CN109783485A - Distribution historical metrology data bearing calibration based on data mining and support vector machines - Google Patents
Distribution historical metrology data bearing calibration based on data mining and support vector machines Download PDFInfo
- Publication number
- CN109783485A CN109783485A CN201811646621.1A CN201811646621A CN109783485A CN 109783485 A CN109783485 A CN 109783485A CN 201811646621 A CN201811646621 A CN 201811646621A CN 109783485 A CN109783485 A CN 109783485A
- Authority
- CN
- China
- Prior art keywords
- data
- support vector
- regression model
- vector machines
- distribution
- 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
- 238000012706 support-vector machine Methods 0.000 title claims abstract description 52
- 238000007418 data mining Methods 0.000 title claims abstract description 19
- 238000012937 correction Methods 0.000 claims abstract description 14
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims abstract description 11
- 238000000034 method Methods 0.000 claims description 16
- 238000005457 optimization Methods 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 7
- 238000013480 data collection Methods 0.000 claims description 6
- 230000009977 dual effect Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000002950 deficient Effects 0.000 claims 1
- JEIPFZHSYJVQDO-UHFFFAOYSA-N ferric oxide Chemical compound O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 claims 1
- 230000000717 retained effect Effects 0.000 claims 1
- 230000006641 stabilisation Effects 0.000 abstract description 2
- 238000011105 stabilization Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 238000005065 mining Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The distribution historical metrology data bearing calibration based on data mining and support vector machines that the present invention relates to a kind of, its technical characterstic is: the following steps are included: step 1, the metric data read in distribution main website historical data base in a period of time record, the metric data such as voltage, electric current, active power, reactive power in historical data base being carried out sliding-model control；Step 2 passes through Apriori algorithm, excavates the Strong association rule of frequent item set in historical metrology data and picks out suspicious bad data；Step 3, building eliminate the data acquisition system of suspicious bad metric data, and the regression model of Training Support Vector Machines, substitute into regression model at the time of suspicious bad data is corresponded to, calculate match value, instead of suspicious bad data, complete Data correction.The present invention is conducive to distribution scheduling control decision, reduces the influence that bad data runs distribution, and then facilitates efficient electrical power distribution automatization system, stabilization, safety, intelligently runs.
Description
Technical field
The invention belongs to data mining technology fields, are related to distribution historical metrology data bearing calibration, are based especially on number
According to the distribution historical metrology data bearing calibration excavated with support vector machines.
Background technique
Currently, the historical data that distribution main website saves may be because a variety of causes there are bad data, bad data can shadow
The accuracy for ringing distribution scheduling control, is unfavorable for electrical power distribution automatization system and safely and steadily runs.
Existing least square method and neural network are also commonly used for data fitting, but the deviation of least square method fitting has
When can be excessive, generalization ability is not strong；For neural network in the case where data are few, fitting effect is very poor, when overabundance of data
The problem of there may be " overfittings ".
Therefore, there is an urgent need to data correcting methods appropriate carries out identification and school to the bad data in historical data base
Just, so as to which the control decision for distribution provides better support.
Summary of the invention
It is an object of the invention to overcome the deficiencies in the prior art, propose that a kind of design is reasonable, accuracy is high, working efficiency
Height improves the convenient and low-cost distribution historical metrology data bearing calibration based on data mining and support vector machines.
The present invention solves its realistic problem and adopts the following technical solutions to achieve:
A kind of distribution historical metrology data bearing calibration based on data mining and support vector machines, comprising the following steps:
Step 1, the metric data read in distribution main website historical data base in a period of time record, and to metric data base
Classify in time series, the sampling period is set, by voltage, electric current, active power, reactive power equivalent in historical data base
Measured data carries out sliding-model control, so as to subsequent carry out data mining.
Step 2, the pretreated data based on step 1 are excavated in historical metrology data frequently by Apriori algorithm
The Strong association rule of item collection, on the basis of Strong association rule, rescan database measurement record, pick out it is suspicious not
Good data；
Step 3, building eliminate the data acquisition system of suspicious bad metric data, based on treated data acquisition system, instruction
The regression model for practicing support vector machines, substitutes into regression model at the time of suspicious bad data is corresponded to, calculates match value, instead of
Suspicious bad data completes Data correction.
Moreover, the specific steps of the step 2 include:
(1) pretreated data are based on, data mining is carried out using Apriori algorithm；
(2) the initial support support threshold value of setting, scan data, by Calculate the support of every record；
Wherein, I indicates total transaction set, if support is lower than support threshold, excludes the transaction item, the item of reservation
Collection is used as frequent item set；
(3) confidence level of frequent item set is calculated by Confidence (X → Y)=P (Y | X)=P (X ∪ Y)/P (X), and
Judge whether the correlation rule of frequent item set has positive correlation in conjunction with promotion degree Lift (X → Y)=P (Y | X)/P (Y), if
Promotion degree is greater than 1, then retains the correlation rule of positive correlation, otherwise, correlation rule is abandoned；
(4) actual principle and institutional framework for combining distribution operation reject the Strong association rule for not meeting actual logic,
Retain reasonable Strong association rule；
(5) historical data base is scanned, is judged in conjunction with Strong association rule, finds out the data for not meeting Strong association rule
, it is classified as suspicious bad data.
Moreover, the specific steps of the step 3 include:
(1) it constructs one and eliminates the historical metrology data collection of suspicious bad data collection；
(2) selected part data return support vector machines as sample from the set for eliminating suspicious bad data
Return model to be trained, obtain the regression model of support vector machines, is fitted for data；
(3) at the time of correspondence based on suspicious bad data, by the regression model of support vector machines, match value is calculated,
Suspicious bad data is replaced with match value, completes Data correction.
Moreover, the specific method of step 3 (2) step is:
Measuring value { (x of the selected part based on time seriesi, yi), i=1,2 ..., N }, the training as support vector machines
Sample；
Wherein, xiFor input vector, yiFor output vector, SVM (Support Vector Machine) by f (x)=
ω φ (x)+b carries out function regression estimation；
Kernel function φ (x) may be selectedOriginal input space can be mapped
To new feature space, so that the sample of script linearly inseparable can divide in nuclear space；
Regression model optimization objective function beThe target of regression model is to make each point in training set most
Amount is fitted to linear model yi=w φ (xi)+b；
A constant offset ε (ε > 0) is defined in SVM, model output is calculated when being greater than ε with true output absolute value of the difference and damaged
It loses, the objective function of SVM regression model are as follows:
Regression model is to each sample (xi,yi) slack variable is introduced, it is defined asThe then loss function of regression model
Measurement becomes after slack variable is added:
Introduce Lagrange multiplierIt converts objective optimization function to without about
The form of beam, as follows:
The target of the primitive form of the objective function of SVM regression model is:
Optimization aim meets KKT condition, can pass through
Lagrange duality converts dual problem of equal value for optimization problem to solve:
First seek majorized function pairIt is minimum
Value, is acquired by partial derivative:
Above-mentioned relation formula is substituted intoIt eliminatesIt obtains:
It is available to objective function minimizing:
It can be found out with SMO algorithm correspondingAnd then system w, the b of regression model are found out, obtain support vector machines
Regression model.
The advantages of the present invention:
1, the distribution historical metrology data bearing calibration based on data mining and support vector machines that the invention discloses a kind of,
The identification that suspicious bad data is first carried out to historical metrology data is then based on and eliminates the set choosing of suspicious bad metric data
Take partial data as sample, the regression model of Training Support Vector Machines is used to correct, phase by means of Regression Model Simulator data
Compared with existing method, the accuracy and efficiency of correction can be improved.This method includes being recorded based on the measurement in historical data base,
Strong association rule is found using Apriori algorithm, rejects the correlation rule for obviously not meeting actual logic, is based on existing association
Rule, recognizes the suspicious bad data in historical data base, and building one eliminates the set of suspicious bad metric data.From picking
In addition to selected part data are based on back as sample, the regression model of Training Support Vector Machines in the set of suspicious bad data
Return model, calculate the match value that suspicious bad data corresponds to the moment, replaces suspicious data with match value, complete metric data
Correction.Historical metrology data is corrected, distribution scheduling control decision is conducive to, reduces the shadow that bad data runs distribution
It is loud, and then facilitate efficient electrical power distribution automatization system, stabilization, safety, intelligently run.
2, a kind of distribution historical metrology data bearing calibration based on data mining and support vector machines of the invention utilizes
Apriori algorithm is good at the characteristics of incidence relation between mining data, by excavating Strong association rule, can recognize well
Bad data in historical metrology data.Support vector machines is not in the case where data volume is very big available preferable time yet
Return effect, there is very strong generalization ability.Bad data is rejected from historical metrology data, selected part data as sample,
Can the regression models of more preferable Training Support Vector Machines pass through so that the accuracy and efficiency of data fitting all improves a lot
Historical metrology data unreasonable in distribution is recognized and corrected, and then is more high-quality to that can be provided for Dispatching Control System
The data of amount be used for Analysis of Policy Making, convenient for making more accurately scheduling controlling, facilitate ensure distribution it is safe and stable, it is efficient,
Intelligently run.
3, the present invention is the improvement carried out under conditions of existing network deployment, and it is convenient to improve, at low cost.
Detailed description of the invention
Fig. 1 is the process flow diagram of bearing calibration of the invention；
Fig. 2 is the process flow diagram of association rule mining and data identification of the invention；
Fig. 3 is the process flow diagram of the invention based on support vector machines correction data.
Specific embodiment
The embodiment of the present invention is described in further detail below in conjunction with attached drawing:
A kind of distribution historical metrology data bearing calibration based on data mining and support vector machines, as shown in Figure 1, including
Following steps:
Step 1, the metric data read in distribution main website historical data base in a period of time record, and to metric data base
Classify in time series, the sampling period is set, by voltage, electric current, active power, reactive power equivalent in historical data base
Measured data carries out sliding-model control, so as to subsequent carry out data mining.
Step 2, the pretreated data based on step 1 are excavated in historical metrology data frequently by Apriori algorithm
The Strong association rule of item collection, on the basis of Strong association rule, rescan database measurement record, pick out it is suspicious not
Good data；
Step 3, building eliminate the data acquisition system of suspicious bad metric data, based on treated data acquisition system, instruction
The regression model for practicing support vector machines, substitutes into regression model at the time of suspicious bad data is corresponded to, calculates match value, instead of
Suspicious bad data completes Data correction.
As shown in Fig. 2, the specific steps of the step 2 include:
(1) pretreated data are based on, data mining is carried out using Apriori algorithm；
(2) the initial support support threshold value of setting, scan data, by Calculate the support of every record；
Wherein, I indicates total transaction set, if support is lower than support threshold, excludes the transaction item, the item of reservation
Collection is used as frequent item set；
(3) confidence level of frequent item set is calculated by Confidence (X → Y)=P (Y | X)=P (X ∪ Y)/P (X), and
Judge whether the correlation rule of frequent item set has positive correlation in conjunction with promotion degree Lift (X → Y)=P (Y | X)/P (Y), if
Promotion degree is greater than 1, then retains the correlation rule of positive correlation, otherwise, correlation rule is abandoned；
(4) actual principle and institutional framework for combining distribution operation reject the Strong association rule for not meeting actual logic,
Retain reasonable Strong association rule；
(5) historical data base is scanned, is judged in conjunction with Strong association rule, finds out the data for not meeting Strong association rule
, it is classified as suspicious bad data.
As shown in figure 3, the specific steps of the step 3 include:
(1) it constructs one and eliminates the historical metrology data collection of suspicious bad data collection；
(2) selected part data return support vector machines as sample from the set for eliminating suspicious bad data
Return model to be trained, obtain the regression model of support vector machines, is fitted for data；
The specific method of step 3 (2) step is: measuring value { (x of the selected part based on time seriesi,yi), i=
1,2 ..., N }, the training sample as support vector machines；
Wherein, xiFor input vector, yiFor output vector, SVM (Support Vector Machine) by f (x)=
ω φ (x)+b carries out function regression estimation；
Kernel function φ (x) may be selectedOriginal input space can be mapped
To new feature space, so that the sample of script linearly inseparable can divide in nuclear space；
Regression model optimization objective function beThe target of regression model is to allow train each point in gathering
It is fitted to linear model y as far as possiblei=w φ (xi)+b；
A constant offset ε (ε > 0) is defined in SVM, model output is calculated when being greater than ε with true output absolute value of the difference and damaged
It loses, the objective function of SVM regression model are as follows:
Regression model is to each sample (xi,yi) slack variable is introduced, it is defined asThe then loss function of regression model
Measurement becomes after slack variable is added:
Introduce Lagrange multiplierIt converts objective optimization function to without about
The form of beam, as follows:
The target of the primitive form of the objective function of SVM regression model is:
Optimization aim meets KKT condition, can pass through glug
Bright day antithesis converts dual problem of equal value for optimization problem to solve:
First seek majorized function pairMinimum,
It is acquired by partial derivative:
Above-mentioned relation formula is substituted intoIt eliminatesIt obtains:
It is available to objective function minimizing:
It can be found out with SMO algorithm correspondingAnd then system w, the b of regression model are found out, obtain support vector machines
Regression model.
(3) at the time of correspondence based on suspicious bad data, by the regression model of support vector machines, match value is calculated,
Suspicious bad data is replaced with match value, completes Data correction.
The working principle of the invention is:
The present invention relates to data mining, data prediction, correlation rule frequent item set mining, Support vector regression models
Etc. multiple technologies and theory.The present invention utilizes Apriori algorithm, excavates the strong association rule of frequent item set in historical metrology data
Then, it is based on Strong association rule, the suspicious data in historical metrology data is screened, suspicious data is picked from historical metrology data
It removes, based on treated historical metrology data, it is quasi- to carry out data using regression model for the regression model of Training Support Vector Machines
It closes, brings regression model at the time of then corresponding to suspicious bad data and calculate match value, replace suspicious data with match value,
Complete Data correction.Historical metrology data is recognized and is corrected, is conducive to the scheduling controlling for optimizing distribution, promotes distribution
Automatization level.
It is emphasized that embodiment of the present invention be it is illustrative, without being restrictive, therefore the present invention includes
It is not limited to embodiment described in specific embodiment, it is all to be obtained according to the technique and scheme of the present invention by those skilled in the art
Other embodiments, also belong to the scope of protection of the invention.
Claims (4)
1. a kind of distribution historical metrology data bearing calibration based on data mining and support vector machines, it is characterised in that: including
Following steps:
Step 1 reads metric data record in distribution main website historical data base in a period of time, and when being based on to metric data
Between sequence classify, be arranged the sampling period, by historical data base voltage, electric current, active power, reactive power etc. measure number
According to progress sliding-model control；
Step 2, the pretreated data based on step 1 excavate frequent item set in historical metrology data by Apriori algorithm
Strong association rule, on the basis of Strong association rule, rescan database measurement record, pick out suspicious umber of defectives
According to；
Step 3, building eliminate the data acquisition system of suspicious bad metric data, based on treated data acquisition system, training branch
The regression model for holding vector machine substitutes into regression model at the time of corresponding to suspicious bad data, calculates match value, instead of suspicious
Bad data, complete Data correction.
2. a kind of distribution historical metrology data correction side based on data mining and support vector machines according to claim 1
Method, it is characterised in that: the specific steps of the step 2 include:
(1) pretreated data are based on, data mining is carried out using Apriori algorithm；
(2) the initial support support threshold value of setting, scan data, by Calculate the support of every record；
Wherein, I indicates total transaction set, if support is lower than support threshold, excludes the transaction item, the item collection of reservation is made
For frequent item set；
(3) by Confidence (X → Y)=P (Y | X) ,/P (X) calculates the confidence level of frequent item set to=P (X ∪ Y), and combines
Promotion degree Lift (X → Y)=P (Y | X)/P (Y) judges whether the correlation rule of frequent item set has positive correlation, if promoted
Degree is greater than 1, then retains the correlation rule of positive correlation, otherwise, correlation rule is abandoned；
(4) Strong association rule for not meeting actual logic is rejected, is retained by the actual principle and institutional framework for combining distribution operation
Reasonable Strong association rule；
(5) historical data base is scanned, is judged in conjunction with Strong association rule, finds out the data item for not meeting Strong association rule, is arranged
For suspicious bad data.
3. a kind of distribution historical metrology data correction side based on data mining and support vector machines according to claim 1
Method, it is characterised in that: the specific steps of the step 3 include:
(1) it constructs one and eliminates the historical metrology data collection of suspicious bad data collection；
(2) selected part data are as sample from the set for eliminating suspicious bad data, to the recurrence mould of support vector machines
Type is trained, and obtains the regression model of support vector machines, is fitted for data；
(3) based on suspicious bad data correspond at the time of, by the regression model of support vector machines, calculate match value, with intend
Conjunction value replaces suspicious bad data, completes Data correction.
4. a kind of distribution historical metrology data correction side based on data mining and support vector machines according to claim 3
Method, it is characterised in that: the specific method of step 3 (2) step is:
Measuring value { (x of the selected part based on time seriesi,yi), i=1,2 ..., N }, the training sample as support vector machines
This；
Wherein, xiFor input vector, yiFor output vector, SVM (Support Vector Machine) passes through f (x)=ω φ
(x)+b carries out function regression estimation；
Kernel function φ (x) may be selectedOriginal input space can be mapped to new
Feature space, so that the sample of script linearly inseparable can divide in nuclear space；
Regression model optimization objective function beThe target of regression model is to make each point in training set quasi- as far as possible
Close linear model yi=w φ (xi)+b；
A constant offset ε (ε > 0) is defined in SVM, model output is calculated when being greater than ε with true output absolute value of the difference and is lost,
The objective function of SVM regression model are as follows:
Regression model is to each sample (xi,yi) slack variable is introduced, it is defined as ξi,The then loss function measurement of regression model
Become after slack variable is added:
Introduce Lagrange multiplierIt converts objective optimization function to unconfined
Form, as follows:
The target of the primitive form of the objective function of SVM regression model is:
Optimization aim meets KKT condition, can pass through Lagrange
Antithesis converts dual problem of equal value for optimization problem to solve:
First ask majorized function to w, b, ξi,It is minimum
Value, is acquired by partial derivative:
Above-mentioned relation formula is substituted intoEliminate w, b, ξi,It obtains:
It is available to objective function minimizing:
Corresponding α can be found out with SMO algorithmi,And then system w, the b of regression model are found out, obtain returning for support vector machines
Return model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811646621.1A CN109783485A (en) | 2018-12-30 | 2018-12-30 | Distribution historical metrology data bearing calibration based on data mining and support vector machines |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811646621.1A CN109783485A (en) | 2018-12-30 | 2018-12-30 | Distribution historical metrology data bearing calibration based on data mining and support vector machines |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109783485A true CN109783485A (en) | 2019-05-21 |
Family
ID=66499681
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811646621.1A Pending CN109783485A (en) | 2018-12-30 | 2018-12-30 | Distribution historical metrology data bearing calibration based on data mining and support vector machines |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109783485A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111614616A (en) * | 2020-04-17 | 2020-09-01 | 江苏信息职业技术学院 | XSS attack automatic detection method |
| CN111679634A (en) * | 2020-01-20 | 2020-09-18 | 武汉裕大华纺织有限公司 | Intelligent roving management system |
| CN113010508A (en) * | 2021-03-23 | 2021-06-22 | 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 | Power metering data quality monitoring method and device and big data computing platform |
| CN113704321A (en) * | 2021-08-11 | 2021-11-26 | 国电南瑞科技股份有限公司 | Power grid abnormal data identification method, device and system |
| CN116596674A (en) * | 2023-07-18 | 2023-08-15 | 山东省标准化研究院(Wto/Tbt山东咨询工作站) | External trade risk assessment method based on big data analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473480A (en) * | 2013-10-08 | 2013-12-25 | 武汉大学 | Online monitoring data correction method based on improved universal gravitation support vector machine |
| CN103617459A (en) * | 2013-12-06 | 2014-03-05 | 李敬泉 | Commodity demand information prediction method under multiple influence factors |
| CN107067089A (en) * | 2016-08-27 | 2017-08-18 | 薛晓辉 | Efficient power grid data processing method |
| CN108510201A (en) * | 2018-04-08 | 2018-09-07 | 上海金智晟东电力科技有限公司 | Establish the data mining framework algorithm of distribution automation remote control index |
| CN108694470A (en) * | 2018-06-12 | 2018-10-23 | 天津大学 | A kind of data predication method and device based on artificial intelligence |
-
2018
- 2018-12-30 CN CN201811646621.1A patent/CN109783485A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473480A (en) * | 2013-10-08 | 2013-12-25 | 武汉大学 | Online monitoring data correction method based on improved universal gravitation support vector machine |
| CN103617459A (en) * | 2013-12-06 | 2014-03-05 | 李敬泉 | Commodity demand information prediction method under multiple influence factors |
| CN107067089A (en) * | 2016-08-27 | 2017-08-18 | 薛晓辉 | Efficient power grid data processing method |
| CN108510201A (en) * | 2018-04-08 | 2018-09-07 | 上海金智晟东电力科技有限公司 | Establish the data mining framework algorithm of distribution automation remote control index |
| CN108694470A (en) * | 2018-06-12 | 2018-10-23 | 天津大学 | A kind of data predication method and device based on artificial intelligence |
Non-Patent Citations (2)
| Title |
|---|
| 吴越嬴: "基于数据挖掘的电力系统不良数据检测与辨识算法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
| 杨慧霞: "含有历史不良数据的电力负荷预测研究", 《电力系统保护与控制》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111679634A (en) * | 2020-01-20 | 2020-09-18 | 武汉裕大华纺织有限公司 | Intelligent roving management system |
| CN111614616A (en) * | 2020-04-17 | 2020-09-01 | 江苏信息职业技术学院 | XSS attack automatic detection method |
| CN113010508A (en) * | 2021-03-23 | 2021-06-22 | 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 | Power metering data quality monitoring method and device and big data computing platform |
| CN113704321A (en) * | 2021-08-11 | 2021-11-26 | 国电南瑞科技股份有限公司 | Power grid abnormal data identification method, device and system |
| CN116596674A (en) * | 2023-07-18 | 2023-08-15 | 山东省标准化研究院(Wto/Tbt山东咨询工作站) | External trade risk assessment method based on big data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109783485A (en) | Distribution historical metrology data bearing calibration based on data mining and support vector machines | |
| CN103530347B (en) | A kind of Internet resources method for evaluating quality based on big data mining and system | |
| CN101795090A (en) | Method of forecasting the electrical production of a photovoltaic device | |
| CN107185854B (en) | The algorithm of photovoltaic cell acetes chinensis and color classification based on RGB channel | |
| CN110659693B (en) | K-nearest neighbor classification-based power distribution network rapid topology identification method, system and medium | |
| CN108090515B (en) | Data fusion-based environment grade evaluation method | |
| CN106525194A (en) | Main transformer oil level real-time monitoring method based on ultimate oil temperature equivalence method | |
| CN115049163B (en) | Photovoltaic region prediction method and device based on tabu search and deep stable learning | |
| CN117411189B (en) | Monitoring data enhancement method of micro-grid coordination controller | |
| CN111242161A (en) | Non-invasive non-resident user load identification method based on intelligent learning | |
| CN104166999A (en) | Cloud cluster extracting method based on strength layering of foundation cloud pictures | |
| CN108335233A (en) | A kind of intelligent grid metric data processing system based on big data technology | |
| CN116800172B (en) | Photoelectric conversion method and energy storage system | |
| Feng et al. | Identification of disturbance sources based on random forest model | |
| CN113435733A (en) | Water resource risk assessment and prediction system | |
| CN109472813A (en) | It is a kind of based on background weighting Mean Shift algorithm and Kalman prediction fusion block tracking | |
| CN110460085A (en) | A method of consider wind-powered electricity generation and part throttle characteristics to effect on power system | |
| CN107515374B (en) | Method for dynamically adjusting filtering gain applied to AGV SOC estimation | |
| CN116249186A (en) | Data processing method and device of wireless network equipment, storage medium and electronic equipment | |
| CN114818827A (en) | Non-invasive load decomposition method based on seq2point network | |
| CN109388845A (en) | Based on backward learning and the complicated photovoltaic array parameter extracting method evolved of enhancing | |
| CN106408433A (en) | Grain yield prediction method and grain yield prediction device based on differential processing | |
| CN112199854A (en) | Method for constructing efficiency analysis model of power industry | |
| CN111612237A (en) | Load space-time distribution prediction method and system for power supply grid | |
| CN116826977B (en) | Intelligent management system for optical storage direct-flexible micro-grid |
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: 20190521 |
|
| RJ01 | Rejection of invention patent application after publication |