CN106897945A - The clustering method and equipment of wind power generating set - Google Patents
The clustering method and equipment of wind power generating set Download PDFInfo
- Publication number
- CN106897945A CN106897945A CN201710121624.2A CN201710121624A CN106897945A CN 106897945 A CN106897945 A CN 106897945A CN 201710121624 A CN201710121624 A CN 201710121624A CN 106897945 A CN106897945 A CN 106897945A
- Authority
- CN
- China
- Prior art keywords
- cluster
- data
- wind power
- power generating
- acquisition 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.)
- Withdrawn
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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
Abstract
The present invention provides the clustering method and equipment of a kind of wind power generating set.The clustering method comprises the following steps:Tentation data in the data of each wind power generating set in wind power plant is pre-processed to obtain the characteristic of each wind power generating set;By the characteristic of each wind power generating set, and data not being pretreated in the data of each wind power generating set are combined into the data acquisition system of each wind power generating set;Clustered by the data acquisition system of all wind power generating sets in wind power plant, to be clustered to wind power generating set.The clustering method and equipment of wind power generating set according to an embodiment of the invention, data acquisition system according to each wind power generating set is clustered to wind power generating set, is capable of achieving rational classification.
Description
Technical field
The present invention relates to wind power generation field.More particularly, it is related to a kind of clustering method of wind power generating set and sets
It is standby.
Background technology
With the fast development of wind-powered electricity generation industry, the construction of wind power plant expands to complicated landform from level terrain, tradition
Be a cluster using the running status of each unit in unified standard or index evaluation wind power plant with whole wind power plant or go out
Power situation has not met actual conditions.Because under MODEL OVER COMPLEX TOPOGRAPHY, under identical crew qiting parameter, due to difference
Running environment between wind power generating set is each with running status variant, if now speculating its of whole wind field with same standard
His wind-power electricity generation group there is also same running status or incipient fault is then unreasonable.
In the clustering method of existing wind power generating set, typically from unit Equivalent Model is built, in the hope of with more
Few model characterizes some characteristics of whole wind power plant, and mode input is single, it is impossible to adapt to the wind-driven generator in complex environment
Group cluster is analyzed.Therefore, the clustering method of existing wind power generating set is not reasonable, and cluster result is not accurate enough.
The content of the invention
It is an object of the invention to provide the clustering method and equipment of a kind of more efficient and accurate wind power generating set,
To solve the problems, such as there is reason in existing clustering method.
According to an aspect of the present invention, there is provided clustering method described in a kind of clustering method of wind power generating set includes as follows
Step:Obtain the data of every wind power generating set in wind power plant;In extracting the data of every wind power generating set
Tentation data, to form every data acquisition system of wind power generating set, wherein, the tentation data includes:Every wind-power electricity generation
At least two data in the control parameter data of unit, geographic position data, environmental data and service data;Calculated by clustering
Method to wind power plant in the data acquisition system of all wind power generating sets cluster, to realize the cluster to wind power generating set.
Alternatively, the control parameter data include at least one parameter in following parameter:Controller key parameter, machine
Group filter parameter and crew qiting parameter.
Alternatively, the control parameter data are by obtaining the version of the unit initialization files of each wind power generating set
Number obtain, wherein, the version number of the unit initialization files includes the character for indicating the control parameter data.
Alternatively, when the tentation data includes the control parameter data, the tentation data is pre-processed,
Specifically include:The version number is encoded to obtain the data as data acquisition system.
Alternatively, the environmental data includes wind-resources data, ambient humidity and/or the environment temperature of predetermined amount of time.
Alternatively, the wind-resources data of the predetermined amount of time include the wind speed of predetermined amount of time, wherein, when described predetermined
When data include the wind-resources data of the predetermined amount of time, the wind-resources data are pre-processed, specifically included:According to
The wind speed of predetermined amount of time is calculated the occurrence number of each wind speed predetermined amount of time Nei and the turbulence intensity of each wind speed,
As the data of data acquisition system.
Alternatively, the service data includes the pitch angular data of predetermined amount of time.
Alternatively, when the tentation data includes the service data, service data is pre-processed, specific bag
Include:Pitch angular data according to predetermined amount of time is calculated the occurrence number of each propeller pitch angle in predetermined amount of time, as
The data of data acquisition system.
Alternatively, before the step of being clustered the data acquisition system of all wind power generating sets, also include:To each
The data acquisition system of wind power generating set carries out dimension and about subtracts, wherein, by the way that dimension is about subtracted after all wind power generating sets
Data acquisition system is clustered to cluster wind power generating set.
Alternatively, the clustering method also includes:Whether there is between the data acquisition system for verifying all wind power generating sets
Intrinsic cluster, wherein, when there is intrinsic cluster, the data acquisition system of all wind power generating sets is clustered.
Alternatively, all wind power generating sets are verified using Thelma Hopkins statistical method in the step of verifying intrinsic cluster
Whether there is intrinsic cluster between data acquisition system, wherein, when Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.
Alternatively, by clustering algorithm to wind power plant in all wind power generating sets the step that is clustered of data acquisition system
Suddenly include:It is determined that the optimum cluster number for clustering and the Optimal cluster centers for clustering;According to determination for clustering
Optimum cluster number and for cluster Optimal cluster centers, the data acquisition system of all wind power generating sets is clustered.
Optionally it is determined that including for the optimum cluster number for clustering and the step of Optimal cluster centers:Set maximum poly-
Class number and min cluster number;The data acquisition system of all wind power generating sets is carried out the cluster circulation of pre-determined number, is obtained
Cluster is circulated every time optimum cluster number and Optimal cluster centers, wherein, in cluster circulation every time, respectively according to minimum
Cluster number to each the cluster number between maximum cluster number, the data acquisition system to all wind power generating sets gathers
Class, obtains each corresponding cluster result of cluster number and cluster centre, evaluates each corresponding cluster result of cluster number,
Using the best cluster number of cluster result as the optimum cluster number that cluster is circulated every time;Determine the cluster circulation of pre-determined number
The mode of middle optimum cluster number, and the mode that will be determined is used as the optimum cluster number for clustering, and the crowd that will be determined
The best Optimal cluster centers of cluster result are used as the Optimal cluster centers for clustering in the corresponding multiple cluster circulations of number.
Alternatively, the data acquisition system of all wind power generating sets is clustered using improved K mean algorithms, wherein,
The improved K mean algorithms are using apart from farthest sample as initial cluster center.
A kind of cluster equipment of wind power generating set is provided according to another aspect of the present invention, and the cluster equipment includes:
Acquiring unit, obtains the data of every wind power generating set in wind power plant;Extraction unit, extracts described per typhoon power generator
Tentation data in the data of group, to form every data acquisition system of wind power generating set, wherein, the tentation data includes:
At least two numbers in every the control parameter data of wind power generating set, geographic position data, environmental data and service data
According to;Cluster cell, by clustering algorithm to wind power plant in the data acquisition system of all wind power generating sets cluster, to realize
To the cluster of wind power generating set.
Alternatively, the control parameter data include at least one parameter in following parameter:Controller key parameter, machine
Group filter parameter and crew qiting parameter.
Alternatively, the control parameter data are by obtaining the version of the unit initialization files of each wind power generating set
Number obtain, wherein, the version number of the unit initialization files includes the character for indicating the control parameter data.
Alternatively, when the tentation data includes the control parameter data, extraction unit enters to the tentation data
Row pretreatment, specifically includes:Extraction unit is encoded to the version number to obtain the data as data acquisition system.
Alternatively, the environmental data includes wind-resources data, ambient humidity and/or the environment temperature of predetermined amount of time.
Alternatively, the wind-resources data of the predetermined amount of time include the wind speed of predetermined amount of time, wherein, when described predetermined
When data include the wind-resources data of the predetermined amount of time, extraction unit is pre-processed to the wind-resources data, specifically
Including:Extraction unit is calculated in predetermined amount of time the occurrence number of each wind speed and each according to the wind speed of predetermined amount of time
The turbulence intensity of individual wind speed, as the data of data acquisition system.
Alternatively, the service data includes the pitch angular data of predetermined amount of time.
Alternatively, when the tentation data includes the service data, extraction unit is pre-processed to service data,
Specifically include:Extraction unit is calculated going out for each propeller pitch angle in predetermined amount of time according to the pitch angular data of predetermined amount of time
Occurrence number, as the data of data acquisition system.
Alternatively, the cluster equipment also includes:Dimension about subtracts unit, in cluster cell by all wind power generating sets
Before data acquisition system is clustered, dimension is carried out to every data acquisition system of wind power generating set and is about subtracted, wherein, cluster cell leads to
The data acquisition system of all wind power generating sets crossed after dimension is about subtracted is clustered to cluster wind power generating set.
Alternatively, the cluster equipment also includes:Intrinsic cluster authentication unit, verifies the data set of all wind power generating sets
Whether there is intrinsic cluster between conjunction, wherein, when there is intrinsic cluster, cluster cell is by the data acquisition system of all wind power generating sets
Clustered.
Alternatively, intrinsic cluster authentication unit verifies the data set of all wind power generating sets using Thelma Hopkins statistical method
Whether there is intrinsic cluster between conjunction, wherein, when Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.
Alternatively, cluster cell includes:Determination subelement, it is determined that for cluster optimum cluster number and for cluster
Optimal cluster centers;Cluster subelement, the Optimal cluster centers according to the optimum cluster number for determining to be used for and for clustering will
The data acquisition system of all wind power generating sets is clustered.
Optionally it is determined that subelement includes:Setup module, sets maximum cluster number and min cluster number;Cluster is followed
Ring module, the data acquisition system of all wind power generating sets is carried out the cluster circulation of pre-determined number, obtains cluster circulation every time
Optimum cluster number and Optimal cluster centers, wherein, in cluster circulation every time, respectively according to min cluster number to maximum
Each cluster number between cluster number, the data acquisition system to all wind power generating sets is clustered, and obtains each cluster
The corresponding cluster result of number and cluster centre, evaluate each corresponding cluster result of cluster number, and cluster result is best
Cluster number as every time cluster circulation optimum cluster number;Mode determining module, determines the cluster circulation of pre-determined number
The mode of middle optimum cluster number, and the mode that will be determined is used as the optimum cluster number for clustering, and the crowd that will be determined
The best Optimal cluster centers of cluster result are used as the Optimal cluster centers for clustering in the corresponding multiple cluster circulations of number.
Alternatively, cluster cell is gathered using improved K mean algorithms to the data acquisition system of all wind power generating sets
Class, wherein, the improved K mean algorithms are using apart from farthest sample as initial cluster center.
The clustering method and equipment of wind power generating set according to an embodiment of the invention, according to every wind power generating set
Data acquisition system wind power generating set clustered, be capable of achieving rational cluster.
Additionally, the clustering method and equipment of wind power generating set according to an embodiment of the invention, with reference to control parameter
At least two data in data, geographic position data, environmental data and service data, the aggregation of data according to various dimensions considers
Similar situation between wind power generating set, more rationally, cluster result is more accurate for cluster, is the follow-up result based on cluster
Unit allocation is carried out there is provided accurate data basis.
Brief description of the drawings
By the detailed description for carrying out below in conjunction with the accompanying drawings, above and other objects of the present invention, feature and advantage will
Become more fully apparent, wherein:
Fig. 1 shows the flow chart of the clustering method of wind power generating set according to an embodiment of the invention;
Fig. 2 shows to determine according to an embodiment of the invention the optimum cluster number and Optimal cluster centers for clustering
The flow chart of step;
Fig. 3 shows that the frequency for clustering each optimum cluster number in circulation of pre-determined number according to an embodiment of the invention is straight
Square illustrated example;
Fig. 4 shows that dimension according to an embodiment of the invention about subtracts the corresponding relation illustrated example of middle dimension and variance yields;
Fig. 5 shows that dimension about subtracts middle dimension according to an embodiment of the invention and is illustrated with the corresponding relation of accumulation contribution rate
Example;
Fig. 6 shows the Wind turbines group result example in wind power plant according to an embodiment of the invention;
Fig. 7 shows the block diagram of the cluster equipment of wind power generating set according to an embodiment of the invention;
Fig. 8 shows the block diagram of determination subelement according to an embodiment of the invention.
Specific embodiment
Now, different example embodiments are more fully described with reference to the accompanying drawings.
Fig. 1 shows the flow chart of the clustering method of wind power generating set according to an embodiment of the invention.According to the present invention
The clustering method of wind power generating set of embodiment can be used to be grouped the wind power generating set in wind power plant.One wind
Electric field may include multiple wind power generating sets.
In step S101, the data of every wind power generating set in wind power plant are obtained.The data are to wind-driven generator
There are the various data of influence in the cluster of group.In a preferred embodiment of the present invention, the data of every wind power generating set can
Including at least two data in control parameter data, geographic position data, environmental data and service data.Above-mentioned various data
Can be obtained from the database of the data of storage wind power generating set.The data used in above preferred embodiment of the invention
The cluster of more accurately wind power generating set can be realized.
Control parameter data are the various parameters data related to the control of wind power generating set.It is excellent at one of the invention
Selecting embodiment, control parameter data may include at least one of following parameter parameter:Controller key parameter, unit wave filter ginseng
Number and crew qiting parameter.Include for indicating the control parameter data (that is, to control in the version number of unit initialization files
Device key parameter processed, unit filter parameter and crew qiting parameter) character in the case of, can be sent out per typhoon power by obtaining
The version number of the unit initialization files of group of motors obtains the control parameter data.The version number may also include unit initialization
Modification date of file etc., but not limited to this.
Here, the controller key parameter refer to the primary control of wind power generating set parameter in unit allocation
The larger parameter of influence.Unit filter parameter is the parameter of the main wave filter of wind power generating set.Crew qiting parameter is
Refer to the hardware configuration parameter of wind power generating set.
Geographic position data is the data in the geographical position for indicating wind power generating set.For example, geographic position data can be wrapped
Include the data such as the longitude and latitude and height above sea level of unit.
Environmental data is the data with the environmental correclation residing for wind power generating set.For example, environmental data may include it is following
It is at least one:The wind-resources data of predetermined amount of time, ambient humidity and environment temperature etc., but this is not limited only to, for example, described pre-
Section of fixing time can be 1 month, 6 months etc. and can differently as desired be set by user.The wind money of the predetermined amount of time
Source data may include the wind speed of predetermined amount of time, wind direction etc., but not limited to this.
Service data is the data related to the ruuning situation of wind power generating set.For example, the service data may include
The pitch angular data and/or power data of predetermined amount of time, for example, the predetermined amount of time can be 1 month, 6 months etc., and
Can differently as desired be set by user.It should be noted that the predetermined amount of time of service data and above-mentioned wind-resources number
According to predetermined amount of time can be with identical, it is also possible to differ.
In step S102, the tentation data in the data of every wind power generating set is extracted, to form every typhoon power
The data acquisition system of generating set.The tentation data may include every control parameter data, the geographical position of wind power generating set
At least two data in data, environmental data and service data.
Preferably, in order to clean data, data volume is reduced, the partial data in tentation data can be pre-processed, obtained
The data of the principal character of wind power generating set must be embodied, and as the data of data acquisition system.
Different pretreatments can be carried out in step S102 for different tentation datas.
Include controller parameter data in every tentation data of wind power generating set, and by obtaining every wind-power electricity generation
In the case that the version number of the unit initialization files of unit is to obtain the controller parameter data, control parameter data can be entered
Row pretreatment, specifically may include:The version number is encoded to obtain being represented with numerical value as the data of data acquisition system
Version number.Here, sample number can be encoded to using predetermined coding rule for the version number represented with numerical value.
When the tentation data includes the wind-resources data of the predetermined amount of time, and the wind of the predetermined amount of time is provided
Source data includes the wind speed of predetermined amount of time, and the wind-resources data are pre-processed, and specifically includes:According to predetermined amount of time
Wind speed be calculated the occurrence number of each wind speed predetermined amount of time Nei and the turbulence intensity of each wind speed, as data
The data of set.Here, pre- timing can be calculated according to the wind speed of predetermined amount of time using various methods of the prior art
Between in section each wind speed turbulence intensity.
In the case where every tentation data of wind power generating set includes the service data, and the service data bag
When including the propeller pitch angle of predetermined amount of time, service data is pre-processed, specifically included:Propeller pitch angle number according to predetermined amount of time
According to the occurrence number for being calculated each propeller pitch angle in predetermined amount of time, as the data of data acquisition system.
In step S102, the data for the partial data in tentation data pre-process acquisition can be sent out with every typhoon power
The data not being pretreated in the tentation data of group of motors are combined into every data acquisition system of wind power generating set.Do not located in advance
The data of reason may include following at least one:Environment temperature, ambient humidity, the longitude and latitude of unit, the height above sea level of unit.
In step S103, by clustering algorithm to wind field in the data acquisition system of all wind power generating sets cluster,
To realize the cluster to wind power generating set.Here, come right using every data acquisition system of wind power generating set as a sample
The data acquisition system of all wind power generating sets is clustered.
Here, can be clustered come the data acquisition system to all wind power generating sets using various clustering algorithms.For example, can
Clustered using following any algorithm:Improved K mean algorithms, based on hierarchical clustering and has noisy density clustering
Algorithm (DBSCA).The improved K mean algorithms are using apart from farthest sample as initial cluster center.
Preferably, in order to improve the accuracy of cluster result, optimum cluster number for clustering and optimal can first be determined
Cluster centre;Further according to the optimum cluster number and Optimal cluster centers that determine for clustering, by all wind power generating sets
Data acquisition system is clustered.When being clustered, the Optimal cluster centers for clustering will be used for as cluster centre, by all wind-force
The data acquisition system cluster of generating set is the cluster for the optimum cluster number for clustering.
Fig. 2 shows the step of determination of embodiments of the invention is used for the optimum cluster number and the Optimal cluster centers that cluster
Flow chart.
In step S201, maximum cluster number and min cluster number are set.
In step S202, the data acquisition system of all wind power generating sets is carried out the cluster circulation of pre-determined number, obtain every
The optimum cluster number and Optimal cluster centers of secondary cluster circulation.Here, it is poly- according to minimum respectively in cluster circulation every time
Class number to each the cluster number between maximum cluster number, the data acquisition system to all wind power generating sets is clustered,
Each corresponding cluster result of cluster number and cluster centre are obtained, each corresponding cluster result of cluster number is evaluated, will
The best cluster number of cluster result is used as the optimum cluster number that cluster is circulated every time.
Here, various clustering algorithms can be used, according to described each cluster number, to the number of all wind power generating sets
Clustered according to set.For example, can be clustered using following any algorithm:Improved K mean algorithms, based on hierarchical clustering and tool
Noisy density-based algorithms (DBSCA).
Each corresponding cluster result of cluster number can be evaluated using various methods.For example, following any calculation can be used
Method evaluates each cluster corresponding cluster result of number:Bayesian information criterion (BIC), gap statistic algorithm (Gap
) and variance proportion criterion (VRC) etc. statistic.Each corresponding cluster result of cluster number is being evaluated using VRC criterions
When, the bigger corresponding cluster result of cluster number of VRC criterions is better.
In step S203, the mode of optimum cluster number in the cluster circulation of pre-determined number, and the mode that will be determined are determined
As the optimum cluster number for clustering, and in the corresponding multiple cluster circulations of mode that will be determined, cluster result is best
Optimal cluster centers are used as the Optimal cluster centers for clustering.The crowd of optimum cluster number in the cluster circulation of the pre-determined number
Number refers to, in the cluster circulation of pre-determined number, the most optimum cluster number of occurrence number.Fig. 3 shows of the invention
The frequency histogram example of each optimum cluster number in the cluster circulation of the pre-determined number of embodiment.As shown in figure 3, at predetermined time
Number for 100 times cluster circulation in optimum cluster number be 4 situation occur number of times at most, be 31 times, thus mode be 4,
Optimum cluster number for clustering is 4.In a preferred embodiment, in order to reduce data volume, computational efficiency is improved,
Before step S103, dimension can be carried out to every data acquisition system of wind power generating set and about subtracted.Correspondingly, in step s 103,
The data acquisition system of all wind power generating sets after by the way that dimension is about subtracted is clustered to cluster wind power generating set.
In the present embodiment, dimension can be carried out using various methods about to subtract.For example, principal component analytical method can be used
(PCA) about subtract carrying out dimension, in the method, realizes that dimension about subtracts by way of calculating variance contribution ratio, chooses tired
Corresponding dimension about subtracts result as final dimension when product contribution rate is more than predetermined value.Data acquisition system after dimension about subtracts retains
The main information of data acquisition system that dimension about subtracts is not carried out.Fig. 4 shows that dimension about subtracts middle dimension according to an embodiment of the invention
The corresponding relation illustrated example of number and variance yields.As shown in figure 4, abscissa representation dimension about subtract after dimension, the ordinate side of expression
Difference.Fig. 5 shows that dimension according to an embodiment of the invention about subtracts the corresponding relation illustrated example of middle dimension and accumulation contribution rate.Such as
Shown in Fig. 5, abscissa representation dimension about subtract after dimension, ordinate represents accumulation contribution rate, chooses accumulation contribution rate more than pre-
Corresponding dimension 9 about subtracts result as final dimension during definite value (such as 0.95).
In another preferred embodiment, in order to reduce unnecessary computing cost, and classification results more adduction is made
Reason, intrinsic cluster is whether there is between the data acquisition system that before step S103, can verify all wind power generating sets.This be due to
The running environment between wind power generating set in wind power plant is each with operating states of the units variant, if do not existed in data acquisition system
Direct clustering during intrinsic cluster, then the reference significance of cluster result is very limited, only when it is determined that being carried out on the premise of there is intrinsic cluster
Cluster, can just make cluster more reasonable and meaningful.When there is intrinsic cluster, the data set of all wind power generating sets is represented
There is different clusters between conjunction, cluster analysis can be carried out in step S103.During if there is no intrinsic cluster, all wind-force hairs are represented
Do not have different clusters between the data acquisition system of group of motors, without carrying out cluster analysis, all of wind power generating set can be considered
One group, the computing cost that cluster is brought can be removed from.
In the present embodiment, intrinsic cluster checking can be carried out using various verification methods.For example, can be united using Thelma Hopkins
Meter method (Hopkins statistic), the visual evaluation method of cluster trend (Visual Assessment of cluster
) etc. Tendency method carries out intrinsic cluster checking.The situation of intrinsic cluster checking is being carried out using Thelma Hopkins statistical method
Under, if Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.If Thelma Hopkins statistics be more than or
During equal to predetermined threshold, then in the absence of intrinsic cluster, it is not necessary to carry out cluster analysis.
It should be understood that in the present embodiment, in order to reduce data amount of calculation, it may be verified that all wind-force after dimension about subtracts are sent out
Whether there is intrinsic cluster between the data acquisition system of group of motors.
Fig. 6 shows the Wind turbines group result example in wind power plant according to an embodiment of the invention.As shown in fig. 6,
The wind power plant includes 21 wind power generating sets, according to an embodiment of the invention the clustering method of wind power generating set, will
The 21 typhoon power generator group divide into 4 groups.
Fig. 7 shows the block diagram of the cluster equipment of wind power generating set according to an embodiment of the invention.As shown in fig. 7, root
Cluster equipment according to the wind power generating set of embodiments of the invention includes acquiring unit 701, extraction unit 702 and cluster cell
703。
Acquiring unit 701 obtains the data of every wind power generating set in wind power plant.
Extraction unit 702 extracts the tentation data in the data of every wind power generating set, to form every typhoon power
The data acquisition system of generating set.The tentation data may include every control parameter data, the geographical position of wind power generating set
At least two data in data, environmental data and service data.
Preferably, in order to clean data, data volume is reduced, extraction unit 702 can enter to the partial data in tentation data
Row pretreatment, obtains the data of the principal character for embodying wind power generating set, and as the data of data acquisition system.
Extraction unit 702 can carry out different pretreatments for different tentation datas.
Include controller parameter data in every tentation data of wind power generating set, and by obtaining every wind-power electricity generation
In the case that the version number of the unit initialization files of unit is to obtain the controller parameter data, control parameter data can be entered
Row pretreatment, specifically includes:The version number is encoded to obtain as the version represented with numerical value of the data of data acquisition system
This number.Here, sample number can be encoded to using predetermined coding rule for the version number represented with numerical value.
When the tentation data includes the wind-resources data of the predetermined amount of time, and the wind of the predetermined amount of time is provided
Source data includes the wind speed of predetermined amount of time, and the wind-resources data are pre-processed, and specifically includes:According to predetermined amount of time
Wind speed be calculated the occurrence number of each wind speed predetermined amount of time Nei and the turbulence intensity of each wind speed, as data
The data of set.Here, pre- timing can be calculated according to the wind speed of predetermined amount of time using various methods of the prior art
Between in section each wind speed turbulence intensity.
In the case where every tentation data of wind power generating set includes the service data, and the service data bag
When including the propeller pitch angle of predetermined amount of time, service data is pre-processed, specifically included:Propeller pitch angle number according to predetermined amount of time
According to the occurrence number for being calculated each propeller pitch angle in predetermined amount of time, as the data of data acquisition system.
Extraction unit 702 can send out the data for the partial data in tentation data pre-process acquisition with every typhoon power
The data not being pretreated in the tentation data of group of motors are combined into every data acquisition system of wind power generating set.Do not located in advance
The data of reason may include following at least one:Environment temperature, ambient humidity, the longitude and latitude of unit, the height above sea level of unit.
Cluster cell 703 by clustering algorithm to wind field in the data acquisition system of all wind power generating sets cluster,
To realize the cluster to wind power generating set.Here, come right using every data acquisition system of wind power generating set as a sample
The data acquisition system of all wind power generating sets is clustered.
Here, can be clustered come the data acquisition system to all wind power generating sets using various clustering algorithms.For example, can
Clustered using following any algorithm:Improved K mean algorithms, based on hierarchical clustering and has noisy density clustering
Algorithm (DBSCA).The improved K mean algorithms are using apart from farthest sample as initial cluster center.
Preferably, in order to improve the accuracy of cluster result, cluster cell 703 may include that determination subelement and cluster are single
Unit.Determination subelement determines the optimum cluster number and Optimal cluster centers for clustering.Cluster subelement is used for according to determination
The optimum cluster number and Optimal cluster centers of cluster, the data acquisition system of all wind power generating sets is clustered.Carry out
During cluster, cluster subelement will be used for the Optimal cluster centers for clustering as cluster centre, by the number of all wind power generating sets
It is the cluster for the optimum cluster number for clustering according to set cluster.
Fig. 8 shows the block diagram of the determination subelement of embodiments of the invention.As shown in figure 8, embodiments of the invention are really
Stator unit includes setup module 801, cluster loop module 802 and mode determining module 803.
Setup module 801 sets maximum cluster number and min cluster number.
The data acquisition system of all wind power generating sets is carried out cluster loop module 802 the cluster circulation of pre-determined number, is obtained
To the optimum cluster number and Optimal cluster centers of cluster circulation every time.Here, in cluster circulation every time, respectively according to most
Small cluster number to each the cluster number between maximum cluster number, the data acquisition system to all wind power generating sets gathers
Class, obtains each corresponding cluster result of cluster number and cluster centre, evaluates each corresponding cluster result of cluster number,
Using the best cluster number of cluster result as the optimum cluster number that cluster is circulated every time.
Here, various clustering algorithms can be used, according to described each cluster number, to the number of all wind power generating sets
Clustered according to set.For example, can be clustered using following any algorithm:Improved K mean algorithms, based on hierarchical clustering and tool
Noisy density-based algorithms (DBSCA).
Each corresponding cluster result of cluster number can be evaluated using various methods.For example, following any calculation can be used
Method evaluates each cluster corresponding cluster result of number:Bayesian information criterion (BIC), gap statistic algorithm (Gap
) and variance proportion criterion (VRC) etc. statistic.Each corresponding cluster result of cluster number is being evaluated using VRC criterions
When, the bigger corresponding cluster result of cluster number of VRC criterions is better.
Mode determining module 803 determine pre-determined number cluster circulation in optimum cluster number mode, and by determine
Cluster result is most as the optimum cluster number for clustering, and in the corresponding multiple cluster circulations of mode that will be determined for mode
Good Optimal cluster centers are used as the Optimal cluster centers for clustering.Optimum cluster number in the cluster circulation of the pre-determined number
Mode refer to, pre-determined number cluster circulation in, the most optimum cluster number of occurrence number.
In a preferred embodiment, in order to reduce data volume, computational efficiency is improved, according to an embodiment of the invention
The cluster equipment of wind power generating set may also include dimension and about subtract unit (not shown).Dimension about subtracts unit in cluster cell 703
Before being clustered, dimension is carried out to every data acquisition system of wind power generating set and is about subtracted.Correspondingly, cluster cell 703 passes through
The data acquisition system of all wind power generating sets after dimension is about subtracted is clustered to cluster wind power generating set.
In the present embodiment, dimension can be carried out using various methods about to subtract.For example, principal component analytical method can be used
(PCA) about subtract carrying out dimension, in the method, realizes that dimension about subtracts by way of calculating variance contribution ratio, chooses tired
Corresponding dimension about subtracts result as final dimension when product contribution rate is more than predetermined value.Data acquisition system after dimension about subtracts retains
The main information of data acquisition system that dimension about subtracts is not carried out.
In another preferred embodiment, in order to reduce unnecessary computing cost, and classification results more adduction is made
Reason, the cluster equipment of wind power generating set may also include intrinsic cluster authentication unit (not shown) according to an embodiment of the invention.
This is because the running environment between the wind power generating set in wind power plant and operating states of the units are each variant, if in data set
Direct clustering during in the absence of intrinsic cluster is closed, then the reference significance of cluster result is very limited, only before it is determined that there is intrinsic cluster
Put and clustered, can just make cluster more reasonable and meaningful.Intrinsic cluster authentication unit is clustered in cluster cell 703
Before, whether there is intrinsic cluster between the data acquisition system for verifying all wind power generating sets.When there is intrinsic cluster, represent all
There is different clusters, cluster cell 703 can carry out cluster analysis between the data acquisition system of wind power generating set.When in the absence of intrinsic
During cluster, represent between the data acquisition system of all wind power generating sets do not have different clusters, it is all of without carrying out cluster analysis
Wind power generating set can be considered one group, can remove the computing cost that cluster is brought from.
In the present embodiment, intrinsic cluster checking can be carried out using various verification methods.For example, can be united using Thelma Hopkins
Visual evaluation method (the VisualAssessment of cluster of meter method (Hopkins statistic) or cluster trend
) etc. Tendency method carries out intrinsic cluster checking.The situation of intrinsic cluster checking is being carried out using Thelma Hopkins statistical method
Under, if Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.If Thelma Hopkins statistics be more than or
During equal to predetermined threshold, then in the absence of intrinsic cluster, it is not necessary to carry out cluster analysis.
It should be understood that in the present embodiment, in order to reduce data amount of calculation, it may be verified that all wind-force after dimension about subtracts are sent out
Whether there is intrinsic cluster between the data acquisition system of group of motors.
The clustering method and equipment of wind power generating set according to an embodiment of the invention, according to every wind power generating set
Data acquisition system wind power generating set clustered, be capable of achieving rational classification.
Additionally, the clustering method and equipment of wind power generating set according to an embodiment of the invention, with reference to control parameter
At least two data in data, geographic position data, environmental data and service data, the aggregation of data according to various dimensions considers
Similar situation between wind power generating set, more rationally, classification results are more accurate for classification, are the follow-up result based on classification
Unit allocation is carried out there is provided accurate data basis.
Additionally, the clustering method and equipment according to the wind power generating set for embodiments of the invention can be wind farm level
Fault diagnosis provides comparative analysis strategy, when the unit in such as same group breaks down, can compare the fortune of other units in group
The risk level of other units in market condition, or evaluation group, rather than the investigation whole wind power plant whole units of blindness, greatly
Reduce manual maintenance cost.Additionally, can there is the wrong report of certain probability and carry out false early warning in the diagnostic model of unit, it is same
The diagnostic result of the unit in cluster can be mutually authenticated, and reduce diagnostic model misinformation probability.
Additionally, the machine group cluster that the clustering method and equipment of wind power generating set are carried out according to an embodiment of the invention point
Group is conducive to follow-up unit performance to compare.For example, carry out with group internal power curve comparing when, similar wind-resources,
If the situation performance of exerting oneself of unit differs under geography information, unit allocation parameter, what such unit had a potential problems can
Energy property is higher.When carrying out with the contrast of the failure form of expression of unit in group, be conducive to combing fault mode and outer strip
The combing of part or machine unit characteristic.When certain the big part of unit goes wrong in group, more targeted inspection is answered during regular inspection
With other units in group, its potential risks can be higher than the unit in other groups, and regular inspection is safeguarded more purposive.Above-mentioned machine
Group cluster packet can also be prepared for customization service, and such as control parameter is customized, power ascension is customized.
Moreover, it should be understood that the clustering method of wind power generating set can be realized to calculate according to an embodiment of the invention
Computer-readable code on machine readable medium recording program performing.Computer readable recording medium storing program for performing is that can store thereafter can be by computer system
The arbitrary data storage device of the data of reading.The example of computer readable recording medium storing program for performing includes:Read-only storage (ROM), with
Machine accesses memory (RAM), CD-ROM, tape, floppy disk, optical data storage devices and carrier wave (such as through wired or wireless transmission
The data transfer that path passes through internet).Computer readable recording medium storing program for performing also can be distributed in the computer system of connection network, from
And computer-readable code is stored and performed in a distributed manner.Additionally, complete function program of the invention, code and code segment can hold
Change places and explained within the scope of the present invention by the ordinary programmers in field related to the present invention.
Additionally, the unit in the cluster equipment of wind power generating set can be completely by hard according to an embodiment of the invention
Part realizes, such as field programmable gate array (FPGA) or application specific integrated circuit (ASIC);Can also be by hardware and software phase
With reference to mode realize;Can also be realized with software mode by computer program completely.
Although the present invention, those skilled in the art are particularly shown and described with reference to its exemplary embodiment
It should be understood that in the case where the spirit and scope of the present invention that claim is limited are not departed from, form can be carried out to it
With the various changes in details.
Claims (28)
1. a kind of clustering method of wind power generating set, it is characterised in that the clustering method comprises the following steps:
Obtain the data of every wind power generating set in wind power plant;
The tentation data in the data of every wind power generating set is extracted, to form every data set of wind power generating set
Close, wherein, the tentation data includes:Every control parameter data, geographic position data, the environmental data of wind power generating set
With at least two data in service data;
By clustering algorithm to wind power plant in the data acquisition system of all wind power generating sets cluster, to realize sending out wind-force
The cluster of group of motors.
2. clustering method according to claim 1, it is characterised in that the control parameter data are included in following parameter
At least one parameter:Controller key parameter, unit filter parameter and crew qiting parameter.
3. clustering method according to claim 2, it is characterised in that the control parameter data are by obtaining each wind-force
The version number of the unit initialization files of generating set obtains,
Wherein, the version number of the unit initialization files includes the character for indicating the control parameter data.
4. clustering method according to claim 3, it is characterised in that when the tentation data includes the control parameter number
According to when, the tentation data is pre-processed, specifically include:The version number is encoded to obtain as data acquisition system
Data.
5. clustering method according to claim 1, it is characterised in that the environmental data includes that the wind of predetermined amount of time is provided
Source data, ambient humidity and/or environment temperature.
6. clustering method according to claim 5, it is characterised in that the wind-resources data of the predetermined amount of time include pre-
The wind speed of section of fixing time,
Wherein, when the tentation data includes the wind-resources data of the predetermined amount of time, the wind-resources data are carried out
Pretreatment, specifically includes:Wind speed according to predetermined amount of time be calculated the occurrence number of each wind speed in predetermined amount of time with
And the turbulence intensity of each wind speed, as the data of data acquisition system.
7. clustering method according to claim 1, it is characterised in that the service data includes the pitch of predetermined amount of time
Angular data.
8. clustering method according to claim 7, it is characterised in that when the tentation data includes the service data
When, service data is pre-processed, specifically include:Pitch angular data according to predetermined amount of time is calculated predetermined amount of time
The occurrence number of interior each propeller pitch angle, as the data of data acquisition system.
9. clustering method according to claim 1, it is characterised in that enter by the data acquisition system of all wind power generating sets
Before the step of row cluster, also include:Dimension is carried out to the data acquisition system of each wind power generating set about to subtract,
Wherein, by the way that dimension is about subtracted after the data acquisition system of all wind power generating sets clustered come to wind power generating set
Clustered.
10. the clustering method according to claim 1 or 9, it is characterised in that the clustering method also includes:Checking is all
Whether there is intrinsic cluster between the data acquisition system of wind power generating set,
Wherein, when there is intrinsic cluster, the data acquisition system of all wind power generating sets is clustered.
11. clustering methods according to claim 10, it is characterised in that using Hope's gold in the step of verifying intrinsic cluster
Whether there is intrinsic cluster between the data acquisition system of all wind power generating sets of this statistical method checking,
Wherein, when Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.
12. clustering methods according to claim 9, it is characterised in that by clustering algorithm to wind power plant in all wind
The step of data acquisition system of power generator group is clustered includes:
It is determined that the optimum cluster number for clustering and the Optimal cluster centers for clustering;
According to determining for the optimum cluster number for clustering and the Optimal cluster centers for cluster, by all wind-driven generators
The data acquisition system of group is clustered.
13. clustering methods according to claim 12, it is characterised in that it is determined that for the optimum cluster number and most for clustering
The step of excellent cluster centre, includes:
Maximum cluster number and min cluster number are set;
The data acquisition system of all wind power generating sets is carried out the cluster circulation of pre-determined number, the optimal of cluster circulation every time is obtained
Cluster number and Optimal cluster centers, wherein, in cluster circulation every time, clustered according to min cluster number to maximum respectively
Each cluster number between number, the data acquisition system to all wind power generating sets is clustered, and obtains each cluster number
Corresponding cluster result and cluster centre, evaluate each cluster corresponding cluster result of number, by best poly- of cluster result
Class number is used as the optimum cluster number that cluster is circulated every time;
Determine the mode of optimum cluster number in the cluster circulation of pre-determined number, and the mode that will be determined is used as clustering most
Excellent cluster number, and using the best Optimal cluster centers of cluster result in the corresponding multiple cluster circulations of the mode for determining as
For the Optimal cluster centers for clustering.
14. clustering methods according to claim 1, it is characterised in that all wind-force are sent out using improved K mean algorithms
The data acquisition system of group of motors is clustered, wherein, the improved K mean algorithms are using apart from farthest sample as initial clustering
Center.
The cluster equipment of 15. a kind of wind power generating sets, it is characterised in that the cluster equipment includes:
Acquiring unit, obtains the data of every wind power generating set in wind power plant;
Extraction unit, extracts the tentation data in the data of every wind power generating set, to form every typhoon power generator
The data acquisition system of group, wherein, the tentation data includes:Every the control parameter data of wind power generating set, geographical position number
According at least two data in, environmental data and service data;
Cluster cell, by clustering algorithm to wind power plant in the data acquisition system of all wind power generating sets cluster, with reality
Now to the cluster of wind power generating set.
16. cluster equipment according to claim 15, it is characterised in that the control parameter data are included in following parameter
At least one parameter:Controller key parameter, unit filter parameter and crew qiting parameter.
17. cluster equipment according to claim 16, it is characterised in that the control parameter data are by obtaining each wind
The version number of the unit initialization files of power generator group obtains,
Wherein, the version number of the unit initialization files includes the character for indicating the control parameter data.
18. cluster equipment according to claim 17, it is characterised in that when the tentation data includes the control parameter
During data, extraction unit is pre-processed to the tentation data, is specifically included:Extraction unit is encoded to the version number
To obtain the data as data acquisition system.
19. cluster equipment according to claim 15, it is characterised in that the environmental data includes the wind of predetermined amount of time
Resource data, ambient humidity and/or environment temperature.
20. cluster equipment according to claim 19, it is characterised in that the wind-resources data of the predetermined amount of time include
The wind speed of predetermined amount of time,
Wherein, when the tentation data includes the wind-resources data of the predetermined amount of time, extraction unit is to the wind-resources
Data are pre-processed, and are specifically included:Extraction unit is calculated in predetermined amount of time each according to the wind speed of predetermined amount of time
The turbulence intensity of the occurrence number of wind speed and each wind speed, as the data of data acquisition system.
21. cluster equipment according to claim 15, it is characterised in that the service data includes the oar of predetermined amount of time
Elongation data.
22. cluster equipment according to claim 7, it is characterised in that when the tentation data includes the service data
When, extraction unit is pre-processed to service data, is specifically included:Pitch angular data meter of the extraction unit according to predetermined amount of time
Calculation obtains the occurrence number of each propeller pitch angle predetermined amount of time Nei, as the data of data acquisition system.
23. cluster equipment according to claim 15, it is characterised in that the cluster equipment also includes:Dimension about subtracts list
Unit, before cluster cell is clustered the data acquisition system of all wind power generating sets, to every number of wind power generating set
Dimension is carried out according to set about to subtract,
Wherein, the data acquisition system of all wind power generating sets after cluster cell is by the way that dimension is about subtracted is clustered come to wind-force
Generating set is clustered.
The 24. cluster equipment according to claim 15 or 23, it is characterised in that the cluster equipment also includes:Intrinsic cluster
Authentication unit, intrinsic cluster is whether there is between the data acquisition system of all wind power generating sets of checking,
Wherein, when there is intrinsic cluster, cluster cell is clustered the data acquisition system of all wind power generating sets.
25. cluster equipment according to claim 24, it is characterised in that intrinsic cluster authentication unit is counted using Thelma Hopkins
Whether there is intrinsic cluster between the data acquisition system of all wind power generating sets of method validation,
Wherein, when Thelma Hopkins statistics is less than predetermined threshold, there is intrinsic cluster.
26. cluster equipment according to claim 23, it is characterised in that cluster cell includes:
Determination subelement, it is determined that the optimum cluster number for clustering and the Optimal cluster centers for clustering;
Cluster subelement, the Optimal cluster centers according to the optimum cluster number for determining to be used for and for clustering, by all wind-force
The data acquisition system of generating set is clustered.
27. cluster equipment according to claim 26, it is characterised in that determination subelement includes:
Setup module, sets maximum cluster number and min cluster number;
Cluster loop module, the data acquisition system of all wind power generating sets is carried out the cluster circulation of pre-determined number, is obtained every time
The optimum cluster number and Optimal cluster centers of circulation are clustered, wherein, in cluster circulation every time, respectively according to min cluster
Number to each the cluster number between maximum cluster number, the data acquisition system to all wind power generating sets is clustered, obtained
To each corresponding cluster result of cluster number and cluster centre, each corresponding cluster result of cluster number is evaluated, will be poly-
The best cluster number of class result is used as the optimum cluster number that cluster is circulated every time;
Mode determining module, determines the mode of optimum cluster number in the cluster circulation of pre-determined number, and the mode that will be determined is made
It is the optimum cluster number for clustering, and cluster result is best most in the corresponding multiple cluster circulations of mode that will be determined
Excellent cluster centre is used as the Optimal cluster centers for clustering.
28. cluster equipment according to claim 15, it is characterised in that cluster cell uses improved K mean algorithms pair
The data acquisition system of all wind power generating sets is clustered, wherein, the improved K mean algorithms will be made apart from farthest sample
It is initial cluster center.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710121624.2A CN106897945A (en) | 2017-03-02 | 2017-03-02 | The clustering method and equipment of wind power generating set |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710121624.2A CN106897945A (en) | 2017-03-02 | 2017-03-02 | The clustering method and equipment of wind power generating set |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106897945A true CN106897945A (en) | 2017-06-27 |
Family
ID=59185410
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710121624.2A Withdrawn CN106897945A (en) | 2017-03-02 | 2017-03-02 | The clustering method and equipment of wind power generating set |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106897945A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107368556A (en) * | 2017-07-04 | 2017-11-21 | 广西电网有限责任公司电力科学研究院 | A kind of transmission line of electricity multi-source geographic information uniformity matching system |
CN107391615A (en) * | 2017-07-04 | 2017-11-24 | 广西电网有限责任公司电力科学研究院 | A kind of transmission line of electricity multi-source geographic information uniformity matching process |
CN108899930A (en) * | 2018-07-09 | 2018-11-27 | 西华大学 | Wind-powered electricity generation station equivalent modeling method based on Principal Component Analysis Method and hierarchical clustering algorithm |
CN109944740A (en) * | 2017-12-21 | 2019-06-28 | 北京金风科创风电设备有限公司 | Wind farm group control method and device |
CN110874663A (en) * | 2018-08-31 | 2020-03-10 | 北京金风科创风电设备有限公司 | Fault optimization method and device for uninterruptible power supply of wind generating set |
CN111222726A (en) * | 2018-11-23 | 2020-06-02 | 北京金风科创风电设备有限公司 | Method and equipment for identifying abnormality of anemometry data |
CN112459971A (en) * | 2020-11-30 | 2021-03-09 | 吉电(滁州)章广风力发电有限公司 | Abnormal vibration detection method for tower of wind generating set |
CN116151788A (en) * | 2022-04-01 | 2023-05-23 | 港珠澳大桥管理局 | Method, device, equipment and storage medium for classifying electromechanical equipment in tunnel |
CN117056758A (en) * | 2023-10-11 | 2023-11-14 | 中国华能集团清洁能源技术研究院有限公司 | Wind turbine running state identification method, device and equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102760191A (en) * | 2012-06-26 | 2012-10-31 | 西安交通大学 | Double-fed machine set wind power station equivalent modeling system and method based on rotation speed grouping |
CN102945509A (en) * | 2012-10-18 | 2013-02-27 | 中能电力科技开发有限公司 | Wind electric field power prediction method |
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 |
CN103955521A (en) * | 2014-05-08 | 2014-07-30 | 华北电力大学 | Cluster classification method for wind power plant |
CN105303319A (en) * | 2015-11-01 | 2016-02-03 | 国网山东省电力公司聊城供电公司 | Wind power plant dynamic clustering equivalence method based on operating data |
-
2017
- 2017-03-02 CN CN201710121624.2A patent/CN106897945A/en not_active Withdrawn
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102760191A (en) * | 2012-06-26 | 2012-10-31 | 西安交通大学 | Double-fed machine set wind power station equivalent modeling system and method based on rotation speed grouping |
CN102945509A (en) * | 2012-10-18 | 2013-02-27 | 中能电力科技开发有限公司 | Wind electric field power prediction method |
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 |
CN103955521A (en) * | 2014-05-08 | 2014-07-30 | 华北电力大学 | Cluster classification method for wind power plant |
CN105303319A (en) * | 2015-11-01 | 2016-02-03 | 国网山东省电力公司聊城供电公司 | Wind power plant dynamic clustering equivalence method based on operating data |
Non-Patent Citations (2)
Title |
---|
徐立亮等: "基于k-means聚类算法的风电场机群划分方法", 《四川电力技术》 * |
高小力: "大型风电场分组建模方法及其在功率预测中的应用", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107368556B (en) * | 2017-07-04 | 2020-10-20 | 广西电网有限责任公司电力科学研究院 | Power transmission line multi-source geographic information consistency matching system |
CN107391615A (en) * | 2017-07-04 | 2017-11-24 | 广西电网有限责任公司电力科学研究院 | A kind of transmission line of electricity multi-source geographic information uniformity matching process |
CN107368556A (en) * | 2017-07-04 | 2017-11-21 | 广西电网有限责任公司电力科学研究院 | A kind of transmission line of electricity multi-source geographic information uniformity matching system |
CN107391615B (en) * | 2017-07-04 | 2020-06-23 | 广西电网有限责任公司电力科学研究院 | Multi-source geographic information consistency matching method for power transmission line |
CN109944740A (en) * | 2017-12-21 | 2019-06-28 | 北京金风科创风电设备有限公司 | Wind farm group control method and device |
CN109944740B (en) * | 2017-12-21 | 2020-10-27 | 北京金风科创风电设备有限公司 | Wind farm group control method and device |
CN108899930A (en) * | 2018-07-09 | 2018-11-27 | 西华大学 | Wind-powered electricity generation station equivalent modeling method based on Principal Component Analysis Method and hierarchical clustering algorithm |
CN110874663A (en) * | 2018-08-31 | 2020-03-10 | 北京金风科创风电设备有限公司 | Fault optimization method and device for uninterruptible power supply of wind generating set |
CN110874663B (en) * | 2018-08-31 | 2022-07-08 | 北京金风科创风电设备有限公司 | Fault optimization method and device for uninterruptible power supply of wind generating set |
CN111222726A (en) * | 2018-11-23 | 2020-06-02 | 北京金风科创风电设备有限公司 | Method and equipment for identifying abnormality of anemometry data |
CN111222726B (en) * | 2018-11-23 | 2022-07-12 | 北京金风科创风电设备有限公司 | Method and equipment for identifying abnormality of anemometry data |
CN112459971A (en) * | 2020-11-30 | 2021-03-09 | 吉电(滁州)章广风力发电有限公司 | Abnormal vibration detection method for tower of wind generating set |
CN116151788A (en) * | 2022-04-01 | 2023-05-23 | 港珠澳大桥管理局 | Method, device, equipment and storage medium for classifying electromechanical equipment in tunnel |
CN116151788B (en) * | 2022-04-01 | 2023-12-22 | 港珠澳大桥管理局 | Method, device, equipment and storage medium for classifying electromechanical equipment in tunnel |
CN117056758A (en) * | 2023-10-11 | 2023-11-14 | 中国华能集团清洁能源技术研究院有限公司 | Wind turbine running state identification method, device and equipment |
CN117056758B (en) * | 2023-10-11 | 2024-01-30 | 中国华能集团清洁能源技术研究院有限公司 | Wind turbine running state identification method, device and equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106897945A (en) | The clustering method and equipment of wind power generating set | |
CN105182450B (en) | A kind of strong convective weather nowcasting warning system | |
Cui et al. | Wind power ramp event forecasting using a stochastic scenario generation method | |
US20160223600A1 (en) | Power generation performance evaluation method and apparatus for power generator set | |
US8868985B2 (en) | Supervised fault learning using rule-generated samples for machine condition monitoring | |
CN110929918B (en) | 10kV feeder fault prediction method based on CNN and LightGBM | |
CN107909118A (en) | A kind of power distribution network operating mode recording sorting technique based on deep neural network | |
CN111680875B (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
CN108345670B (en) | Service hotspot discovery method for 95598 power work order | |
CN102830626A (en) | Multilevel power system fault diagnosis system based on fault tree | |
CN107392304A (en) | A kind of Wind turbines disorder data recognition method and device | |
CN104182803B (en) | Wind-powered electricity generation data preprocessing method and wind power forecasting method and system | |
CN109325607A (en) | A kind of short-term wind power forecast method and system | |
CN112990330B (en) | User energy abnormal data detection method and device | |
CN111861023A (en) | Statistical-based hybrid wind power prediction method and device | |
CN114723285A (en) | Power grid equipment safety evaluation prediction method | |
CN115526258A (en) | Power system transient stability evaluation method based on Spearman correlation coefficient feature extraction | |
CN115221776A (en) | Training method and device for intelligent sensing diagnosis model of novel power distribution network | |
CN110705831A (en) | Power angle instability mode pre-judgment model construction method after power system fault and application thereof | |
Qu et al. | Research on short‐term output power forecast model of wind farm based on neural network combination algorithm | |
CN107527093B (en) | Wind turbine generator running state diagnosis method and device | |
CN117150359A (en) | Small sample fault diagnosis method, system, device and medium based on model independent element learning | |
CN116502073A (en) | High-reliability intelligent fault diagnosis and health management method for wind generating set | |
Liu et al. | Dynamic security assessment of western Danish power system based on ensemble decision trees | |
CN115758083A (en) | Motor bearing fault diagnosis method based on time domain and time-frequency domain fusion |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20170627 |