CN107248031A - A kind of fast power user classification method for load curve peak-valley difference - Google Patents
A kind of fast power user classification method for load curve peak-valley difference Download PDFInfo
- Publication number
- CN107248031A CN107248031A CN201710392694.1A CN201710392694A CN107248031A CN 107248031 A CN107248031 A CN 107248031A CN 201710392694 A CN201710392694 A CN 201710392694A CN 107248031 A CN107248031 A CN 107248031A
- Authority
- CN
- China
- Prior art keywords
- mrow
- power consumer
- msub
- power
- peak
- 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
Classifications
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention provides a kind of fast power user classification method for load curve peak-valley difference.This method comprises the following steps:1) the daily load data of power consumer are obtained;2) operation is normalized in data;3) corresponding peak interval of time is chosen;4) according to correlation coefficient function, iterate, finally give the similar classification of power customers result of peak interval of time curve.The present invention can cluster to power consumer according to the daily load curve data of user, facilitate following needs company specific aim to generate strategy to reduce peak-valley difference.
Description
Technical field
The invention belongs to power system Demand-side technical field of information management, and in particular to one kind is directed to load curve peak valley
The fast power user classification method of difference.
Background technology
With social development and the raising of living standards of the people, resident's is also becomed more diverse with electrical characteristics, and power consumption is held
It is continuous to rise, the pressure of imbalance between power supply and demand has been aggravated, has caused power network peak-valley difference persistently to increase.And low-carbon is being advocated energetically
Under form that is economic and carrying out clean energy resource, a large amount of new energy access power networks also increase the peak regulation pressure of power network, make power network
Peaking problem becomes increasingly conspicuous.Meanwhile, with the construction of intelligent grid, the demand response mechanism of user side has turned into the stabilization of power grids and transported
The essential valuable source of row.It is then desired to study the demand response behavior of power consumer, it is made full use of to dispatch potentiality,
Peak load shifting, alleviates peak load regulation network pressure.However, different load types has different part throttle characteristics, used in research electric power
Before the demand response behavior at family, power consumer is classified and just seems necessary.
There is certain document report for the method that electric load curve is clustered in the prior art.But, for institute in document
Some load curves all time points are calculated, and clustering method is complicated, the problem of existing computationally intensive.
The content of the invention
It is an object of the invention to provide a kind of fast power user classification method for load curve peak-valley difference.
For achieving the above object, present invention employs following technical scheme:
1) power consumer daily load curve to be sorted is obtained;
2) daily load curve of each power consumer of acquisition is normalized, when then according to the peak valley of selection
Section, obtains the respective load data of each power consumer, goes to step 3);
3) Euclidean distance of still non-classified power consumer load between any two is calculated, is determined as according to Euclidean distance poly-
The power consumer at class center;
4) load and the electricity as cluster centre of remaining power consumer are calculated respectively in still non-classified power consumer
The coefficient correlation of the load of power user, according to coefficient correlation by the similar power consumer of the daily load curve in the peak interval of time
It is divided into a class with the power consumer as cluster centre;
5) by repeat step 3) and 4) be iterated, until reaching stopping criterion for iteration.
The load data includes the electric power on each time point corresponding to the peak interval of time chosen on daily load curve
Normalization result.
The normalized is carried out according to below equation:
In formula:TS represents daily load curve time point set, PitRepresent i-th of power consumer to be sorted t-th of time
Electric power on point, xitRepresent PitData after normalization.
The selection quantity of the peak interval of time studies actual peak valley when formulating peak-valley difference regulating strategy according to Utilities Electric Co.
Period situation is determined.
The step 3) specifically include following steps:
3.1) Euclidean distance of load between power consumer is calculated, n × n dimensions matrix D is is formed:
In formula,dijRepresent i-th power consumer load and j-th of power consumer load it
Between Euclidean distance, i=1 ..., n, j=1 ..., n, T represents the time point set of selected peak interval of time, xikAnd xjkPoint
Biao Shi not the electric power normalization data of i-th and j-th of power consumer on k-th of time point;
3.2) matrix D is is summed by row, obtains each between power consumer and remaining power consumer in n power consumer
Load Euclidean distance and, the power consumer chosen in matrix D is corresponding to distance and minimum row is cluster centre.
The step 4) in, coefficient correlation is calculated according to below equation:
In formula,Represent i-th of power consumer electric power normalization data x in the peak interval of time of selectionikAverage,
I is not the power consumer as cluster centre,The power consumer of cluster centre is denoted as to use in the peak interval of time of selection
Electrical power normalization data xmkAverage.
The similar criterion of the daily load curve is coefficient correlation >=0.8.
The end condition of the iteration is:Still non-classified power consumer number≤2 or remaining described power consumer
Load and the load of the power consumer as cluster centre coefficient correlation it is equal<0.8.
The step 2) in, peak interval of time is chosen by Utilities Electric Co. according to regional total load curve, power consumer to be sorted
It is user in this area.
Beneficial effects of the present invention are embodied in:
Euclidean distance and coefficient correlation are applied in the classification of power customers for peak-valley difference by the present invention, by choosing phase
Corresponding peak interval of time carries out the calculating of Euclidean distance and coefficient correlation, rather than all the period of time calculating, it is quick, accurate to reach
The purpose of classification, and clustering method is simply efficient, is that the demand response behavior of follow-up study power consumer lays the foundation, after convenience
Continuous Utilities Electric Co. pointedly formulates peak-valley difference regulating strategy.
Further, the present invention can be by choosing the lower limit of coefficient correlation, the similarity degree and classification number of controlling curve
Purpose quantity, to reach purpose that classification becomes more meticulous.
Brief description of the drawings
Fig. 1 is the fast power user classification method flow chart for load curve peak-valley difference.
Fig. 2 is the regional total load curve referred in emulating.
Fig. 3 a, Fig. 3 b, Fig. 3 c are the cluster result of different classes of user.
Embodiment
The present invention is elaborated with reference to the accompanying drawings and examples.Described is explanation of the invention, rather than limit
It is fixed.
Fast power user classification method of the present invention for load curve peak-valley difference, relate to power consumer
Daily load data (peak interval of time), can be made and other electric power in the set by being chosen from power consumer set N to be sorted
The minimum power consumer of user's correspondence period daily load data Euclidean distance sum is as cluster centre, then by calculating respectively
The coefficient correlation of other power consumers period daily load data corresponding with the power consumer as cluster centre, selects daily load
Curve has the power consumer of similar features and is classified as a class, iterates, until completing the cluster of power consumer in set N.
It is main to consider following two aspect when carrying out similar load curve cluster according to correlation:
Firstth, by clustering the curve for needing to obtain that there is very strong similitude in peak interval of time, similitude and coefficient correlation
The relation of numerical value is referring to table 1;
Secondth, avoid because coefficient correlation lower limit obtains too big, the situation for causing Classification of Load Curves number excessive, most
The criterion selected eventually is correlation coefficient r >=0.8.
The similarity criteria of table 1.
The present invention's comprises the following steps that, referring to Fig. 1:
Step 1: obtaining s (usual s>2) the daily load data of power consumer to be sorted and the volume of corresponding power consumer
Number, daily load data include the power at corresponding time point.Conventional power consumer daily load data are derived from correspondence power consumer day
The electric power of 24 or 96 time points (interval 1 hour is spaced 15 minutes) on load curve.
Step 2: power consumer daily load data are normalized into operation, calculation formula is as follows:
In formula:A represents power consumer day electric power summation, and TS represents daily load curve time point set, PitRepresent i-th
Electric power of the individual power consumer on t-th of time point, xitRepresent PitData after normalization.Not only disappeared by normalization
Except dimension, and eliminate influence of the varying number level to cluster.
Step 3: due to when Utilities Electric Co. can select the one or more peak valleys of regional total load curve according to actual conditions
Duan Jinhang is studied, so needing the corresponding peak interval of time in selected daily load curve, peak interval of time is in the literature without strict
Definition, when being all that Utilities Electric Co. formulates electricity price, defined according to local area situation, peak interval of time includes a peak load
Period and a paddy lotus period.For example, one section of (time point t in the regional total load curve of selection in maximum peak load1~t2Between) with
And one section of (time point t in maximum valley lotus3~t4Between) it is used as a peak interval of time.After selected peak interval of time, according to time sequencing
Retain the data at each time point in selected peak interval of time, the data at remaining time point on daily load curve are removed, by reservation
Data are reintegrated according to time sequencing, form the power consumer load number that n power consumers of not yet classifying each occupy a line
According to matrix, often capable data element number is depending on time point sum in selected peak interval of time, and n initial value is s.
Step 4: calculating the Euclidean distance of power consumer load data, matrix is obtained:
In formula,dijRepresent i-th power consumer load and j-th of power consumer load it
Between Euclidean distance, i=1 ..., n, j=1 ..., n, T represents the time point set of peak interval of time.
Step 5: matrix D is is summed by row, the minimum row of selection Euclidean distance sum (due to the limitation of computational accuracy,
Equal and when being minimum row when there are two Euclidean distance sums, any selection wherein a line continues following step), will
Power consumer corresponding to the row is designated as m.The load and use of power consumer corresponding to remaining n-1 row in difference calculating matrix Dis
The coefficient correlation of family m loads:
In formula:Represent the average of i-th of power consumer electric power normalization data in selected peak interval of time, i
≠ m,Represent the average of power consumer m electric power normalization datas in selected peak interval of time;
The load data for choosing correlation coefficient r >=0.8 is expert at, by the corresponding power consumer of these rows and power consumer m
A class is classified as, and records the corresponding initial line number m of these power consumers1、m2..., the result of record is represented with vector V:
V=[m1,m2,…]
Step 6: when vectorial V is empty set, then jumping to step 8, otherwise continuing executing with next step.
Step 7, as n≤2, then jumps to step 9.Otherwise, vector V is removed from power consumer load data matrix
The data that element is expert at, return to step four.
Step 8: giving separate marking by user corresponding to remaining n-1 row (has not been able to cluster, it is understood that there may be Acquisition Error
The problems such as), continue executing with next step.
Step 9: output program result (cluster result and the user for having not been able to cluster).
Simulation example
With the daily load data instance of 800, somewhere user, regional total load curve is shown in Fig. 2, and the wherein paddy lotus period is
When 11~13, when the peak load period is 21~24, the time points included altogether are 7.The user's classification finally given has 10 classes.
It is can be seen that from Fig. 3 a~c by classification, there is daily load curve the user of similar peak interval of time to have obtained effective classification.
In summary, the present invention uses Euclidean distance, and cluster centre point is looked for from the overall situation, then by coefficient correlation, will be negative
The similar user of lotus curve peak interval of time, is classified as a class, makes cluster result more reasonable.The present invention can be negative according to the day of user
Lotus data (peak interval of time), realize quick user's classification, are that further research user power utilization behavior and demand response behavior are established
Fixed basis, facilitates following needs company specific aim to generate strategy to reduce peak-valley difference.
Claims (9)
1. a kind of fast power user classification method for load curve peak-valley difference, it is characterised in that:Comprise the following steps:
1) power consumer daily load curve to be sorted is obtained;
2) daily load curve of each power consumer of acquisition is normalized, then according to the peak interval of time of selection,
Obtain the respective load data of each power consumer;
3) Euclidean distance of still non-classified power consumer load between any two is calculated, is determined as according to Euclidean distance in cluster
The power consumer of the heart;
4) coefficient correlation of the load and the load of the power consumer as cluster centre of remaining power consumer is calculated respectively, according to
Coefficient correlation divides the similar power consumer of the daily load curve in the peak interval of time with the power consumer as cluster centre
For a class;
5) by repeat step 3) and 4) be iterated, until reaching stopping criterion for iteration.
2. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The load data includes the normalization result of the electric power on daily load curve correspondence time point.
3. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The normalized is carried out according to below equation:
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>a</mi>
</mfrac>
<mo>&CenterDot;</mo>
<msub>
<mi>P</mi>
<mrow>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>,</mo>
<mi>a</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>t</mi>
<mo>&Element;</mo>
<mi>T</mi>
<mi>S</mi>
</mrow>
</munder>
<msub>
<mi>P</mi>
<mrow>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
</mrow>
In formula:TS represents daily load curve time point set, PitRepresent i-th of power consumer to be sorted on t-th of time point
Electric power, xitRepresent PitData after normalization.
4. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The selection quantity of the peak interval of time studies actual peak interval of time feelings when formulating peak-valley difference regulating strategy according to Utilities Electric Co.
Condition is determined.
5. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The step 3) specifically include following steps:
3.1) Euclidean distance of load between power consumer is calculated, n × n dimensions matrix D is is formed:
In formula,dijRepresent between the load of i-th power consumer and the load of j-th of power consumer
Euclidean distance, i=1 ..., n, j=1 ..., n, T represent the time point set of selected peak interval of time, xikAnd xjkDifference table
Show i-th and electric power normalization data of j-th of power consumer on k-th of time point;
3.2) matrix D is is summed by row, obtains each negative between power consumer and remaining power consumer in n power consumer
The Euclidean distance of lotus and, the power consumer chosen in matrix D is corresponding to distance and minimum row is cluster centre.
6. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The step 4) in, coefficient correlation is calculated according to below equation:
<mrow>
<mi>r</mi>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>&Element;</mo>
<mi>T</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>x</mi>
<mi>m</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mrow>
<mo>(</mo>
<msqrt>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>&Element;</mo>
<mi>T</mi>
</mrow>
</munder>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
</mrow>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<msqrt>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>&Element;</mo>
<mi>T</mi>
</mrow>
</munder>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>m</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>x</mi>
<mi>m</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
</mrow>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>)</mo>
</mrow>
</mrow>
1
In formula,Represent i-th of power consumer electric power normalization data x in the peak interval of time of selectionikAverage, i is not
For as the power consumer of cluster centre,It is denoted as the power consumer of cluster centre and uses electric work in the peak interval of time of selection
Rate normalization data xmkAverage, T represents the time point set of selected peak interval of time.
7. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The similar criterion of the daily load curve is coefficient correlation >=0.8.
8. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The end condition of the iteration is:Still non-classified power consumer number≤2 or the load of remaining power consumer
Coefficient correlation with the load of the power consumer as cluster centre is equal<0.8.
9. a kind of fast power user classification method for load curve peak-valley difference according to claim 1, its feature exists
In:The step 2) in, peak interval of time is chosen according to regional total load curve, and power consumer to be sorted is to be used in this area
Family.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710392694.1A CN107248031B (en) | 2017-05-27 | 2017-05-27 | Rapid power consumer classification method aiming at load curve peak-valley difference |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710392694.1A CN107248031B (en) | 2017-05-27 | 2017-05-27 | Rapid power consumer classification method aiming at load curve peak-valley difference |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107248031A true CN107248031A (en) | 2017-10-13 |
CN107248031B CN107248031B (en) | 2021-07-13 |
Family
ID=60017805
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710392694.1A Active CN107248031B (en) | 2017-05-27 | 2017-05-27 | Rapid power consumer classification method aiming at load curve peak-valley difference |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107248031B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110210755A (en) * | 2019-05-30 | 2019-09-06 | 国网山东省电力公司泰安供电公司 | A kind of user demand responding ability appraisal procedure based on K_means clustering algorithm |
CN111738340A (en) * | 2020-06-24 | 2020-10-02 | 西安交通大学 | Distributed K-means power user classification method, storage medium and classification equipment |
CN111754090A (en) * | 2020-06-08 | 2020-10-09 | 深圳供电局有限公司 | Power consumer demand side regulation and control method based on big data |
CN111754091A (en) * | 2020-06-08 | 2020-10-09 | 深圳供电局有限公司 | Power consumer demand side regulation and control system |
CN111967723A (en) * | 2020-07-24 | 2020-11-20 | 南昌大学 | User peak regulation potential analysis method based on data mining |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120252430A1 (en) * | 2009-08-21 | 2012-10-04 | Imes Kevin R | Establishing proximity detection using 802.11 based networks |
CN103914621A (en) * | 2014-04-03 | 2014-07-09 | 国家电网公司 | Method for analyzing distribution and transformation daily load typical curves on basis of structure features |
CN104680261A (en) * | 2015-03-16 | 2015-06-03 | 朗新科技股份有限公司 | Power load operation control method based on load curve clustering of major clients |
-
2017
- 2017-05-27 CN CN201710392694.1A patent/CN107248031B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120252430A1 (en) * | 2009-08-21 | 2012-10-04 | Imes Kevin R | Establishing proximity detection using 802.11 based networks |
CN103914621A (en) * | 2014-04-03 | 2014-07-09 | 国家电网公司 | Method for analyzing distribution and transformation daily load typical curves on basis of structure features |
CN104680261A (en) * | 2015-03-16 | 2015-06-03 | 朗新科技股份有限公司 | Power load operation control method based on load curve clustering of major clients |
Non-Patent Citations (1)
Title |
---|
朱晓清: "电力负荷的分类方法及其应用", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110210755A (en) * | 2019-05-30 | 2019-09-06 | 国网山东省电力公司泰安供电公司 | A kind of user demand responding ability appraisal procedure based on K_means clustering algorithm |
CN110210755B (en) * | 2019-05-30 | 2023-04-18 | 国网山东省电力公司泰安供电公司 | User demand response capability assessment method based on K _ means clustering algorithm |
CN111754090A (en) * | 2020-06-08 | 2020-10-09 | 深圳供电局有限公司 | Power consumer demand side regulation and control method based on big data |
CN111754091A (en) * | 2020-06-08 | 2020-10-09 | 深圳供电局有限公司 | Power consumer demand side regulation and control system |
CN111738340A (en) * | 2020-06-24 | 2020-10-02 | 西安交通大学 | Distributed K-means power user classification method, storage medium and classification equipment |
CN111738340B (en) * | 2020-06-24 | 2022-05-20 | 西安交通大学 | Distributed K-means power user classification method, storage medium and classification equipment |
US11768858B2 (en) | 2020-06-24 | 2023-09-26 | Xi'an Jiaotong University | Method of power user classification based on distributed K-means, storage medium and classification device |
CN111967723A (en) * | 2020-07-24 | 2020-11-20 | 南昌大学 | User peak regulation potential analysis method based on data mining |
CN111967723B (en) * | 2020-07-24 | 2022-05-20 | 南昌大学 | User peak regulation potential analysis method based on data mining |
Also Published As
Publication number | Publication date |
---|---|
CN107248031B (en) | 2021-07-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107248031A (en) | A kind of fast power user classification method for load curve peak-valley difference | |
CN103559655B (en) | The Forecasting Methodology of the novel feeder line load of microgrid based on data mining | |
Huang et al. | A clustering based grouping method of nearly zero energy buildings for performance improvements | |
Sohn et al. | Decision tree based on data envelopment analysis for effective technology commercialization | |
Shao et al. | A new electricity price prediction strategy using mutual information-based SVM-RFE classification | |
CN105117810A (en) | Residential electricity consumption mid-term load prediction method under multistep electricity price mechanism | |
CN106446967A (en) | Novel power system load curve clustering method | |
CN112184070B (en) | Multi-objective optimization scheduling method and system for cascade hydropower station with cooperative ecological flow demand | |
CN106408008A (en) | Load curve distance and shape-based load classification method | |
CN102999791A (en) | Power load forecasting method based on customer segmentation in power industry | |
CN107133652A (en) | Electricity customers Valuation Method and system based on K means clustering algorithms | |
Zhang et al. | Analysis of power consumer behavior based on the complementation of K-means and DBSCAN | |
CN109117872A (en) | A kind of user power utilization behavior analysis method based on automatic Optimal Clustering | |
CN104239968A (en) | Short-term load predicting method based on quick fuzzy rough set | |
CN110109971A (en) | A kind of low-voltage platform area user power utilization Load Characteristic Analysis method | |
CN111460001B (en) | Power distribution network theoretical line loss rate evaluation method and system | |
CN108428055A (en) | A kind of load characteristics clustering method considering load vertical characteristics | |
Campbell et al. | Methodologies for load forecasting | |
CN106777005A (en) | User power utilization behavior analysis method based on big data technological improvement clustering algorithm | |
CN107748944A (en) | Decontrol user side demand response method under environment in a kind of sale of electricity side | |
CN104376402B (en) | A kind of load classification and integrated approach based on frequency-domain index | |
Wang et al. | Big data analytics for price forecasting in smart grids | |
CN105653670A (en) | Intelligent electricity consumption data mining method based on manifold learning clustering algorithm | |
CN104820942A (en) | Electricity market trade evaluation criterion measuring and calculating method based on hierarchical clustering | |
CN112465403A (en) | Regional power demand side response potential evaluation method |
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 |