CN107784015A - A kind of Data Reduction method based on the online historical data of power system - Google Patents
A kind of Data Reduction method based on the online historical data of power system Download PDFInfo
- Publication number
- CN107784015A CN107784015A CN201610771830.3A CN201610771830A CN107784015A CN 107784015 A CN107784015 A CN 107784015A CN 201610771830 A CN201610771830 A CN 201610771830A CN 107784015 A CN107784015 A CN 107784015A
- Authority
- CN
- China
- Prior art keywords
- sample
- mrow
- msub
- data
- power system
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 230000009467 reduction Effects 0.000 title claims abstract description 17
- 238000012216 screening Methods 0.000 claims abstract description 26
- 238000000605 extraction Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000003068 static effect Effects 0.000 claims description 6
- 238000013480 data collection Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 description 11
- 238000004364 calculation method Methods 0.000 description 4
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000006641 stabilisation Effects 0.000 description 2
- 238000011105 stabilization Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Human Resources & Organizations (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a kind of Data Reduction method based on the online historical data of power system, the method includes:I, raw data set is generated according to online historical data;II, feature extraction is carried out to raw data set;III, screening sample is carried out to raw data set.Technical scheme provided by the invention from the horizontal and vertical of data set while can carry out Data Reduction, especially in terms of screening sample, can screen out influences the noise sample of classification, and can screens out does not have contributive redundant samples to classification, therefore while classification accuracy is kept, space filling rate is greatly reduced again.
Description
Technical field
The invention belongs to power system big data analysis technical field, in particular to one kind is gone through online based on power system
The Data Reduction method of history data.
Background technology
With gradually forming for alternating current-direct current series-parallel connection bulk power grid general layout, the complexity more and more higher of power network, this gives power system
Safe and stable operation adds difficulty.Based on this case, it is necessary to constantly extend on-line security and stability analysis function, constantly meet
The technical need that bulk power grid fast development works management and running.
For big system in line computation, the historical data scale of construction accumulated is extremely huge, and elapses can also over time
It is quick to produce.Substantial amounts of management and running data are have accumulated before regulation and control centraleses at different levels, a safety on line was carried out every 15 minutes
Stability analysis, the data of calculating and the data volume of result can be annual enough of about 1G, the data volume about 96G of daily 96 points
The huge data volume of TB magnitudes can be reached.Made under the conditions of existing software and hardware of these data and quick sentence stability analysis and show
By the way of so can not simply taking " all use ", it is necessary to take into account " big " and " fast " both sides feature.The processing of big data
Thought and technology provide technical support for the analysis of a large amount of online historical datas.
Therefore need to provide a kind of data to the deficiency of research in terms of online historical data processing for current power system
The data preprocessing method of yojan reduces the time of data training, improves data and dig to reduce data scale, reduce memory space
The efficiency of pick;Remove the noise data in data, improve data classification accuracy.
The content of the invention
For a large amount of online historical datas, the invention provides a kind of data based on the online historical data of power system about
Simple method implementation method.
A kind of Data Reduction method based on the online historical data of power system, it is characterised in that methods described includes:
I, raw data set is generated according to online historical data;
II, feature extraction is carried out to raw data set;
III, screening sample is carried out to raw data set.
Further, historical data includes described in the step I:Power network shape as the attribute variable of analyze data collection
The critical clearing time of state variable and class variable as analyze data collection.
Further, the electric network state variable includes the quantity of state and electrical quantity of static device, by region or plant stand
Statistic.
Further, feature extraction includes described in the step II:Extraction to class variable it is moderate related and with
On feature of the attribute variable as subsequent analysis, remove weak related and unrelated attribute variable, using shown in following formula
Pearson correlation coefficient γA,BCarry out feature extraction:
Wherein:
The number of n-tuple;
aiNumerical value of-tuple the i on A;
biNumerical value of-tuple the i on B;
A-A average;
B-B average;
σA- A standard deviation;
σB- B standard deviation.
Further, the screening of sample includes described in the step III:
(1) repeatability screening;
(2) screening sample based on SVM algorithm;
(3) screening sample based on nearest sample pair.
Further, the step (1) includes:Identical sample is deleted, only retains one of sample.
Further, the step (2) includes:
Sample classification training is carried out with SVM algorithm, the classification function being shown below:
Wherein, sgn is sign function, and K is kernel function, and a and b are the parameters obtained by training, xiAnd yiRespectively support to
Amount and its classification (yi=1 or -1), x is sample to be discriminated;
Kernel function K is shown below:
Wherein, σ is parameter.
Further, the screening sample based on SVM algorithm includes:
Training sample by SVM methods, is obtained hyperplane by (4-1);
(4-2) inputs training sample as sample to be sorted, obtains classification results;
(4-3) deletes the sample of classification error in training sample;
(4-4) repeats above step, until the sample of no classification error.
Further, the sample recently is shown to being defined as follows:
It is β that the nearest samples of distance sample α, which are calculated, using Euclidean distance formula, if sample β nearest sample is α,
Then sample α and sample β nearest sample pair each other.
Further, the screening based on nearest sample pair includes:
(5-1) calculates the nearest sample x ' of each sample x in data set S, labeled as nearest sample;
The nearest sample of all samples in (5-2) mark S;
(5-3) finds the sample of all samples pair nearest each other in S;
(5-4) judges whether each sample centering classification is identical, if identical, deletes one of sample;If differing,
Then two samples are all deleted;
The S that (5-5) is obtained is data set after compressing.
With immediate prior art ratio, technical scheme provided by the invention has following excellent effect:
1st, algorithm of the invention embodies the characteristics of carrying out analysis work using the online historical data of magnanimity.In order to improve meter
The complexity of calculation to big measure feature and sample, it is necessary to effectively be screened, to reach the more accurate purpose of data analysis faster.
2nd, the present invention from the horizontal and vertical of data set while can carry out Data Reduction, especially in terms of screening sample,
Can screen out influences the noise sample of classification, and and can screens out does not have contributive redundant samples to classification, is keeping classification accuracy
While, greatly reduce space filling rate.
Brief description of the drawings
Fig. 1 is that correlation coefficient threshold of the present invention chooses contrast.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, the present invention is carried out below further
Describe in detail.
The method of the invention generates according to following steps:
Step 1, the online historical data of magnanimity is chosen, generate raw data set.
The main attribute variable for choosing electric network state variable as analyze data collection, including under static state each equipment quantity of state
And electrical quantity, and the statistic carried out by region or plant stand.Why equipment variables static state under are selected, and main cause is uncommon
Hope to shorten as much as possible and sentence the steady time, if the variable in selection transient process, need the temporary steady emulation meter of a period of time after all
Calculate, can so reduce and sentence steady speed.Because the relation between static variable and stabilization of power grids degree there is no clear and definite final conclusion, because
This, this problem more widely chooses static variable, the electric network state mainly included as far as possible on the premise of computing capability allows
Amount and statistic are as shown in table 1.
The electric network state amount of table 1 and statistic list
Class variable of the critical clearing time (CCT, critical clearing time) as analyze data collection is chosen,
Critical clearing time represents the stable and unstable border of system, occurs to specify the stabilization of failure available for power system is characterized
Degree, critical clearing time is bigger, represents that the short trouble is smaller to systematic influence, system is more stable.If critical excision
Time is less than normal operating time of protection, then illustrates that the failure can cause system unstability, i.e. system has potential safety hazard.
All samples for choosing certain month 2015 ultimately generate 2484 samples, each sample bag as raw data set
Include 9815 attribute variables.
Step 2, feature extraction is carried out using coefficient correlation.
Coefficient correlation is used to investigate the degree of correlation between two things (we term it variable in data).It is most normal
Method is to calculate Pearson correlation coefficient, and calculation formula is as follows.
In formula:
The number of n-tuple;
aiNumerical value of-tuple the i on A;
biNumerical value of-tuple the i on B;
- A average;
- B average;
σA- A standard deviation;
σB- B standard deviation.
Coefficient correlation value between -1 to+1, more than 0 interval scale, two sequence positive correlations, i.e. A values with the increase of B values and
Increase, numerical value is closer to+1, then degree of correlation is stronger;Otherwise it is less than 0 interval scale, two sequence negative correlation, closer to -1,
Then degree of correlation is stronger;It is uncorrelated equal to 0 interval scale, two sequences.
The critical clearing time after certain line fault is chosen as class variable, calculates all properties variable and class variable
Coefficient correlation.
Extract to the variable that class variable is moderate related and attribute variable of the above is as subsequent analysis, remove weak phase
Close and unrelated attribute variable.Contrast separately below using coefficient correlation as 0.4,0.5 and 0.6 as threshold value when, variable number with
And the situation of classification accuracy is as shown in Figure 1.As can be seen that as the increase of correlation coefficient threshold, the number of attribute variable are anxious
Reduce sharply less, but classification accuracy situation of change is more complicated.Finally draw, it is relatively fewer and classification is accurate with attribute variable's number
Rate is up to target, and extraction coefficient correlation is the attribute variable of 0.5 and the above, and final 574 attribute variables of extraction carry out follow-up
Analysis.
Step 3, the data set obtained according to step 2, carry out the repeatability screening of first time sample.
The operation of power system is present periodically, and this periodicity is relatively stable in a short time, and drawing over time
Long, system operation mode inevitably produces increasing difference.Raw data set is after feature extraction, due to attribute
Variable greatly reduces, it may appear that the situation that differences between samples disappear, the feature for causing sample similar in the time to include are identical.
At this moment, it is necessary to carry out carrying out repeated screening to sample, a reservation is selected in identical sample, and by remaining identical sample
All delete.
Step 4, the data set obtained according to step 3, second of screening sample is carried out using SVM algorithm.
Sample classification training is carried out using SVM algorithm, final classification function can be write:
Wherein, sgn is that for the final classification for determining sample, (the positive and negative of such as result of calculation corresponds respectively to sign function
Two classification results), K is kernel function, and α and b are the parameters obtained by training, xiAnd yiFor supporting vector and its classification (yi=1
Or -1), x is sample to be discriminated.
Kernel function K uses Gaussian form (RBF):
σ therein is parameter.
In the training process, in addition to it need to specify the parameter of kernel function, generally also need to specify a coefficient of relaxation c, i.e.,
The degree for allowing classification to malfunction, for strengthening SVM generalization ability.
Need it is clear that, in hands-on model, if based on the data analysis tool of maturation, such as R language or
Matlab, its disaggregated model training function provided are usually constructed with default parameters.For SVM, usual σ value be characterized to
The inverse of amount, c are often taken as 1.
Screening sample method based on SVM algorithm is:
(1) using training sample S training SVM, hyperplane is obtained;
(2) inputted training sample S as classification samples, obtain classification results R;
(3) sample of classification error is deleted from S;
(4) repeat step (1)-(3), untill no misclassification sample.
This method flow is the continuous process for removing sample, and this method can be used for removing noise sample so that classification boundaries
Become apparent from.
Step 5, the data set obtained according to step 4, using nearest sample to carrying out third time screening sample.
The present invention proposes the concept of nearest sample pair, is defined as:For a sample a in data set, using Euclidean away from
From formula, the nearest samples of distance a (being set to b) are calculated, if b nearest sample is also a, claim a and b nearest samples pair each other.
Screening sample method based on nearest sample pair is:
(1) the nearest sample x ' of each sample x in data set S is calculated, labeled as nearest sample;
(2) next sample in S is continued with, turns to step (1), until the sample in S is all labeled;
(3) S is traveled through, finds all samples nearest each other in S;
(4) whether each sample centering classification searched out in judgment step (3) is identical, if identical, by any one
Sample is deleted;If differing, two samples are all deleted;
(5) S obtained is data set after compressing.
This method can be used for removing noise sample, can also remove the sample of redundancy in class.
Sample calculation analysis is as follows:
Data set of the initial data after feature extraction is described as follows:
Sample number | Characteristic (after screening) | Classification number |
2484 | 574 | 9 |
80% sample is randomly selected in each classification respectively as training set, remaining sample is as test set.And
There is provided 10 kinds of random numbers, try to achieve the final result averaged after result as this method.Following table gives screening sample side
The implementation effect of method, classification accuracy and space filling rate before and after screening sample are compared for, it is as shown in the table.
It can be drawn by upper table, the data set after the algorithm process is more former on classification accuracy and storage compression ratio
Beginning sample set is obviously improved.Improved while classification accuracy is kept and averagely store compression ratio more than 35%, and completely protected
The data distribution characteristics of original sample collection are stayed, compromise is obtained on classification accuracy and storage compression ratio.
Finally it should be noted that:The above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof, to the greatest extent
The present invention is described in detail with reference to above-described embodiment for pipe, those of ordinary skills in the art should understand that:Still
The embodiment of the present invention can be modified or equivalent substitution, and without departing from any of spirit and scope of the invention
Modification or equivalent substitution, it all should cover among scope of the presently claimed invention.
Claims (10)
- A kind of 1. Data Reduction method based on the online historical data of power system, it is characterised in that methods described includes:I, raw data set is generated according to online historical data;II, feature extraction is carried out to raw data set;III, screening sample is carried out to raw data set.
- A kind of 2. Data Reduction method based on the online historical data of power system as claimed in claim 1, it is characterised in that Historical data includes described in the step I:As the electric network state variable of the attribute variable of analyze data collection and as analysis The critical clearing time of the class variable of data set.
- A kind of 3. Data Reduction method based on the online historical data of power system as claimed in claim 2, it is characterised in that The electric network state variable includes the quantity of state and electrical quantity of static device, by region or the statistic of plant stand.
- A kind of 4. Data Reduction method based on the online historical data of power system as claimed in claim 1, it is characterised in that Feature extraction includes described in the step II:Utilize the Pearson correlation coefficient γ shown in following formulaA,BCarry out feature extraction:<mrow> <msub> <mi>&gamma;</mi> <mrow> <mi>A</mi> <mo>,</mo> <mi>B</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <msubsup> <mi>&Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </msubsup> <mrow> <mo>(</mo> <msub> <mi>a</mi> <mi>i</mi> </msub> <mo>-</mo> <mover> <mi>A</mi> <mo>&OverBar;</mo> </mover> <mo>)</mo> </mrow> <mrow> <mo>(</mo> <msub> <mi>b</mi> <mi>i</mi> </msub> <mo>-</mo> <mover> <mi>B</mi> <mo>&OverBar;</mo> </mover> <mo>)</mo> </mrow> </mrow> <mrow> <msub> <mi>n&sigma;</mi> <mi>A</mi> </msub> <msub> <mi>&sigma;</mi> <mi>B</mi> </msub> </mrow> </mfrac> </mrow>Wherein:The number of n-tuple;aiNumerical value of-tuple the i on A;biNumerical value of-tuple the i on B;- A average;- B average;σA- A standard deviation;σB- B standard deviation.
- A kind of 5. Data Reduction method based on the online historical data of power system as claimed in claim 1, it is characterised in that The screening of sample includes described in the step III:(1) repeatability screening;(2) screening sample based on SVM algorithm;(3) screening sample based on nearest sample pair.
- A kind of 6. Data Reduction method based on the online historical data of power system as claimed in claim 5, it is characterised in that The step (1) includes:Identical sample is deleted, only retains one of sample.
- A kind of 7. Data Reduction method based on the online historical data of power system as claimed in claim 5, it is characterised in that The step (2) includes:Sample classification training is carried out with SVM algorithm, the classification function being shown below:<mrow> <mi>f</mi> <mrow> <mo>(</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>sgn</mi> <mo>{</mo> <munderover> <mo>&Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msub> <mi>ay</mi> <mi>i</mi> </msub> <mi>K</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>&CenterDot;</mo> <mi>x</mi> <mo>)</mo> </mrow> <mo>+</mo> <mi>b</mi> <mo>}</mo> </mrow>Wherein, sgn is sign function, and K is kernel function, and a and b are the parameters obtained by training, xiAnd yiRespectively supporting vector and Its (y that classifiesi=1 or -1), x is sample to be discriminated;Kernel function K is shown below:<mrow> <mi>K</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>)</mo> </mrow> <mo>=</mo> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <mfrac> <mrow> <mo>|</mo> <mo>|</mo> <mi>x</mi> <mo>-</mo> <msub> <mi>x</mi> <mi>i</mi> </msub> <mo>|</mo> <msup> <mo>|</mo> <mn>2</mn> </msup> </mrow> <msup> <mi>&sigma;</mi> <mn>2</mn> </msup> </mfrac> <mo>)</mo> </mrow> </mrow>Wherein, σ is parameter.
- A kind of 8. Data Reduction method based on the online historical data of power system as claimed in claim 7, it is characterised in that The screening sample based on SVM algorithm includes:Training sample by SVM methods, is obtained hyperplane by (4-1);(4-2) inputs training sample as sample to be sorted, obtains classification results;(4-3) deletes the sample of classification error in training sample;(4-4) repeats above step, until the sample of no classification error.
- A kind of 9. Data Reduction method based on the online historical data of power system as claimed in claim 5, it is characterised in that The sample recently is shown to being defined as follows:It is β that the nearest samples of distance sample α, which are calculated, using Euclidean distance formula, if sample β nearest sample is α, sample This α and sample β nearest sample pair each other.
- 10. a kind of Data Reduction method based on the online historical data of power system as claimed in claim 9, its feature exist In the screening based on nearest sample pair includes:(5-1) calculates the nearest sample x ' of each sample x in data set S, labeled as nearest sample;The nearest sample of all samples in (5-2) mark S;(5-3) finds the sample of all samples pair nearest each other in S;(5-4) judges whether each sample centering classification is identical, if identical, deletes one of sample;, will if differing Two samples are all deleted;The S that (5-5) is obtained is data set after compressing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610771830.3A CN107784015B (en) | 2016-08-30 | 2016-08-30 | Data reduction method based on online historical data of power system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610771830.3A CN107784015B (en) | 2016-08-30 | 2016-08-30 | Data reduction method based on online historical data of power system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107784015A true CN107784015A (en) | 2018-03-09 |
CN107784015B CN107784015B (en) | 2022-04-05 |
Family
ID=61450678
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610771830.3A Active CN107784015B (en) | 2016-08-30 | 2016-08-30 | Data reduction method based on online historical data of power system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107784015B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110533265A (en) * | 2019-09-20 | 2019-12-03 | 云南电网有限责任公司电力科学研究院 | A kind of power distribution network lightning stroke method for early warning and power distribution network are struck by lightning prior-warning device |
CN111027612A (en) * | 2019-12-04 | 2020-04-17 | 国网天津市电力公司电力科学研究院 | Energy metering data feature reduction method and device based on weighted entropy FCM |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103744935A (en) * | 2013-12-31 | 2014-04-23 | 华北电力大学(保定) | Rapid mass data cluster processing method for computer |
US20150269233A1 (en) * | 2012-08-06 | 2015-09-24 | National Instruments Corporation | Data Reduction With Specified Constraints |
CN105406476A (en) * | 2015-12-31 | 2016-03-16 | 国家电网公司 | Historical data-based power system stability fast judging method |
CN105512799A (en) * | 2015-11-26 | 2016-04-20 | 中国电力科学研究院 | Mass online historical data-based power system transient stability evaluation method |
CN105897517A (en) * | 2016-06-20 | 2016-08-24 | 广东电网有限责任公司信息中心 | Network traffic abnormality detection method based on SVM (Support Vector Machine) |
-
2016
- 2016-08-30 CN CN201610771830.3A patent/CN107784015B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150269233A1 (en) * | 2012-08-06 | 2015-09-24 | National Instruments Corporation | Data Reduction With Specified Constraints |
CN103744935A (en) * | 2013-12-31 | 2014-04-23 | 华北电力大学(保定) | Rapid mass data cluster processing method for computer |
CN105512799A (en) * | 2015-11-26 | 2016-04-20 | 中国电力科学研究院 | Mass online historical data-based power system transient stability evaluation method |
CN105406476A (en) * | 2015-12-31 | 2016-03-16 | 国家电网公司 | Historical data-based power system stability fast judging method |
CN105897517A (en) * | 2016-06-20 | 2016-08-24 | 广东电网有限责任公司信息中心 | Network traffic abnormality detection method based on SVM (Support Vector Machine) |
Non-Patent Citations (1)
Title |
---|
王快妮 等: "基于支持向量机的异常值问题的研究现状", 《计算机工程与设计》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110533265A (en) * | 2019-09-20 | 2019-12-03 | 云南电网有限责任公司电力科学研究院 | A kind of power distribution network lightning stroke method for early warning and power distribution network are struck by lightning prior-warning device |
CN111027612A (en) * | 2019-12-04 | 2020-04-17 | 国网天津市电力公司电力科学研究院 | Energy metering data feature reduction method and device based on weighted entropy FCM |
CN111027612B (en) * | 2019-12-04 | 2024-01-30 | 国网天津市电力公司电力科学研究院 | Energy metering data feature reduction method and device based on weighted entropy FCM |
Also Published As
Publication number | Publication date |
---|---|
CN107784015B (en) | 2022-04-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111523785A (en) | Power system dynamic security assessment method based on generation countermeasure network | |
CN108008244A (en) | A kind of small current grounding fault progressive classifying identification method at many levels | |
CN107478963A (en) | Single-phase ground fault line selecting method of small-electric current grounding system based on power network big data | |
CN105354210A (en) | Mobile game payment account behavior data processing method and apparatus | |
CN102832617B (en) | Large power grid transient state stabilization analyzing method based on precision pattern discrimination | |
CN107122327A (en) | The method and training system of a kind of utilization training data training pattern | |
CN107679994A (en) | Core protects creation method, device, computer equipment and the storage medium of decision tree | |
CN104881741B (en) | Electric power system transient stability determination methods based on SVMs | |
CN105893388A (en) | Text feature extracting method based on inter-class distinctness and intra-class high representation degree | |
CN104281779A (en) | Abnormal data judging and processing method and device | |
CN105005878B (en) | A kind of comprehensive estimation method of strong intelligent grid | |
CN110532596A (en) | A kind of method and system generating grid simulation sample based on LSTM algorithm | |
CN107784015A (en) | A kind of Data Reduction method based on the online historical data of power system | |
CN105868900A (en) | Enterprise pre-cooperative partner classification method based on SVM (support vector machine) | |
CN105306438A (en) | Network security situation assessment method based on fuzzy rough set | |
Ravi et al. | Hybrid classification and regression models via particle swarm optimization auto associative neural network based nonlinear PCA | |
CN110348540A (en) | Electrical power system transient angle stability Contingency screening method and device based on cluster | |
KR101666740B1 (en) | Method for generating assocication rules for data mining based on semantic analysis in big data environment | |
CN106485036A (en) | Based on the method graded to asset securitization Assets Pool by Survival Models | |
CN116401890A (en) | Method, device, equipment and storage medium for determining composition information of load model | |
CN116400168A (en) | Power grid fault diagnosis method and system based on depth feature clustering | |
CN115564324A (en) | Electric power fingerprint identification method, device and equipment based on event detection | |
CN107491417A (en) | A kind of document structure tree method under topic model based on particular division | |
CN112884363A (en) | Nuclear power project economic evaluation probability risk analysis method | |
JP6818658B2 (en) | Power system monitoring system, power system monitoring method, and program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |