CN110135471A - Wind power output typical scene generation method based on BIRCH algorithm - Google Patents
Wind power output typical scene generation method based on BIRCH algorithm Download PDFInfo
- Publication number
- CN110135471A CN110135471A CN201910337008.XA CN201910337008A CN110135471A CN 110135471 A CN110135471 A CN 110135471A CN 201910337008 A CN201910337008 A CN 201910337008A CN 110135471 A CN110135471 A CN 110135471A
- Authority
- CN
- China
- Prior art keywords
- scene
- wind
- power output
- powered electricity
- electricity generation
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 235000018185 Betula X alpestris Nutrition 0.000 title claims abstract 20
- 235000018212 Betula X uliginosa Nutrition 0.000 title claims abstract 20
- 230000005611 electricity Effects 0.000 claims description 48
- 239000012141 concentrate Substances 0.000 claims description 11
- 238000012360 testing method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/231—Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram
-
- 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)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Wind Motors (AREA)
Abstract
The wind power output typical scene generation method based on BIRCH algorithm that the invention discloses a kind of, comprising steps of obtaining the initial scene collection of wind power output；Reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, typical scene collection is obtained.Present invention application BIRCH algorithm is iterated reduction to initial scene collection, obtains typical scene collection.Method provided by the present invention can quickly and accurately carry out the reduction of wind power output scene, compare other methods, and this method has greater advantage on calculating time and storage size.
Description
Technical field
The present invention relates to field of power systems, more particularly to a kind of wind power output typical field examined based on BIRCH algorithm
Scape generation method.
Background technique
In recent years, Chinese wind-power electricity generation scale rapid growth becomes the third-largest power source of China.Wind-power electricity generation exists
The problem of scene collection scale of sampling is excessive, influences computational efficiency.Therefore, it is significant to carry out scene reduction.
Existing scene reduction method have after to "flop-out" method, fast forword back-and-forth method, synchronous back substitution null method, scene tree structure
Build method etc..But current scene reduction method requires to calculate when carrying out scene reduction all possible to be merged into apart from group
Row compares, and when scene quantity is larger, there are the disadvantages such as computationally intensive, computational efficiency is low.
The computational efficiency for how improving scene reduction is that current wind power output typical scene generation needs that furthers investigate to ask
Topic.
Summary of the invention
The present invention solves the technical problem of BIRCH algorithm is used, a kind of generation wind power output typical scene is provided
Method.
The following technical solution is employed by the present invention:
Reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, typical scene collection is obtained.
Specifically, comprising steps of
Reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, typical scene collection, packet are obtained
It includes:
Wind power output scene refers to the practical wind power curve of output at each moment in one day.Obtaining, wind power output is initial
After scene collection, the present invention clusters these wind power output scenes by BIRCH algorithm.
BIRCH (Balanced Iterative Reducing and Clustering Using Hierarchies) is
The hierarchical clustering algorithm of one increment type only needs single pass wind-powered electricity generation scene collection that can effectively be clustered.BIRCH algorithm uses
Bottom-up strategy, and relocated by iteration and improve the situation as a result, larger suitable for scene amount and classification number.Wherein,
Cluster feature (Cluster Feature, CF) and clustering tree (CF-Tree) are two key concepts of BIRCH algorithm.
Define scene collection comprising N number of d dimension wind-powered electricity generation scene: { zi(i=1,2 ..., N), then the wind-powered electricity generation scene collection
Cluster feature CF is a triple:
CF=(N, LS, SS)
In formula, N represents the scene number of wind-powered electricity generation scene concentration；LS represents the vector sum that wind-powered electricity generation scene concentrates each dimension of sceneSS represents the quadratic sum that wind-powered electricity generation scene concentrates each dimension of sample point
Cluster feature CF can reflect the essential information of wind-powered electricity generation scene.Wherein, LS can reflect that scene concentrates all kinds of scenes
Cluster centre:
z0=LS/N
In formula, z0For the center of wind-powered electricity generation scene cluster, it can be used for calculating the distance between each scene cluster；SS can reflect scene
Concentrate the average distance of all kinds of scenes:
CF meets linear relationship, it may be assumed that
CFa+CFb=(Na+Nb,LSa+LSb,SSa+SSb)
The property illustrates that the CF triple numerical value of each father node in CF-Tree is equal to all sons pointed by the father node
The sum of triple of node, so as to be used to improve the update efficiency of CF-Tree.
CF-Tree is the balanced tree for reflecting wind-powered electricity generation scene clustering situation.The form of tree is reflected by three parameters: non-leaf segment
Point branch parameter B, leaf node branch parameter L and wind-powered electricity generation scene cluster maximum radius threshold value T.Wherein, B is root node and nonleaf node
Number maximum value；L is the number maximum value of leaf node and wind-powered electricity generation scene cluster, and each leaf node may include multiple scene clusters；T is
The maximum sample radius of wind-powered electricity generation scene cluster, it is ensured that the compactness of wind-powered electricity generation scene cluster.
The method of the present invention application BIRCH algorithm carries out scene reduction and obtains typical wind-powered electricity generation scene collection, it is possible to reduce when calculating
Between and storage size.
Technical solution provided by the invention the utility model has the advantages that
As the popularity rate of the renewable energy such as wind-powered electricity generation is higher and higher, the uncertainty of electric system is also increasing.?
In previous Power System Planning and operation, the method for scenario analysis is generallyd use to solve the problems, such as this, but due to calculation amount
Greatly, validity is restricted.Therefore, the present invention provides a kind of carries out wind-powered electricity generation typical case based on the method for BIRCH algorithm
Scene generates, i.e., is iterated reduction to initial scene collection using BIRCH algorithm, obtains typical scene collection.It is provided by the present invention
Method can quickly and accurately carry out the reduction of wind power output scene, compare other methods, this method is calculating the time and depositing
There is greater advantage in storage scale.
Detailed description of the invention
Fig. 1 is the wind power output typical scene generation method flow chart based on BIRCH algorithm；
Fig. 2 is the CF-Tree schematic diagram of BIRCH algorithm.
Specific embodiment
Purpose, technical solution and technical effect for a better understanding of the present invention, below in conjunction with attached drawing to the present invention
Carry out further explaining illustration.
A kind of wind power output typical scene generation method based on BIRCH algorithm, it is characterized in that: comprising steps of
Obtain the initial scene collection of wind power output；
Reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, typical scene collection is obtained.
The application BIRCH algorithm is iterated reduction to initial scene collection, split, again reduces, and obtains typical scene collection,
Include:
Wind power output scene refers to the practical wind power curve of output at each moment in one day；Obtaining, wind power output is initial
After scene collection, these wind power output scenes are clustered by BIRCH algorithm；
BIRCH is the hierarchical clustering algorithm of an increment type, only needs single pass wind-powered electricity generation scene collection that can effectively be gathered
Class；BIRCH algorithm uses bottom-up strategy, and is relocated and improved as a result, being suitable for scene amount and classification number by iteration
Larger situation；Wherein, cluster feature；And clustering tree；It is two key concepts of BIRCH algorithm；
Define scene collection comprising N number of d dimension wind-powered electricity generation scene: { zi(i=1,2 ..., N), then the wind-powered electricity generation scene collection
Cluster feature CF is a triple:
CF=(N, LS, SS)
In formula, N represents the scene number of wind-powered electricity generation scene concentration；LS represents the vector sum that wind-powered electricity generation scene concentrates each dimension of sceneSS represents the quadratic sum that wind-powered electricity generation scene concentrates each dimension of sample point
Cluster feature CF can reflect the essential information of wind-powered electricity generation scene；Wherein, LS can reflect that scene concentrates all kinds of scenes
Cluster centre:
z0=LS/N
In formula, z0For the center of wind-powered electricity generation scene cluster, it can be used for calculating the distance between each scene cluster；SS can reflect scene
Concentrate the average distance of all kinds of scenes:
CF meets linear relationship, it may be assumed that
CFa+CFb=(Na+Nb,LSa+LSb,SSa+SSb)
The property illustrates that the CF triple numerical value of each father node in CF-Tree is equal to all sons pointed by the father node
The sum of triple of node, so as to be used to improve the update efficiency of CF-Tree；
CF-Tree is the balanced tree for reflecting wind-powered electricity generation scene clustering situation；The form of tree is reflected by three parameters: non-leaf segment
Point branch parameter B, leaf node branch parameter L and wind-powered electricity generation scene cluster maximum radius threshold value T；Wherein, B is root node and nonleaf node
Number maximum value；L is the number maximum value of leaf node and wind-powered electricity generation scene cluster, and each leaf node may include multiple scene clusters；T is
The maximum sample radius of wind-powered electricity generation scene cluster, it is ensured that the compactness of wind-powered electricity generation scene cluster.
For a further understanding of the present invention, below with the typical wind-powered electricity generation data instance in Chinese somewhere, to explain the present invention
Practical application.
After the initial scene collection for obtaining this area, 24 moment are divided into 4 subintervals, each subinterval packet first
Containing 6 moment, M quantile is carved with when each.In this way, M will be only generated in each subinterval6A initial scene, first to every
M in a subinterval6A initial scene is reduced, this 4 subintervals are then reconnected, to form the typical case at 24 moment
Scene collection.When scene number reaches given threshold value after final reduction, classical wind power output scene collection can be obtained.In order to compare
The superiority and inferiority of the method for the present invention and other methods, table 1 are listed respectively when scene is reduced to 1000,500,300 and 100 respectively,
Based on BIRCH algorithm, k-means cluster and Agglomerative hierarchical clustering (aggregative hierarchy clustering,
AHC scene) reduces speed.Test environment is Windows 7, MATLAB2014a with Python 3.7；Hardware is Core
i5 7400@3.00GHz,RAM 8GB.Quantile M takes 6.The method based on AHP always can not be can as can be seen from Table 1
It provides in the time (< 1h) of receiving as a result, this is because during AHP, needs to generate 66×66Double type apart from square
Battle array, occupies 6 altogether6×66× 8=1.74 × 1010B ≈ 16.2GB has been more than the free memory for testing computer.Based on BIRCH and k-
The method of means can be provided in finite time as a result, on the one hand they only generate several 66× 1 vector, it is another
Their time complexity of aspect is all O (n) rank.Method ratio as can also be seen from Table 1 based on BIRCH algorithm is based on k-
Means algorithm more faster, this is because BIRCH algorithm only needs a scan data set that can set up typical scene
Clustering tree, and k-means algorithm needs iterate.
Table 1 is compared with the speed of other methods
Therefore, the mentioned method of the present invention and k-means method are substantially better than the side AHP on the memory space that scene is reduced
Method.Compared with k-means method, the mentioned method of the present invention in calculating speed advantageously.
Claims (6)
1. a kind of wind power output typical scene generation method based on BIRCH algorithm, it is characterized in that: comprising steps of
Obtain the initial scene collection of wind power output；
Reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, typical scene collection is obtained.
2. the wind power output typical scene generation method according to claim 1 based on BIRCH algorithm, it is characterized in that: institute
It states and reduction is iterated to initial scene collection using BIRCH algorithm, split, is again reduced, obtain typical scene collection, comprising:
Wind power output scene refers to the practical wind power curve of output at each moment in one day；Obtaining the initial scene of wind power output
After collection, these wind power output scenes are clustered by BIRCH algorithm；
BIRCH is the hierarchical clustering algorithm of an increment type, only needs single pass wind-powered electricity generation scene collection that can effectively be clustered；
BIRCH algorithm uses bottom-up strategy, and is relocated and improved as a result, larger suitable for scene amount and classification number by iteration
The case where；Wherein, cluster feature；And clustering tree；It is two key concepts of BIRCH algorithm；
Define scene collection comprising N number of d dimension wind-powered electricity generation scene: { zi(i=1,2 ..., N), then the cluster of the wind-powered electricity generation scene collection
Feature CF is a triple:
CF=(N, LS, SS)
In formula, N represents the scene number of wind-powered electricity generation scene concentration；LS represents the vector sum that wind-powered electricity generation scene concentrates each dimension of scene
SS represents the quadratic sum that wind-powered electricity generation scene concentrates each dimension of sample pointCluster feature CF can reflect the basic of wind-powered electricity generation scene
Information；Wherein, LS can reflect that scene concentrates the cluster centre of all kinds of scenes:
z0=LS/N
In formula, z0For the center of wind-powered electricity generation scene cluster, it can be used for calculating the distance between each scene cluster；SS can reflect that scene is concentrated
The average distance of all kinds of scenes:
CF meets linear relationship, it may be assumed that
CFa+CFb=(Na+Nb,LSa+LSb,SSa+SSb)
The property illustrates that the CF triple numerical value of each father node in CF-Tree is equal to all child nodes pointed by the father node
The sum of triple, so as to be used to improve the update efficiency of CF-Tree；
CF-Tree is the balanced tree for reflecting wind-powered electricity generation scene clustering situation；The form of tree is reflected by three parameters: nonleaf node point
Branch parameter B, leaf node branch parameter L and wind-powered electricity generation scene cluster maximum radius threshold value T；Wherein, B is of root node and nonleaf node
Number maximum value；L is the number maximum value of leaf node and wind-powered electricity generation scene cluster, and each leaf node may include multiple scene clusters；T is wind-powered electricity generation
The maximum sample radius of scene cluster, it is ensured that the compactness of wind-powered electricity generation scene cluster.
3. the wind power output typical scene generation method according to claim 1 based on BIRCH algorithm, it is characterized in that: CF-
The CF triple numerical value of each father node is equal to the sum of the triple of all child nodes pointed by the father node in Tree, thus
The update efficiency of CF-Tree can be used to improve.
4. the wind power output typical scene generation method according to claim 1 based on BIRCH algorithm, it is characterized in that: wind
Electricity power output scene refers to the practical wind power curve of output at each moment in one day；Obtain the initial scene collection of wind power output it
Afterwards, these wind power output scenes are clustered by BIRCH algorithm.
5. the wind power output typical scene generation method according to claim 1 based on BIRCH algorithm, it is characterized in that:
BIRCH is the hierarchical clustering algorithm of an increment type, only needs single pass wind-powered electricity generation scene collection that can effectively be clustered；BIRCH
Algorithm uses bottom-up strategy, and is relocated and improved as a result, being suitable for scene amount and the biggish feelings of classification number by iteration
Condition；Wherein, cluster feature；And clustering tree；It is two key concepts of BIRCH algorithm.
6. the wind power output typical scene generation method according to claim 1 based on BIRCH algorithm, it is characterized in that: CF-
Tree is the balanced tree for reflecting wind-powered electricity generation scene clustering situation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910337008.XA CN110135471A (en) | 2019-04-25 | 2019-04-25 | Wind power output typical scene generation method based on BIRCH algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910337008.XA CN110135471A (en) | 2019-04-25 | 2019-04-25 | Wind power output typical scene generation method based on BIRCH algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110135471A true CN110135471A (en) | 2019-08-16 |
Family
ID=67570969
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910337008.XA Withdrawn CN110135471A (en) | 2019-04-25 | 2019-04-25 | Wind power output typical scene generation method based on BIRCH algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110135471A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110929399A (en) * | 2019-11-21 | 2020-03-27 | 国网江苏省电力有限公司南通供电分公司 | Wind power output typical scene generation method based on BIRCH clustering and Wasserstein distance |
-
2019
- 2019-04-25 CN CN201910337008.XA patent/CN110135471A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110929399A (en) * | 2019-11-21 | 2020-03-27 | 国网江苏省电力有限公司南通供电分公司 | Wind power output typical scene generation method based on BIRCH clustering and Wasserstein distance |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110569316A (en) | low-voltage distribution area user topology identification method based on t-SNE dimension reduction technology and BIRCH clustering | |
| WO2019056753A1 (en) | Dynamic equivalent modeling method for distributed photovoltaic power station cluster | |
| CN109524993A (en) | The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling | |
| CN107679687A (en) | A kind of photovoltaic output modeling method and Generation System Reliability appraisal procedure | |
| CN109086527A (en) | A kind of practical equivalent modeling method based on running of wind generating set state | |
| Liu et al. | Clustering methods of wind turbines and its application in short-term wind power forecasts | |
| CN110929399A (en) | Wind power output typical scene generation method based on BIRCH clustering and Wasserstein distance | |
| CN108549962A (en) | Wind power forecasting method based on history fragment sequence search and sequential rarefaction | |
| CN109978042A (en) | A kind of adaptive quick K-means clustering method of fusion feature study | |
| CN109409575A (en) | Wind power plant group of planes division methods based on Gap Statistic | |
| CN110909786A (en) | New user load identification method based on characteristic index and decision tree model | |
| CN110135471A (en) | Wind power output typical scene generation method based on BIRCH algorithm | |
| CN107276093B (en) | The Probabilistic Load calculation method cut down based on scene | |
| CN106355511A (en) | Active power distribution network reconstruction method taking new energy and electric vehicle access into consideration | |
| CN114021818A (en) | Wind power multistep prediction method considering space-time distribution characteristics | |
| Zhang et al. | Joint forecasting of regional wind and solar power based on attention neural network | |
| CN117332963A (en) | Dynamic optimization scheduling method and system for virtual power plant with collaborative source network and load storage | |
| CN112085065A (en) | Low-voltage user-to-home phase identification method based on voltage and active power reading | |
| CN116662844A (en) | Method and device for generating new energy output power typical scene | |
| CN109638892A (en) | A kind of photovoltaic plant equivalent modeling method based on improvement fuzzy clustering algorithm | |
| CN104200106A (en) | Longitudinal time axis clustering method in generalized load modeling on basis of seasonality | |
| Wu et al. | t-SNE based on Sobol sequence initialized exchange market algorithm | |
| Peng et al. | Stochastic scenario generation for wind power and photovoltaic system based on CGAN | |
| Dong et al. | Nontechnical loss detection of electricity based on neural architecture search in distribution power networks | |
| Liu et al. | Topology Identification of Low-voltage Transformer Area Based on Improved Particle Swarm Algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190816 |
|
| WW01 | Invention patent application withdrawn after publication |