CN105956318B - Based on the wind power plant group of planes division methods for improving division H-K clustering method - Google Patents
Based on the wind power plant group of planes division methods for improving division H-K clustering method Download PDFInfo
- Publication number
- CN105956318B CN105956318B CN201610334655.1A CN201610334655A CN105956318B CN 105956318 B CN105956318 B CN 105956318B CN 201610334655 A CN201610334655 A CN 201610334655A CN 105956318 B CN105956318 B CN 105956318B
- Authority
- CN
- China
- Prior art keywords
- cluster
- wind power
- division
- power plant
- point
- 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.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
Abstract
A kind of wind power plant group of planes division methods based on improvement division H-K clustering method, comprising: use split-merge algorithm algorithm, the parameter of blower is divided into k cluster, wherein k is that wind power plant divides number；The cluster C with maximum gauge is found in all clusters；It finds out in cluster C with the maximum point p of other average dissimilarities, and puts it into a new cluster；Target point is found in cluster C, the distance of point of the target point into new cluster is not more than the point nearest into cluster C, and target point is put into new cluster；Silhouette coefficient when different value of K is calculated, optimum clustering number is determined according to silhouette coefficient；In the state of optimum clustering number, cluster centre is calculated to replace the stochastic clustering center in k means clustering algorithm to be clustered；Calculate cluster C in left point to each cluster centre distance and by each left point distribute enter nearest classification in, new cluster centre is calculated after being assigned；Then cluster result of being subject to establishes new farm model and carries out analogue simulation.
Description
Technical field
The present invention relates to wind generator system fields, it is more particularly related to a kind of based on improvement division H-K
The wind power plant group of planes division methods of (Hierarchical K-means) clustering method, can be applied to establish Large Scale Wind Farm Integration etc.
Be worth model, with silhouette coefficient and be subject to suitable criterion solves the problems, such as a group of planes division number, utilize cluster proposed by the present invention
Method completes wind power plant cluster, finally establishes Equivalent Model.
Background technique
With the continuous development of wind power technology and the continuous reduction of cost of electricity-generating, the degree of concern that wind generating technology is subject to
It is higher and higher, but for traditional energy, wind-powered electricity generation can be described as " power supply inferior " because wind-powered electricity generation be with it is intermittent and with
The particular power source of machine fluctuation, therefore its output power also has very strong uncertainty.When large-scale wind farm grid-connected fortune
After row tremendous influence can be brought to the stability and power quality of power grid.It is generated after analysis large-scale wind power field is grid-connected at present
Influence, it is necessary first to solve the problems, such as be large-scale wind power field modeling.Early stage researcher is to every typhoon in wind power plant
Machine is modeled in detail, but this method gradually starts to become more and more difficult with the continuous expansion of wind power plant scale, because
A model that is relatively easy and can portraying wind power plant is established in this research becomes critically important, that is, establishes wind power plant to the outer of power grid
Characteristic model.Mainly using simplified Equivalent Model for wind power plant dynamic model at present, there are mainly two types of mode i.e. single machine equivalences
Model and multimachine Equivalent Model.Comparatively the accuracy of multimachine Equivalent Model is higher.
But with regard at present for, foundations of Equivalent Model all there is a problem of it is corresponding, for a group of planes divide number not clearly
Standard, use is all that empirical method is in the majority.Accordingly, it is desirable to provide a kind of division mode and simplicity for meeting actual demand
Classification Index, it is ensured that establish the accuracy of model.
Summary of the invention
It is a kind of based on cluster the technical problem to be solved by the present invention is to provide for drawbacks described above exists in the prior art
The large-scale wind power field of algorithm models, and optimal clusters number, and detailed analysis fan operation mechanism are determined by silhouette coefficient
Suitable clustering target is selected, multimachine model complexity is reduced while guaranteeing that model is accurate.About wheel at present
Wide coefficient is not described later in detail in terms of wind power plant equivalent modeling, is not probed into deeply in related fields.
In order to achieve the above technical purposes, according to the present invention, it provides a kind of based on the wind for improving division H-K clustering method
Electric field group of planes division methods, comprising:
First step: using split-merge algorithm algorithm, the parameter of blower be divided into k cluster, and wherein k is that wind power plant is drawn
Score mesh；
Second step: the division that there is the cluster C of maximum gauge to be used for data is found in all clusters；
Third step: find out in cluster C with the maximum point p of other average dissimilarities, and put it into one it is new
In cluster；
Four steps: finding target point in cluster C, and the distance of point of the target point into new cluster is not more than into cluster C
Nearest point, and target point is put into new cluster；
5th step: silhouette coefficient when different value of K is calculated, optimum clustering number is determined according to silhouette coefficient；
6th step: in the state of optimum clustering number, calculating cluster centre, (k mean value is poly- to replace k-means cluster
Class) stochastic clustering center in algorithm to be to be clustered；
7th step: calculate cluster C in left point to each cluster centre distance and by each left point distribution entrance
In nearest classification, new cluster centre is calculated after being assigned；
8th step: and then be subject to cluster result and establish new farm model and carry out analogue simulation.
Preferably, in the 5th step silhouette coefficient calculation formula are as follows:
Wherein niIt is the element x in clusteriAverage distance between elements other in this cluster, miIt is the element x in clusteriIt arrives
The minimum value of element average distance in other clusters.
Preferably, the codomain of S (i) is between [- 1,1].
Preferably, k is the integer not less than 2.
Preferably, in all clusters, there are maximum two points of Euclidean distance in cluster C.
Preferably, terminate four steps when the target point being not allocated in new cluster in cluster C and execute the 5th step.
It is according to the present invention to be suitble on a large scale based on the wind power plant group of planes division methods for improving division H-K clustering method
Wind power plant Equivalent Model, the wind power plant modeling based on H-K clustering algorithm, is by k-means clustering method and hierarchical clustering side
Advantage and disadvantage of the method in data processing are combined together, and first obtain initial information using hierarchical clustering algorithm, then use k-
Means clustering algorithm further improves cluster process.It is clustered in the present invention by H-K clustering method and using silhouette coefficient optimization
Both numbers combine, and the number of a clearly division group of planes can make the farm model established be more in line with reality in this way
Situation.
The present invention is based on the modelings of the wind power plant of H-K clustering algorithm, give a group of planes and divide one specific standard of establishment, with this
Based on define the number that large-scale wind power field carries out required equivalent type when multimachine equivalent modeling, third selection is closed afterwards
Suitable clustering target carries out wind power plant cluster modeling with H-K clustering algorithm.Ensure to model efficiently and accurately.
Detailed description of the invention
In conjunction with attached drawing, and by reference to following detailed description, it will more easily have more complete understanding to the present invention
And its adjoint advantage and feature is more easily to understand, in which:
Fig. 1 schematically shows according to the preferred embodiment of the invention based on the wind-powered electricity generation for improving division H-K clustering method
The flow chart of field group of planes division methods.
Fig. 2 schematically shows the profile value according to the preferred embodiment of the invention according to Variable cluster.
It should be noted that attached drawing is not intended to limit the present invention for illustrating the present invention.Note that indicating that the attached drawing of structure can
It can be not necessarily drawn to scale.Also, in attached drawing, same or similar element indicates same or similar label.
Specific embodiment
In order to keep the contents of the present invention more clear and understandable, combined with specific embodiments below with attached drawing in of the invention
Appearance is described in detail.
Holistic approach is described in detail, builds simulation model in MATLAB first, establishes the detailed of a wind power plant
Thin model.The analogue simulation of wind field is carried out under different wind velocity conditions, and saves the relevant curve of output of wind power plant, is deeply being divided
Select suitable index as being calculated in detail in primary data input algorithm after analysing the operation mechanism of blower, attached drawing 1 is
The flow chart of algorithm part, algorithm are handled in accordance with the following steps in the present invention:
First step S1: split-merge algorithm algorithm is used, the parameter of blower is divided into k cluster, wherein k is wind power plant
Divide number.Wherein in order to be bonded reality and reduce operand, k value can select that the upper limit is arranged according to wind field scale since 2；
That is, k is the integer not less than 2.
Second step S2: (that is, in all clusters, there are Europe in cluster C by cluster C of the searching with maximum gauge in all clusters
Formula is apart from maximum two points) it is used for the division of data.
Third step S3: find out in cluster C with the maximum point p of other average dissimilarities, and put it into one it is new
Cluster in.
Four steps S4: finding target point in cluster C, and the distance of point of the target point into new cluster is not more than to cluster C
In nearest point, and target point is put into new cluster.Terminate the 4th when the target point being not allocated in new cluster in cluster C
Step S4 simultaneously executes the 5th step S5.
5th step S5: silhouette coefficient when different value of K is calculated, optimum clustering number is determined according to silhouette coefficient, wherein taking turns
The calculation formula of wide coefficient are as follows:
Wherein niIt is the element x in clusteriAverage distance between elements other in this cluster, miIt is the element x in clusteriIt arrives
The minimum value of element average distance in other clusters, and the codomain of S (i) is between [- 1,1]；
Silhouette coefficient illustrates the element in a close independent collective closer to 1；Illustrate member closer to -1
Element is not belonging to the cluster；Show that nature clustering architecture is not present in sample set if silhouette coefficient is 0, but equally distributed structure.
For example, under different value of K state, profile value of the wind turbine parameter relative to this cluster center is calculated, and can be with
A profile value curve graph is established, and opposing curves maximum using the sum of profile value compare the k value of tendency 1 as optimum cluster
Number.
6th step S6: in the state of optimum clustering number, that is, the optimal wind power plant determined divides under number, calculates
Cluster centre replaces the stochastic clustering center in k-means clustering algorithm to be clustered.
7th step S7: calculate cluster C in left point to each cluster centre distance and by each left point distribute into
Enter in nearest classification, new cluster centre is calculated after being assigned.Preferably, if the cluster centre that the 7th step S7 is calculated
It changes, recalculates cluster centre and carry out left point distribution again until convergence.
8th step S8: and then be subject to cluster result and establish new farm model and carry out analogue simulation, and preferably
Ground can also be compared and analyzed with the curve of output of the detailed model saved before.
Below with reference to specific example, the present invention is described further, a wind power plant is constituted with 24 Fans and is divided
Analysis, the setting of blower primary data are shown in Table 1 (data are per unit value), determine wind speed and refer to that in 12m/s be blower output power, two
Kind wind speed mode operation, the first is the superposition of basic wind and progressive wind；Second is fluctuation wind speed with constant wind speed and Gauss
White noise signal synthesis.Selection clustering target is clustered, and the profile value under different cluster numbers is as shown in Figure 2.
1 blower initial parameter of table
Calculated result according to Fig.2, shows, it is optimal to be polymerized to result when 3 class, and it is shared then to set equivalent blower model
For 3 machine Equivalent Models, Equivalent Model parameter is calculated to be carried out according to equivalent mode.
It should be noted that unless stated otherwise or point out, the otherwise term " first " in specification, " second ", "
The descriptions such as three " are used only for distinguishing various components, element, step etc. in specification, without being intended to indicate that various components, member
Logical relation or ordinal relation between element, step etc..
It is understood that although the present invention has been disclosed in the preferred embodiments as above, above-described embodiment not to
Limit the present invention.For any person skilled in the art, without departing from the scope of the technical proposal of the invention,
Many possible changes and modifications all are made to technical solution of the present invention using the technology contents of the disclosure above, or are revised as
With the equivalent embodiment of variation.Therefore, anything that does not depart from the technical scheme of the invention are right according to the technical essence of the invention
Any simple modifications, equivalents, and modifications made for any of the above embodiments still fall within the range of technical solution of the present invention protection
It is interior.
Claims (6)
1. a kind of based on the wind power plant group of planes division methods for improving division H-K clustering method, characterized by comprising:
First step: using split-merge algorithm algorithm, the parameter of every Fans be divided into k cluster, and wherein k is that wind power plant is drawn
Score mesh；
Second step: the division that there is the cluster C of maximum gauge to be used for data is found in all clusters；
Third step: it finds out in cluster C with the maximum point p of other average dissimilarities, and puts it into a new cluster；
Four steps: finding target point in cluster C, and the distance of point of the target point into new cluster is not more than into cluster C recently
Point, and target point is put into new cluster；
5th step: the silhouette coefficient of every Fans when k value is calculated；
First step is repeated several times after adding 1 to k again to four steps, until determining according to the silhouette coefficient of all blowers optimal poly-
Class number；
6th step: in the state of optimum clustering number, it is random poly- in k means clustering algorithm to replace to calculate cluster centre
Class center is to be clustered；
7th step: calculate cluster C in left point to each cluster centre distance and by each left point distribution enter recently
Classification in, new cluster centre is calculated after being assigned；
8th step: and then be subject to cluster result and establish new farm model and carry out analogue simulation.
2. according to claim 1 based on the wind power plant group of planes division methods for improving division H-K clustering method, feature exists
In the calculation formula of silhouette coefficient in the 5th step are as follows:
Wherein niIt is the element x in clusteriAverage distance between elements other in this cluster, miIt is the element x in clusteriTo other clusters
The minimum value of middle element average distance.
3. according to claim 2 based on the wind power plant group of planes division methods for improving division H-K clustering method, feature exists
In the codomain of S (i) is between [- 1,1].
4. it is according to claim 1 or 2 based on the wind power plant group of planes division methods for improving division H-K clustering method, it is special
Sign is that k is the integer not less than 2.
5. it is according to claim 1 or 2 based on the wind power plant group of planes division methods for improving division H-K clustering method, it is special
Sign is, in all clusters, there are maximum two points of Euclidean distance in cluster C.
6. it is according to claim 1 or 2 based on the wind power plant group of planes division methods for improving division H-K clustering method, it is special
Sign is, when target point being not allocated in new cluster in cluster C terminates four steps and executes the 5th step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610334655.1A CN105956318B (en) | 2016-05-19 | 2016-05-19 | Based on the wind power plant group of planes division methods for improving division H-K clustering method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610334655.1A CN105956318B (en) | 2016-05-19 | 2016-05-19 | Based on the wind power plant group of planes division methods for improving division H-K clustering method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105956318A CN105956318A (en) | 2016-09-21 |
| CN105956318B true CN105956318B (en) | 2019-03-26 |
Family
ID=56912938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610334655.1A Expired - Fee Related CN105956318B (en) | 2016-05-19 | 2016-05-19 | Based on the wind power plant group of planes division methods for improving division H-K clustering method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105956318B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685139A (en) * | 2018-12-25 | 2019-04-26 | 刘万里 | Based on the gastroesophageal reflux disease risk factor extracting method precisely clustered and system |
| CN113408579A (en) * | 2021-05-13 | 2021-09-17 | 桂林电子科技大学 | Internal threat early warning method based on user portrait |
| CN117708613A (en) * | 2023-12-25 | 2024-03-15 | 北京中微盛鼎科技有限公司 | Industrial chain collaborative operation-oriented digital resource matching method |
| CN117476165B (en) * | 2023-12-26 | 2024-03-12 | 贵州维康子帆药业股份有限公司 | Intelligent management method and system for Chinese patent medicine medicinal materials |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103401236A (en) * | 2013-07-30 | 2013-11-20 | 华北电力大学 | Wind power farm generator unit grouping method based on flow correlation of wind power farm |
| CN103558768A (en) * | 2013-08-12 | 2014-02-05 | 清华大学 | Equivalent modeling method based on wind speed distribution characteristics in wind power plant |
| CN103955521A (en) * | 2014-05-08 | 2014-07-30 | 华北电力大学 | Cluster classification method for wind power plant |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101519483B1 (en) * | 2013-09-27 | 2015-05-12 | 한국전력공사 | Apparatus for simulating wind power farm |
-
2016
- 2016-05-19 CN CN201610334655.1A patent/CN105956318B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103401236A (en) * | 2013-07-30 | 2013-11-20 | 华北电力大学 | Wind power farm generator unit grouping method based on flow correlation of wind power farm |
| CN103558768A (en) * | 2013-08-12 | 2014-02-05 | 清华大学 | Equivalent modeling method based on wind speed distribution characteristics in wind power plant |
| CN103955521A (en) * | 2014-05-08 | 2014-07-30 | 华北电力大学 | Cluster classification method for wind power plant |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105956318A (en) | 2016-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105956318B (en) | Based on the wind power plant group of planes division methods for improving division H-K clustering method | |
| CN101672637B (en) | Digitizing detection method of complicated curved face | |
| CN106779148B (en) | A kind of method for forecasting wind speed of high speed railway line of multi-model multiple features fusion | |
| CN104317997A (en) | Optimized design method for modelling of end wall of high load fan/compressor | |
| CN109766604B (en) | Blade high-rigidity design method based on random isogeometric analysis | |
| CN104376214A (en) | Fluctuating wind velocity simulation method based on data driving | |
| CN106897826A (en) | A kind of street accidents risks appraisal procedure and system | |
| CN107784380A (en) | The optimization method and optimization system of a kind of inspection shortest path | |
| CN103268572A (en) | A micro-siting method of wind detecting network of ten-million-kilowatt-class large wind power base | |
| CN107300853B (en) | Pulsating wind speed time-course input method for calculating spatial correlation in partition mode | |
| CN103714166B (en) | laser radar power line point cloud data fitting method | |
| CN109992889B (en) | Wind power plant model building method and system and wake value calculating method and system | |
| CN104732545A (en) | Texture image segmentation method combined with sparse neighbor propagation and rapid spectral clustering | |
| CN106503279B (en) | A kind of modeling method for transient stability evaluation in power system | |
| CN109492796A (en) | A kind of Urban Spatial Morphology automatic Mesh Partition Method and system | |
| CN109492748A (en) | A kind of Mid-long term load forecasting method for establishing model of the electric system based on convolutional neural networks | |
| CN106372801A (en) | Wind power field wind region division method and method for selecting wind detection tower position of wind power field wind regions | |
| CN110532666A (en) | Urban micro-climate evolution analysis method based on wind environment variation | |
| CN110263834A (en) | A kind of detection method of new energy power quality exceptional value | |
| CN104463251A (en) | Cancer gene expression profile data identification method based on integration of extreme learning machines | |
| CN108509718B (en) | Far-field wake two-dimensional analytic model based on mass conservation | |
| CN201514207U (en) | Digitized detection system for complex curved surface | |
| CN107391950A (en) | A kind of turbulent flow wall is apart from quick calculation method | |
| CN115392133A (en) | Wind power plant optimal clustering equivalence method adopting Gaussian mixture model | |
| CN107909194A (en) | System level testing designs Multipurpose Optimal Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190326 |