-RF |
CN111722046A (en) * | 2020-07-01 | 2020-09-29 | 昆明理工大学 | Transformer fault diagnosis method based on deep forest model |
CN112016696A (en) * | 2020-08-14 | 2020-12-01 | 武汉大学 | PM integrating satellite observation and ground observation1Concentration inversion method and system |
CN113798315A (en) * | 2021-10-16 | 2021-12-17 | 北京航空航天大学 | Machine learning-based heat-strengthened SVE (singular value Environment) technology gas emission control method |
CN113869578A (en) * | 2021-09-26 | 2021-12-31 | 合肥通用机械研究院有限公司 | Intelligent prediction and diagnosis method for salt content of crude oil after removal of electric desalting system of atmospheric and vacuum distillation unit |
CN114184695A (en) * | 2021-11-09 | 2022-03-15 | 国网内蒙古东部电力有限公司电力科学研究院 | Parameter optimization-based method and system for predicting gas concentration in random forest oil |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104535865A (en) * | 2014-12-30 | 2015-04-22 | 西安工程大学 | Comprehensive diagnosing method for operation troubles of power transformer based on multiple parameters |
CN107545307A (en) * | 2017-07-28 | 2018-01-05 | 上海交通大学 | Predicting model for dissolved gas in transformer oil method and system based on depth belief network |
CN108537683A (en) * | 2018-04-13 | 2018-09-14 | 贵州电网有限责任公司 | A kind of load forecasting method based on similar day selection and random forests algorithm |
CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
-
2019
- 2019-08-09 CN CN201910735129.XA patent/CN110428113A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104535865A (en) * | 2014-12-30 | 2015-04-22 | 西安工程大学 | Comprehensive diagnosing method for operation troubles of power transformer based on multiple parameters |
CN107545307A (en) * | 2017-07-28 | 2018-01-05 | 上海交通大学 | Predicting model for dissolved gas in transformer oil method and system based on depth belief network |
CN108537683A (en) * | 2018-04-13 | 2018-09-14 | 贵州电网有限责任公司 | A kind of load forecasting method based on similar day selection and random forests algorithm |
CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
Non-Patent Citations (3)
Title |
---|
刘云鹏,许自强,董王英等: "基于经验模态分解和长短期记忆神经网络的变压器油中溶解气体浓度预测方法", 《中国电机工程学报》 * |
包灿: "基于随机森林算法的牵引变压器和可控高压电抗器保护研究", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
邵强: "基于随机森林的正例与未标注学习研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111461263A (en) * | 2020-05-19 | 2020-07-28 | 昆明理工大学 | Method for predicting concentration of dissolved gas in transformer oil based on EMD-RF |
CN111722046A (en) * | 2020-07-01 | 2020-09-29 | 昆明理工大学 | Transformer fault diagnosis method based on deep forest model |
CN111722046B (en) * | 2020-07-01 | 2021-05-18 | 昆明理工大学 | Transformer fault diagnosis method based on deep forest model |
CN112016696A (en) * | 2020-08-14 | 2020-12-01 | 武汉大学 | PM integrating satellite observation and ground observation1Concentration inversion method and system |
CN112016696B (en) * | 2020-08-14 | 2022-10-04 | 武汉大学 | PM integrating satellite observation and ground observation 1 Concentration inversion method and system |
CN113869578A (en) * | 2021-09-26 | 2021-12-31 | 合肥通用机械研究院有限公司 | Intelligent prediction and diagnosis method for salt content of crude oil after removal of electric desalting system of atmospheric and vacuum distillation unit |
CN113798315A (en) * | 2021-10-16 | 2021-12-17 | 北京航空航天大学 | Machine learning-based heat-strengthened SVE (singular value Environment) technology gas emission control method |
CN114184695A (en) * | 2021-11-09 | 2022-03-15 | 国网内蒙古东部电力有限公司电力科学研究院 | Parameter optimization-based method and system for predicting gas concentration in random forest oil |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110428113A (en) | A kind of predicting model for dissolved gas in transformer oil method based on random forest | |
CN109543737B (en) | Information system health degree evaluation method based on FAHP-FCA combined empowerment | |
CN107179503B (en) | Wind turbine generator fault intelligent diagnosis and early warning method based on random forest | |
CN110441500A (en) | A kind of predicting model for dissolved gas in transformer oil method based on shot and long term memory network | |
CN108832619A (en) | Transient stability evaluation in power system method based on convolutional neural networks | |
CN102928720B (en) | Defect rate detecting method of oil immersed type main transformer | |
CN109270390A (en) | Diagnosis Method of Transformer Faults based on Gaussian transformation Yu global optimizing SVM | |
CN110718910A (en) | Transient stability evaluation method for Bayesian optimization LightGBM | |
WO2012029154A1 (en) | 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 | |
CN110503305A (en) | A kind of transformer performance appraisal procedure | |
CN106597154A (en) | Transformer fault diagnosis lifting method based on DAG-SVM | |
de Aquino et al. | A fuzzy system for detection of incipient fault in power transformers based on gas-in-oil analysis | |
Ghosh et al. | Ensemble machine learning methods for better dynamic assessment of transformer status | |
CN116842433A (en) | Power distribution network fault classification method based on ensemble learning and longhorn beetle whisker algorithm | |
CN111239484A (en) | Non-invasive load electricity consumption information acquisition method for non-resident users | |
CN109523026A (en) | It cites a precedent inference method and system | |
Vilakazi et al. | Bushing fault detection and diagnosis using extension neural network | |
Wei et al. | Short text data model of secondary equipment faults in power grids based on LDA topic model and convolutional neural network | |
Qian et al. | Research on Power Transformer Fault Prediction Model Based on LSTM Neural Network | |
Chen et al. | Research and application of a real-time operation risk assessment method for power grid based on random forest algorithm | |
CN113535820A (en) | Electrical operating personnel attribute presumption method based on convolutional neural network | |
CN113283479A (en) | Feature extraction and diagnosis method suitable for power transformer faults | |
Li et al. | Research on the transformer fault diagnosis method based on LSTM artificial neural network and DGA | |
Chao et al. | Health assessment model of power transformer based on dissolved gas analysis by support vector machine | |
Zhang et al. | Research on main transformer defect detection methods based on Conditional Inference Tree and AdaBoost Algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191108 |
|
RJ01 | Rejection of invention patent application after publication |