CN116378897A - Wind farm yaw angle control method and device - Google Patents
Wind farm yaw angle control method and device Download PDFInfo
- Publication number
- CN116378897A CN116378897A CN202310492612.6A CN202310492612A CN116378897A CN 116378897 A CN116378897 A CN 116378897A CN 202310492612 A CN202310492612 A CN 202310492612A CN 116378897 A CN116378897 A CN 116378897A
- Authority
- CN
- China
- Prior art keywords
- wind
- wind turbine
- sets
- generating set
- group
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000010248 power generation Methods 0.000 claims abstract description 84
- 238000011156 evaluation Methods 0.000 claims abstract description 7
- 239000002245 particle Substances 0.000 claims abstract description 7
- 238000012549 training Methods 0.000 claims description 44
- 238000003062 neural network model Methods 0.000 claims description 33
- 238000010845 search algorithm Methods 0.000 claims description 15
- 238000012423 maintenance Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000010408 sweeping Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 235000006629 Prosopis spicigera Nutrition 0.000 description 1
- 240000000037 Prosopis spicigera Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D7/00—Controlling wind motors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
- F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
- F05B2270/00—Control
- F05B2270/70—Type of control algorithm
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
- F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
- F05B2270/00—Control
- F05B2270/70—Type of control algorithm
- F05B2270/709—Type of control algorithm with neural networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/72—Wind turbines with rotation axis in wind direction
Abstract
The application provides a wind farm yaw angle control method and device, wherein the method comprises the following steps: dividing the wind power generator sets in the wind power plant according to wake interference relation between any two wind power generator sets in the wind power plant to obtain a plurality of wind power generator groups; generating multiple groups of random yaw angles corresponding to each wind turbine group aiming at each wind turbine group, and calculating the total power generation and average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles; inputting each group of random yaw angles corresponding to the wind turbine groups and total power generation and average fatigue load corresponding to each group of random yaw angles into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting target yaw angles of each wind turbine group; and adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle. By the mode, the total power generation amount of the wind power plant is improved, and the fatigue load of the wind generating set is reduced.
Description
Technical Field
The application relates to the technical field of wind power generation, in particular to a method and a device for controlling a yaw angle of a wind power plant.
Background
The power generation principle of the wind generating set is that wind power is utilized to drive blades to rotate, and then the rotating speed is increased through a speed increaser so as to promote a generator to generate power.
In a wind farm, in order to improve land utilization, as many wind power generation units as possible may be installed on a limited land, thereby obtaining as much power generation as possible. In order to increase the power generation amount of each wind turbine, the blades of each wind turbine in the wind farm are usually aligned with the incoming wind direction (the wind turbine is in a windward state) so that the yaw angle of each wind turbine is 0. For a single wind generating set, the yaw angle is adjusted to be 0, so that the generating capacity of the wind generating set is the highest. However, for a wind farm, the wind farm includes a plurality of wind power generating sets, and wake effects may exist between the wind power generating sets, so that wake interference occurs.
The wake effect refers to the formation of a wake zone of decreasing wind speed downstream of a wind turbine generator system while harvesting wind energy from the wind. If the downstream wind generating set is positioned in the wake zone, the input wind speed of the downstream wind generating set is lower than that of the upstream fan. The wake effect causes uneven wind speed distribution in the wind power plant, and the generated energy of each wind generating set in the wind power plant is easily influenced and extra fatigue load is caused.
Disclosure of Invention
In view of the above, an object of the present application is to provide a method and an apparatus for controlling yaw angle of a wind farm, so as to avoid wake interference between wind turbine generator sets, reduce wake influence, improve total power generation amount generated by the wind farm, and reduce fatigue load of the wind turbine generator sets in the wind farm.
In a first aspect, embodiments of the present application provide a method for controlling a yaw angle of a wind farm, the method including:
dividing wind power generator groups in a wind power plant according to wake interference relation between any two wind power generator groups in the wind power plant to obtain a plurality of wind power generator groups; no wake interference exists between the wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine cluster and at least one wind generating set in the wind turbine cluster;
generating multiple groups of random yaw angles corresponding to each wind turbine group aiming at each wind turbine group so as to calculate the total power generation and average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles;
Inputting each group of random yaw angles corresponding to the wind turbine groups and the total power generated by each group of random yaw angles and the average fatigue load into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting a target yaw angle of each wind turbine group in the wind turbine groups;
and adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle.
With reference to the first aspect, an embodiment of the present application provides a first possible implementation manner of the first aspect, where, according to a wake interference relationship between any two wind turbine groups in a wind farm, wind turbine groups of the wind turbine groups in the wind farm are divided, and before obtaining a plurality of wind turbine groups, the method further includes:
inputting relative position coordinates of two wind power generator sets and incoming wind speed of the wind power plant into a pre-trained classified neural network model aiming at any two wind power generator sets in the wind power plant, and outputting the wake interference relationship between the two wind power generator sets; the wake interference relationship is the presence or absence of wake interference.
With reference to the first possible implementation manner of the first aspect, the present embodiment provides a second possible implementation manner of the first aspect, wherein the classification neural network model is trained by:
acquiring a training sample set comprising a plurality of training samples; each training sample comprises sample relative position coordinates between two sample wind generating sets, sample incoming wind speed and a label for representing wake interference relation between the two sample wind generating sets;
inputting the relative position coordinates of the samples in the training samples and the incoming wind speeds of the samples into an initial classification neural network model for each training sample, and outputting a predicted wake interference relationship between the two sample wind generating sets;
and calculating a cross entropy loss function according to the predicted wake interference relation and the label to obtain a loss value, so as to train the learnable parameters in the initial classified neural network model by using the loss value, and stopping training until training is performed for a preset number of times by using the training sample set, thus obtaining the classified neural network model.
With reference to the first possible implementation manner of the first aspect, the embodiment of the present application provides a third possible implementation manner of the first aspect, wherein, before the inputting, for any two wind turbine groups in the wind farm, the relative position coordinates of the two wind turbine groups and the incoming wind speed of the wind farm into a pre-trained classification neural network model, the method further includes:
Inputting the position coordinates of any two wind generating sets in the wind power plant and the included angle between the incoming wind direction and the designated direction into the following formula to obtain the relative position coordinates of the two wind generating sets:
wherein, (x) ij ,y ij ) Is the relative position coordinate between wind generating set i and wind generating set j, (x) i ,y i ) Is the position coordinate of the wind generating set i, (x) j ,y j ) Is the position coordinate of the wind generating set j, and theta is the included angle between the incoming wind direction and the appointed direction.
With reference to the first aspect, an embodiment of the present application provides a fourth possible implementation manner of the first aspect, wherein the dividing a wind turbine group of wind turbine groups of any two wind turbine groups in a wind farm according to a wake interference relationship between the wind turbine groups in the wind farm to obtain a plurality of wind turbine groups includes:
generating a wind power plant wake interference relation topological graph according to wake interference relations between any two wind power generator sets in the wind power plant;
and inputting an adjacency matrix for representing the wind farm wake interference relation topological graph into a depth-first search algorithm to obtain a plurality of wind turbine clusters.
With reference to the first aspect, the embodiment of the present application provides a fifth possible implementation manner of the first aspect, wherein inputting an adjacency matrix for characterizing the wind farm wake interference relationship topological graph into a depth-first search algorithm, to obtain a plurality of wind turbine groups, includes:
in the depth-first search algorithm, any wind generating set is selected from wind generating sets which are not divided into wind generating sets, the wind generating sets which have direct wake interference and indirect wake interference with the wind generating sets are divided into the same wind generating set, and the selection of any wind generating set from the wind generating sets which are not divided into the wind generating sets and the subsequent steps are continuously executed until all the wind generating sets in the wind power plant are divided into the respective wind generating sets to stop.
With reference to the first aspect, an embodiment of the present application provides a sixth possible implementation manner of the first aspect, wherein, for each wind turbine group, a plurality of groups of random yaw angles corresponding to the wind turbine group are generated, so as to calculate total power generated by the wind turbine group when yaw angles of each wind turbine group in the wind turbine group are each random yaw angle in each group of random yaw angles, where the method includes:
The total power generated by the wind turbine group is calculated by the following formula:
wherein P is n For the power generation of the nth wind generating set, ρ is the air density, C P Is the wind energy utilization coefficient of the wind generating set, S is the wind sweeping area of the impeller of the wind generating set, v is the incoming wind speed, and gamma n Is the random yaw angle of the wind generating set;
wherein P is WG And N is the number of wind generating sets in the wind turbine group, wherein N is the total power of the wind turbine group.
With reference to the sixth possible implementation manner of the first aspect, the embodiment of the present application provides a seventh possible implementation manner of the first aspect, wherein, for each wind turbine group, generating multiple sets of random yaw angles corresponding to the wind turbine group, so as to calculate an average fatigue load of the wind turbine group when yaw angles of each wind turbine group in the wind turbine group are each random yaw angle in each set of random yaw angles, where the method includes:
the average fatigue load of the wind farm is calculated by the following formula:
F n ＝f p +f t
wherein F is n Is the fatigue load of the nth wind generating set, f p Is fatigue load related to the generated power, f t Is fatigue load related to turbulence, P rated Is rated power of the nth wind generating set, T life The method is characterized in that the method is the estimated service life of an nth wind generating set, t is a preset time step, P (tau) is the real-time power of the nth wind generating set, W is a weight coefficient of turbulence fatigue, r is a maintenance compensation coefficient of the nth wind generating set, and I (tau) is the real-time turbulence of the nth wind generating set;
wherein F is WG Is the average fatigue load of the wind turbine group.
With reference to the first aspect, the embodiments of the present application provide an eighth possible implementation manner of the first aspect, wherein the method further includes:
acquiring a first wind farm power generation and a first operation cost of the wind farm in a designated time step after the yaw angles of the wind turbine generator sets in the wind turbine generator group are adjusted by using the target yaw angle; acquiring second wind farm power generation and second operation cost of the wind farm in a designated time step before the yaw angle of each wind turbine generator set in the wind turbine group is adjusted by using the target yaw angle;
when the first wind farm power is smaller than the second wind farm power, and the first operation cost is greater than the second operation cost, continuing to use the target yaw angle;
And when the power generated by the first wind farm is not less than the power generated by the second wind farm, and/or the first operation cost is not greater than the second operation cost, re-executing the step to divide wind power groups of the wind power generator groups in the wind farm according to the wake interference relationship between any two wind power generator groups in the wind farm to obtain a plurality of wind power groups and subsequent steps.
In a second aspect, embodiments of the present application further provide a wind farm yaw control device, the device including:
the division module is used for dividing wind power generation sets in the wind power plant into wind power generation sets according to wake interference relation between any two wind power generation sets in the wind power plant to obtain a plurality of wind power generation sets; no wake interference exists between the wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine cluster and at least one wind generating set in the wind turbine cluster;
the generation module is used for generating a plurality of groups of random yaw angles corresponding to each wind turbine group so as to calculate the total power generation and the average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles;
The first input module is used for inputting each group of random yaw angles corresponding to the wind turbine groups and the total power generation and the average fatigue load corresponding to each group of random yaw angles into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting the target yaw angle of each wind turbine group in the wind turbine groups;
and the adjusting module is used for adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle.
With reference to the second aspect, embodiments of the present application provide a first possible implementation manner of the second aspect, where the method further includes:
the second input module is used for dividing wind power generation sets in a wind power plant into wind power generation sets according to wake flow interference relation between any two wind power generation sets in the wind power plant, inputting relative position coordinates of the two wind power generation sets and incoming flow wind speed of the wind power plant into a pre-trained classification neural network model for any two wind power generation sets in the wind power plant before obtaining a plurality of wind power generation sets, and outputting the wake flow interference relation between the two wind power generation sets; the wake interference relationship is the presence or absence of wake interference.
With reference to the first possible implementation manner of the second aspect, the present application examples provide a second possible implementation manner of the second aspect, where the method further includes:
the first acquisition module is used for acquiring a training sample set containing a plurality of training samples; each training sample comprises sample relative position coordinates between two sample wind generating sets, sample incoming wind speed and a label for representing wake interference relation between the two sample wind generating sets;
the third input module is used for inputting the relative position coordinates of the samples in the training samples and the incoming wind speeds of the samples into an initial classification neural network model for each training sample, and outputting a predicted wake interference relationship between the two sample wind generating sets;
and the training module is used for calculating a cross entropy loss function according to the predicted wake interference relation and the label to obtain a loss value so as to train the learnable parameters in the initial classified neural network model by using the loss value, and stopping training until training is performed for preset times by using the training sample set to obtain the classified neural network model.
With reference to the first possible implementation manner of the second aspect, the embodiment of the present application provides a third possible implementation manner of the second aspect, where the method further includes:
The fourth input module is configured to input, for any two of the wind power generator sets in the wind farm, the relative position coordinates of the two wind power generator sets and the incoming wind speed of the wind power farm into a pre-trained classification neural network model, and input, for any two of the wind power generator sets in the wind farm, the position coordinates of the two wind power generator sets and the included angle between the incoming wind direction and the specified direction into the following formula before outputting the wake interference relationship between the two wind power generator sets, to obtain the relative position coordinates of the two wind power generator sets:
wherein, (x) ij ,y ij ) Is the relative position coordinate between wind generating set i and wind generating set j, (x) i ,y i ) Is the position coordinate of the wind generating set i, (x) j ,y j ) Is the position coordinate of the wind generating set j, and theta is the included angle between the incoming wind direction and the appointed direction.
With reference to the second aspect, an embodiment of the present application provides a fourth possible implementation manner of the second aspect, where the dividing module is configured to divide wind power generator groups in a wind farm according to a wake interference relationship between any two wind power generator groups in the wind farm, so as to obtain a plurality of wind power generator groups, where the dividing module is specifically configured to:
Generating a wind power plant wake interference relation topological graph according to wake interference relations between any two wind power generator sets in the wind power plant;
and inputting an adjacency matrix for representing the wind farm wake interference relation topological graph into a depth-first search algorithm to obtain a plurality of wind turbine clusters.
With reference to the fourth possible implementation manner of the second aspect, the embodiment of the present application provides a fifth possible implementation manner of the second aspect, where the dividing module is specifically configured to, when inputting, into a depth-first search algorithm, an adjacency matrix for characterizing the wind farm wake interference relationship topology map, obtain a plurality of wind turbine groups:
in the depth-first search algorithm, any wind generating set is selected from wind generating sets which are not divided into wind generating sets, the wind generating sets which have direct wake interference and indirect wake interference with the wind generating sets are divided into the same wind generating set, and the selection of any wind generating set from the wind generating sets which are not divided into the wind generating sets and the subsequent steps are continuously executed until all the wind generating sets in the wind power plant are divided into the respective wind generating sets to stop.
With reference to the second aspect, an embodiment of the present application provides a sixth possible implementation manner of the second aspect, where the generating module is configured to, for each of the wind turbine groups, generate a plurality of groups of random yaw angles corresponding to the wind turbine group, so as to calculate a total power generated by the wind turbine group when a yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles, where the total power generated by the wind turbine group is specifically:
the total power generated by the wind turbine group is calculated by the following formula:
wherein P is n For the power generation of the nth wind generating set, ρ is the air density, C P Is the wind energy utilization coefficient of the wind generating set, S is the wind sweeping area of the impeller of the wind generating set, v is the incoming wind speed, and gamma n Is the random yaw angle of the wind generating set;
wherein P is WG And N is the number of wind generating sets in the wind turbine group, wherein N is the total power of the wind turbine group.
With reference to the sixth possible implementation manner of the second aspect, the embodiment of the present application provides a seventh possible implementation manner of the second aspect, where the generating module is configured to, for each of the wind turbine groups, generate a plurality of groups of random yaw angles corresponding to the wind turbine group, so as to calculate an average fatigue load of the wind turbine group when a yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles:
The average fatigue load of the wind farm is calculated by the following formula:
F n ＝f p +f t
wherein F is n Is the fatigue load of the nth wind generating set, f p Is fatigue load related to the generated power, f t Is fatigue load related to turbulence, P rated Is rated power of the nth wind generating set, T life The method is characterized in that the method is the estimated service life of an nth wind generating set, t is a preset time step, P (tau) is the real-time power of the nth wind generating set, W is a weight coefficient of turbulence fatigue, r is a maintenance compensation coefficient of the nth wind generating set, and I (tau) is the real-time turbulence of the nth wind generating set;
wherein F is WG Is the average fatigue load of the wind turbine group.
With reference to the second aspect, embodiments of the present application provide an eighth possible implementation manner of the second aspect, where the method further includes:
a second obtaining module, configured to obtain a first wind farm power generation and a first operation cost of the wind farm in a specified time step after the yaw angles of the wind turbine groups are adjusted by using the target yaw angle; acquiring second wind farm power generation and second operation cost of the wind farm in a designated time step before the yaw angle of each wind turbine generator set in the wind turbine group is adjusted by using the target yaw angle;
The continuous use module is used for continuously using the target yaw angle when the power generated by the first wind power plant is smaller than the power generated by the second wind power plant and the first operation and maintenance cost is greater than the second operation and maintenance cost;
and the re-executing module is used for dividing wind power generation groups of any two wind power generation units in the wind power plant according to wake interference relation between any two wind power generation units in the wind power plant when the power generation of the first wind power plant is not less than the power generation of the second wind power plant and/or the first operation and maintenance cost is not greater than the second operation and maintenance cost, so as to obtain a plurality of wind power generation groups and subsequent steps.
According to the yaw angle control method and device for the wind power plant, through adjusting the yaw angle of each wind generating set in the wind power plant, the wake area of each wind generating set is changed, interference of wake of each wind generating set to other wind generating sets is avoided, wake influence among the wind generating sets is reduced, and therefore the total power generation amount generated by the wind power plant is improved, and fatigue load of the wind generating sets in the wind power plant is reduced. Moreover, in this embodiment, through dividing the wind generating set in the wind farm, obtain a plurality of wind turbine clusters, adjust the yaw angle of wind generating set in every wind turbine cluster respectively, compare in adjusting the yaw angle of all wind generating sets in the wind farm simultaneously, the mode of this application is favorable to reducing the adjustment degree of difficulty. And because the wind generating set in the wind power plant is divided into a plurality of wind power clusters, parallel calculation can be carried out on each wind power cluster at the same time, and the adjustment time is shortened.
In order to make the above objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered limiting the scope, and that other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 illustrates a flowchart of a method for wind farm yaw control provided by an embodiment of the present application;
FIG. 2 shows a schematic diagram of wake interference relationship between two wind turbine generator sets provided by embodiments of the present application;
FIG. 3 illustrates a schematic diagram of the same wind farm provided by an embodiment of the present application;
fig. 4 shows a schematic structural diagram of a wind farm yaw control device according to an embodiment of the present application.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, are intended to be within the scope of the present application.
Considering that wake interference exists among wind generating sets, the generated energy of each wind generating set in a wind power plant is influenced, and extra fatigue load is caused. Based on this, the embodiment of the application provides a wind farm yaw angle control method and device, so as to avoid wake interference between wind generating sets, reduce wake influence, improve total generated energy of the wind farm, and reduce fatigue load of the wind generating sets in the wind farm.
Embodiment one:
for the sake of understanding the present embodiment, a method for controlling yaw angle of wind farm disclosed in the present embodiment will be described in detail first. Fig. 1 shows a flowchart of a wind farm yaw control method according to an embodiment of the present application, as shown in fig. 1, including the following steps S101 to S104:
s101: dividing wind power generator groups in the wind power plant according to wake interference relation between any two wind power generator groups in the wind power plant to obtain a plurality of wind power generator groups; wake interference does not exist among wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine group and at least one wind generating set in the wind turbine group.
In this embodiment, the wind farm includes a plurality of wind power generating sets, and after the wind power generating sets in the wind farm are divided into wind power generating sets, each wind power generating set includes at least one wind power generating set. The number of wind turbines included in different wind turbine clusters may or may not be the same.
For example, assuming 20 wind power generation sets in a wind power plant, after cluster division, 3 wind power generation sets A, B, C are obtained, wherein wind power generation set a contains 8 wind power generation sets, wind power generation set B contains 6 wind power generation sets, and wind power generation set C contains 6 wind power generation sets. Any wind generating set in the wind power generation group A and each wind generating set in the wind power generation group B have no wake interference, and each wind generating set in the wind power generation group C has no wake interference.
In this embodiment, fig. 2 shows a schematic diagram of a wake interference relationship between two wind generating sets provided in the embodiment of the present application, and as shown in fig. 2, a wind generating set a2 is located in a wake area of a wind generating set a1, so that wake interference exists between the wind generating set a1 and the wind generating set a 2. The wind generating set a3 is not located in the wake area of the wind generating set a1, and the wind generating set a1 is not located in the wake area of the wind generating set a3, so that wake interference does not exist between the wind generating set a1 and the wind generating set a 3.
Fig. 3 shows a schematic diagram of the same wind turbine group provided in the embodiment of the present application, where, as shown in fig. 3, the wind turbine group includes wind turbine groups b1, b2, b3, b4, b5, and for any wind turbine group in the wind turbine group, for example, wind turbine group b3, wake interference exists between wind turbine group b3 and wind turbine group b1 and wind turbine group b 5.
In one possible implementation, before performing step S101, it may further be that: inputting relative position coordinates of two wind power generator sets and incoming wind speed of the wind power plant into a pre-trained classified neural network model aiming at any two wind power generator sets in the wind power plant, and outputting wake interference relation between the two wind power generator sets; the wake interference relationship is the presence or absence of wake interference.
In this embodiment, the wake interference relationship between two wind turbine generator sets is distinguished by the classification neural network model, and compared with the traditional physical calculation method, the method in this embodiment is more efficient and more accurate.
In one possible embodiment, the relative position coordinates of two wind power units can be calculated by:
For any two wind generating sets in a wind power plant, inputting the position coordinates of the two wind generating sets and the included angle between the incoming wind direction and the designated direction into the following formula to obtain the relative position coordinates of the two wind generating sets:
wherein, (x) ij ,y ij ) Is the relative position coordinate between wind generating set i and wind generating set j, (x) i ,y i ) Is the position coordinate of the wind generating set i, (x) j ,y j ) Is the position coordinate of the wind generating set j, and theta is the included angle between the incoming wind direction and the appointed direction.
In this embodiment, the specified direction may be a north-positive direction, a south-positive direction, a west-positive direction, an east-positive direction, or the like, which is not limited in this application.
In one possible implementation, the classification neural network model is trained by:
acquiring a training sample set comprising a plurality of training samples; each training sample comprises sample relative position coordinates between two sample wind generating sets, sample incoming wind speed and a label for representing wake interference relation between the two sample wind generating sets;
for each training sample, inputting the sample relative position coordinates and sample incoming wind speed in the training sample into an initial classification neural network model, and outputting a predicted wake interference relationship between the two sample wind generating sets;
And calculating a cross entropy loss function according to the predicted wake interference relation and the label to obtain a loss value, training the learnable parameters in the initial classified neural network model by using the loss value, and stopping training until a training sample set is used for training for preset times to obtain the classified neural network model.
In this embodiment, the loss function takes the form of a cross entropy loss. The cross entropy may describe the distance between two probability distributions, with a smaller cross entropy indicating a closer approach between the two, calculated as:
wherein P is a determined probability distribution, i.e. a probability distribution to be approximated; q is the probability distribution represented by the neural network; x is x m Is the sample relative position coordinates and sample incoming wind speed contained in the mth training sample.
The cross entropy represents the distance between two probability distributions, and for the classified neural network model, a probability distribution capable of representing the discrimination condition needs to be output, so that the result obtained by forward propagation of the classified neural network model is changed into the probability distribution by adopting a Softmax regression method at the output layer of the classified neural network model. The formula is as follows:
wherein z is m 、z k The original output values of the m and k classification neural network models are Sof respectively m Is the mth output value through Softmax, i.e. Q (x m ) K is the number of output values and e is the natural exponent.
In this embodiment, the classified neural network model includes an input layer, a hidden layer, and an output layer, wherein the number of hidden layer neurons is set to 50, and the learning rate is set to 0.012.
In one possible implementation, when performing step S101, it may specifically be that: generating a wind power plant wake interference relation topological graph according to wake interference relations between any two wind power generator sets in the wind power plant; and inputting an adjacency matrix for representing the wind power plant wake interference relation topological graph into a depth-first search algorithm to obtain a plurality of wind power clusters.
In one possible implementation, when the adjacency matrix for representing the wind farm wake interference relation topological graph is input into the depth-first search algorithm to obtain a plurality of wind turbine groups, the method specifically can be as follows: in the depth-first search algorithm, any wind generating set is selected from wind generating sets which are not divided into wind generating sets, the wind generating sets which have direct wake interference and indirect wake interference with the wind generating sets are divided into the same wind generating set, and the selection of any wind generating set from the wind generating sets which are not divided into the wind generating sets and the subsequent steps are continuously executed until all the wind generating sets in the wind farm are divided into the respective wind generating sets to stop.
In this embodiment, the direct wake interference means that one of the two wind power units is located in the wake zone of the other wind power unit, and as shown in fig. 3, there is a direct wake interference between the wind power unit b1 and the wind power unit b 2.
The indirect wake interference means that direct wake interference exists between the two wind generating sets and other wind generating sets; when the other wind generating sets are multiple, wake interference exists among the other wind generating sets; when the number of other wind generating sets is 1, direct wake interference exists between the two wind generating sets and the same other wind generating sets. Illustratively, as shown in FIG. 3, there is an indirect wake interference between wind turbine b2 and wind turbine b2, where wind turbine b2 and wind turbine b2 have a direct wake interference with wind turbine b 1. An indirect wake interference exists between the wind generating set b4 and the wind generating set b5, and at this time, an (indirect) wake interference exists between the wind generating set b4 and the wind generating set b1, and an (indirect) wake interference also exists between the wind generating set b5 and the wind generating set b1, so that an indirect wake interference also exists between the wind generating set b4 and the wind generating set b 5.
S102: and generating a plurality of groups of random yaw angles corresponding to each wind turbine group aiming at each wind turbine group so as to calculate the total power generation and the average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles.
And randomly generating a plurality of groups of random yaw angles corresponding to each wind turbine group aiming at each wind turbine group, wherein the number of the random yaw angles contained in each group of random yaw angles corresponding to the wind turbine group is the same as the number of wind generating sets contained in the wind turbine group. Each random yaw angle contained in each group of random yaw angles corresponding to the wind turbine group corresponds to each wind turbine group in a one-to-one mode. Each group of random yaw angles corresponds to one total generated power and one fatigue load, and each wind turbine group corresponds to a plurality of total generated powers and a plurality of fatigue loads.
For example, the wind power generator group A comprises 8 wind power generator groups, the wind power generator group A corresponds to 5 groups of random yaw angles, each group of random yaw angles comprises 8 random yaw angles, and the 8 random yaw angles correspond to the 8 wind power generator groups one by one. The wind turbine group a corresponds to 5 total power generation and 5 fatigue loads.
In this embodiment, the yaw angle refers to the angle between the normal line of the impeller plane and the incoming wind direction, and when the impeller plane is perpendicular to the incoming wind direction, the yaw angle is 0.
In one possible implementation manner, when performing step S102, the total power generated by the wind turbine group corresponding to each set of random yaw angles may be calculated specifically by:
the total power generated by the wind turbine group is calculated by the following formula:
wherein P is n For the power generation of the nth wind generating set, ρ is the air density, C P Is the wind energy utilization coefficient of the wind generating set, S is the wind sweeping area of the impeller of the wind generating set, v is the incoming wind speed, and gamma n Is the random yaw angle of the wind generating set;
wherein P is WG And N is the number of wind generating sets in the wind turbine group, wherein N is the total power of the wind turbine group.
The above-mentioned process is only the total power generated by the wind power generation group calculated by using a set of random yaw angles, and the total power generated by the wind power generation group corresponding to each set of random yaw angles is calculated by using each set of random yaw angles respectively.
In a possible embodiment, when performing step S102, the average fatigue load of the wind turbine group corresponding to each set of random yaw angles may be calculated in particular by:
The average fatigue load of the wind farm is calculated by the following formula:
F n ＝f p +f t
wherein F is n Is the fatigue load of the nth wind generating set, f p Is fatigue load related to the generated power, f t Is fatigue load related to turbulence, P rated Is rated power of the nth wind generating set, T life The method is characterized in that the method is the estimated service life of an nth wind generating set, t is a preset time step, P (tau) is the real-time power of the nth wind generating set, W is a weight coefficient of turbulence fatigue, r is a maintenance compensation coefficient of the nth wind generating set, and I (tau) is the real-time turbulence of the nth wind generating set;
wherein F is WG Is the average fatigue load of the wind turbine group.
In this embodiment, turbulence refers to the steady state of the wind, with no turbulence. Similarly, the above-mentioned process is only the average fatigue load of the wind turbine group calculated by using a set of random yaw angles, and by the above-mentioned process, the average fatigue load of the wind turbine group corresponding to each set of random yaw angles is calculated by using each set of random yaw angles, respectively.
S103: and inputting each group of random yaw angles corresponding to the wind turbine groups and total power generation and average fatigue load corresponding to each group of random yaw angles into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting the target yaw angle of each wind turbine group in the wind turbine groups.
In this embodiment, each wind turbine in the wind farm corresponds to a respective target yaw angle.
S104: and adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle.
In this embodiment, according to the corresponding relationship between the target yaw angle and each wind turbine group, each target yaw angle is issued to each corresponding wind turbine group, so that each yaw angle is adjusted by using each corresponding target yaw angle of each wind turbine group.
In one possible implementation, after step S104 is performed, it may further: acquiring first wind farm power generation and first operation cost of a wind farm in a designated time step after yaw angles of all wind turbine generator sets in a wind turbine group are adjusted by using a target yaw angle; acquiring second wind farm power generation and second operation cost of the wind farm in a designated time step before the yaw angles of the wind turbine groups are adjusted by using the target yaw angles;
when the power generated by the first wind farm is smaller than that generated by the second wind farm, and the first operation cost is greater than the second operation cost, continuing to use the target yaw angle;
And when the power generated by the first wind farm is not less than the power generated by the second wind farm, and/or the first operation cost is not greater than the second operation cost, the step of re-executing is performed to divide wind power generation groups in the wind farm according to the wake interference relationship between any two wind power generation groups in the wind farm, so as to obtain a plurality of wind power generation groups and subsequent steps.
In this embodiment, by comparing the power generated by the first wind farm with the power generated by the second wind farm and comparing the magnitude relation between the first operation cost and the second operation cost, it is determined whether the yaw angle of the wind turbine in the wind farm needs to be readjusted.
Embodiment two:
based on the same technical concept, the present application further provides a wind farm yaw angle control device, and fig. 4 shows a schematic structural diagram of the wind farm yaw angle control device provided by the embodiment of the present application, as shown in fig. 4, where the device includes:
the division module 401 is configured to divide wind power generator groups in a wind power plant according to wake interference relationships between any two wind power generator groups in the wind power plant, so as to obtain a plurality of wind power generator groups; no wake interference exists between the wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine cluster and at least one wind generating set in the wind turbine cluster;
A generating module 402, configured to generate, for each of the wind turbine groups, a plurality of groups of random yaw angles corresponding to the wind turbine group, so as to calculate a total power generated by the wind turbine group and an average fatigue load when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles;
the first input module 403 is configured to input each set of random yaw angles corresponding to the wind turbine groups and the total generated power and the average fatigue load corresponding to each set of random yaw angles into an adaptive evaluation multi-target particle swarm algorithm, and output a target yaw angle of each wind turbine group in the wind turbine groups;
an adjustment module 404 for adjusting yaw angles of individual wind turbines in the wind turbine farm using the target yaw angle.
Optionally, the method further comprises:
the second input module is configured to, before the dividing module 401 divides wind power generation sets in the wind power plant into wind power generation sets according to wake interference relationships between any two wind power generation sets in the wind power plant to obtain a plurality of wind power generation sets, input relative position coordinates of the two wind power generation sets and incoming wind speeds of the wind power plant into a pre-trained classification neural network model for any two wind power generation sets in the wind power plant, and output the wake interference relationships between the two wind power generation sets; the wake interference relationship is the presence or absence of wake interference.
Optionally, the method further comprises:
the first acquisition module is used for acquiring a training sample set containing a plurality of training samples; each training sample comprises sample relative position coordinates between two sample wind generating sets, sample incoming wind speed and a label for representing wake interference relation between the two sample wind generating sets;
the third input module is used for inputting the relative position coordinates of the samples in the training samples and the incoming wind speeds of the samples into an initial classification neural network model for each training sample, and outputting a predicted wake interference relationship between the two sample wind generating sets;
and the training module is used for calculating a cross entropy loss function according to the predicted wake interference relation and the label to obtain a loss value so as to train the learnable parameters in the initial classified neural network model by using the loss value, and stopping training until training is performed for preset times by using the training sample set to obtain the classified neural network model.
Optionally, the method further comprises:
the fourth input module is configured to input, for any two of the wind power generator sets in the wind farm, the relative position coordinates of the two wind power generator sets and the incoming wind speed of the wind power farm into a pre-trained classification neural network model, and input, for any two of the wind power generator sets in the wind farm, the position coordinates of the two wind power generator sets and the included angle between the incoming wind direction and the specified direction into the following formula before outputting the wake interference relationship between the two wind power generator sets, to obtain the relative position coordinates of the two wind power generator sets:
Wherein, (x) ij ,y ij ) Is the relative position coordinate between wind generating set i and wind generating set j, (x) i ,y i ) Is the position coordinate of the wind generating set i, (x) j ,y j ) Is the position seat of the wind generating set jAnd θ is the angle between the incoming wind direction and the designated direction.
Optionally, the dividing module 401 is configured to divide wind power generator groups in a wind farm according to wake interference relationships between any two wind power generator groups in the wind farm, so as to obtain a plurality of wind power generator groups, and is specifically configured to:
generating a wind power plant wake interference relation topological graph according to wake interference relations between any two wind power generator sets in the wind power plant;
and inputting an adjacency matrix for representing the wind farm wake interference relation topological graph into a depth-first search algorithm to obtain a plurality of wind turbine clusters.
Optionally, the dividing module 401 is specifically configured to, when inputting the adjacency matrix for characterizing the wind farm wake interference relationship topological graph into a depth-first search algorithm to obtain a plurality of wind turbine groups:
in the depth-first search algorithm, any wind generating set is selected from wind generating sets which are not divided into wind generating sets, the wind generating sets which have direct wake interference and indirect wake interference with the wind generating sets are divided into the same wind generating set, and the selection of any wind generating set from the wind generating sets which are not divided into the wind generating sets and the subsequent steps are continuously executed until all the wind generating sets in the wind power plant are divided into the respective wind generating sets to stop.
Optionally, when the generating module 402 is configured to generate, for each of the wind turbine groups, a plurality of sets of random yaw angles corresponding to the wind turbine group, so as to calculate yaw angles of each wind turbine group in the wind turbine group as each random yaw angle in each set of random yaw angles, the generating module is specifically configured to:
the total power generated by the wind turbine group is calculated by the following formula:
wherein P is n For the power generation of the nth wind generating set, ρ is the air density, C P Is the wind energy utilization coefficient of the wind generating set, S is the wind sweeping area of the impeller of the wind generating set, v is the incoming wind speed, and gamma n Is the random yaw angle of the wind generating set;
wherein P is WG And N is the number of wind generating sets in the wind turbine group, wherein N is the total power of the wind turbine group.
Optionally, when the generating module 402 is configured to generate, for each of the wind turbine groups, a plurality of sets of random yaw angles corresponding to the wind turbine group, so as to calculate yaw angles of each wind turbine group in the wind turbine group as each random yaw angle in each set of random yaw angles, the average fatigue load of the wind turbine group is specifically configured to:
the average fatigue load of the wind farm is calculated by the following formula:
F n ＝f p +f t
Wherein F is n Is the fatigue load of the nth wind generating set, f p Is fatigue load related to the generated power, f t Is fatigue load related to turbulence, P rated Is rated power of the nth wind generating set, T life Is the expected service life of the nth wind generating set, t is a preset time step, P (tau) is the real-time power of the nth wind generating set, W is the weight coefficient of turbulent fatigue, and r is the nth wind powerThe maintenance compensation coefficient of the generator set, I (tau) is the real-time turbulence of the nth wind generator set;
wherein F is WG Is the average fatigue load of the wind turbine group.
Optionally, the method further comprises:
a second obtaining module, configured to obtain a first wind farm power generation and a first operation cost of the wind farm in a specified time step after the yaw angles of the wind turbine groups are adjusted by using the target yaw angle; acquiring second wind farm power generation and second operation cost of the wind farm in a designated time step before the yaw angle of each wind turbine generator set in the wind turbine group is adjusted by using the target yaw angle;
the continuous use module is used for continuously using the target yaw angle when the power generated by the first wind power plant is smaller than the power generated by the second wind power plant and the first operation and maintenance cost is greater than the second operation and maintenance cost;
And the re-executing module is used for dividing wind power generation groups of any two wind power generation units in the wind power plant according to wake interference relation between any two wind power generation units in the wind power plant when the power generation of the first wind power plant is not less than the power generation of the second wind power plant and/or the first operation and maintenance cost is not greater than the second operation and maintenance cost, so as to obtain a plurality of wind power generation groups and subsequent steps.
It will be clear to those skilled in the art that, for convenience and brevity of description, reference may be made to the corresponding process in the foregoing method embodiment for the specific working process of the apparatus described above, which is not described herein again.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described apparatus embodiments are merely illustrative, and the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, and for example, multiple modules or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Finally, it should be noted that: the foregoing examples are merely specific embodiments of the present application, and are not intended to limit the scope of the present application, but the present application is not limited thereto, and those skilled in the art will appreciate that while the foregoing examples are described in detail, the present application is not limited thereto. Any person skilled in the art may modify or easily conceive of the technical solution described in the foregoing embodiments, or make equivalent substitutions for some of the technical features within the technical scope of the disclosure of the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method of controlling yaw angle of a wind farm, the method comprising:
dividing wind power generator groups in a wind power plant according to wake interference relation between any two wind power generator groups in the wind power plant to obtain a plurality of wind power generator groups; no wake interference exists between the wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine cluster and at least one wind generating set in the wind turbine cluster;
Generating multiple groups of random yaw angles corresponding to each wind turbine group aiming at each wind turbine group so as to calculate the total power generation and average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles;
inputting each group of random yaw angles corresponding to the wind turbine groups and the total power generated by each group of random yaw angles and the average fatigue load into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting a target yaw angle of each wind turbine group in the wind turbine groups;
and adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle.
2. The method according to claim 1, wherein the wind turbine generator sets in the wind farm are divided into wind turbine groups according to wake interference relationships between any two wind turbine generator sets in the wind farm, and before obtaining the plurality of wind turbine groups, the method further comprises:
inputting relative position coordinates of two wind power generator sets and incoming wind speed of the wind power plant into a pre-trained classified neural network model aiming at any two wind power generator sets in the wind power plant, and outputting the wake interference relationship between the two wind power generator sets; the wake interference relationship is the presence or absence of wake interference.
3. The method of claim 2, wherein the classification neural network model is trained by:
acquiring a training sample set comprising a plurality of training samples; each training sample comprises sample relative position coordinates between two sample wind generating sets, sample incoming wind speed and a label for representing wake interference relation between the two sample wind generating sets;
inputting the relative position coordinates of the samples in the training samples and the incoming wind speeds of the samples into an initial classification neural network model for each training sample, and outputting a predicted wake interference relationship between the two sample wind generating sets;
and calculating a cross entropy loss function according to the predicted wake interference relation and the label to obtain a loss value, so as to train the learnable parameters in the initial classified neural network model by using the loss value, and stopping training until training is performed for a preset number of times by using the training sample set, thus obtaining the classified neural network model.
4. The method according to claim 2, wherein the method further comprises, for any two of the wind power plants, before inputting the relative position coordinates of the two wind power plants and the incoming wind speed of the wind power plant into a pre-trained classification neural network model:
Inputting the position coordinates of any two wind generating sets in the wind power plant and the included angle between the incoming wind direction and the designated direction into the following formula to obtain the relative position coordinates of the two wind generating sets:
wherein, (x) ij ,y ij ) Is the relative position coordinate between wind generating set i and wind generating set j, (x) i ,y i ) Is the position coordinate of the wind generating set i, (x) j ,y j ) Is the position coordinate of the wind generating set j, and theta is the included angle between the incoming wind direction and the appointed direction.
5. The method according to claim 1, wherein the dividing the wind turbine generator sets in the wind farm into a plurality of wind turbine generator sets according to wake interference relation between any two wind turbine generator sets in the wind farm includes:
generating a wind power plant wake interference relation topological graph according to wake interference relations between any two wind power generator sets in the wind power plant;
and inputting an adjacency matrix for representing the wind farm wake interference relation topological graph into a depth-first search algorithm to obtain a plurality of wind turbine clusters.
6. The method of claim 5, wherein inputting an adjacency matrix characterizing the wind farm wake interference relationship topology into a depth-first search algorithm results in a plurality of the wind clusters, comprising:
In the depth-first search algorithm, any wind generating set is selected from wind generating sets which are not divided into wind generating sets, the wind generating sets which have direct wake interference and indirect wake interference with the wind generating sets are divided into the same wind generating set, and the selection of any wind generating set from the wind generating sets which are not divided into the wind generating sets and the subsequent steps are continuously executed until all the wind generating sets in the wind power plant are divided into the respective wind generating sets to stop.
7. The method of claim 1, wherein generating, for each of the wind turbine clusters, a plurality of sets of random yaw angles corresponding to the wind turbine cluster to calculate a total generated power for the wind turbine cluster when the yaw angle of each wind turbine cluster is each random yaw angle of each set of random yaw angles, comprises:
the total power generated by the wind turbine group is calculated by the following formula:
wherein P is n For the power generation of the nth wind generating set, ρ is the air density, C P Is the wind energy utilization coefficient of the wind generating set, S is the wind sweeping area of the impeller of the wind generating set, v is the incoming wind speed, and gamma n Is the random yaw angle of the wind generating set;
wherein P is WG And N is the number of wind generating sets in the wind turbine group, wherein N is the total power of the wind turbine group.
8. The method of claim 7, wherein generating, for each of the wind turbine clusters, a plurality of sets of random yaw angles corresponding to the wind turbine cluster to calculate an average fatigue load for each wind turbine cluster when the yaw angle of each wind turbine cluster is each of the plurality of sets of random yaw angles, comprises:
the average fatigue load of the wind farm is calculated by the following formula:
F n ＝f p +f t
wherein F is n Is the fatigue load of the nth wind generating set, f p Is fatigue load related to the generated power, f t Is fatigue load related to turbulence, P rated Is rated power of the nth wind generating set, T life The method is characterized in that the method is the estimated service life of an nth wind generating set, t is a preset time step, P (tau) is the real-time power of the nth wind generating set, W is a weight coefficient of turbulence fatigue, r is a maintenance compensation coefficient of the nth wind generating set, and I (tau) is the real-time turbulence of the nth wind generating set;
wherein F is WG Is the average fatigue load of the wind turbine group.
9. The method according to claim 1, wherein the method further comprises:
acquiring a first wind farm power generation and a first operation cost of the wind farm in a designated time step after the yaw angles of the wind turbine generator sets in the wind turbine generator group are adjusted by using the target yaw angle; acquiring second wind farm power generation and second operation cost of the wind farm in a designated time step before the yaw angle of each wind turbine generator set in the wind turbine group is adjusted by using the target yaw angle;
when the first wind farm power is smaller than the second wind farm power, and the first operation cost is greater than the second operation cost, continuing to use the target yaw angle;
and when the power generated by the first wind farm is not less than the power generated by the second wind farm, and/or the first operation cost is not greater than the second operation cost, re-executing the step to divide wind power groups of the wind power generator groups in the wind farm according to the wake interference relationship between any two wind power generator groups in the wind farm to obtain a plurality of wind power groups and subsequent steps.
10. A wind farm yaw control device, the device comprising:
the division module is used for dividing wind power generation sets in the wind power plant into wind power generation sets according to wake interference relation between any two wind power generation sets in the wind power plant to obtain a plurality of wind power generation sets; no wake interference exists between the wind generating sets in different wind generating sets; wake interference exists between any wind generating set in the same wind turbine cluster and at least one wind generating set in the wind turbine cluster;
the generation module is used for generating a plurality of groups of random yaw angles corresponding to each wind turbine group so as to calculate the total power generation and the average fatigue load of each wind turbine group when the yaw angle of each wind turbine group in the wind turbine group is each random yaw angle in each group of random yaw angles;
the first input module is used for inputting each group of random yaw angles corresponding to the wind turbine groups and the total power generation and the average fatigue load corresponding to each group of random yaw angles into a self-adaptive evaluation multi-target particle swarm algorithm, and outputting the target yaw angle of each wind turbine group in the wind turbine groups;
And the adjusting module is used for adjusting the yaw angle of each wind generating set in the wind generating set by using the target yaw angle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310492612.6A CN116378897B (en) | 2023-05-04 | 2023-05-04 | Wind farm yaw angle control method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310492612.6A CN116378897B (en) | 2023-05-04 | 2023-05-04 | Wind farm yaw angle control method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116378897A true CN116378897A (en) | 2023-07-04 |
| CN116378897B CN116378897B (en) | 2023-12-26 |
Family
ID=86973401
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310492612.6A Active CN116378897B (en) | 2023-05-04 | 2023-05-04 | Wind farm yaw angle control method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116378897B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110094297A (en) * | 2018-01-31 | 2019-08-06 | 北京金风科创风电设备有限公司 | Control method and control system of wind generating set based on sectors |
| EP3792484A1 (en) * | 2019-09-16 | 2021-03-17 | Siemens Gamesa Renewable Energy A/S | Wind turbine yaw offset control based on reinforcement learning |
| US20210246881A1 (en) * | 2020-02-11 | 2021-08-12 | Inventus Holdings, Llc | Wind turbine drivetrain wear detection using azimuth variation clustering |
| CN114021346A (en) * | 2021-11-04 | 2022-02-08 | 中南大学 | Wind power plant distributed operation optimization method based on wake directed graph |
| CN114692397A (en) * | 2022-03-16 | 2022-07-01 | 中山大学 | Cluster distributed capture method based on multi-mechanism combination strategy |
| CN114899876A (en) * | 2022-07-12 | 2022-08-12 | 华北电力大学 | Wind power plant operation control method, device, equipment and storage medium |
| CN115333168A (en) * | 2022-09-20 | 2022-11-11 | 浙江大学 | Offshore wind farm field level control strategy based on distributed rolling optimization |
| CN116025512A (en) * | 2023-03-02 | 2023-04-28 | 华北电力大学 | Wind driven generator group control method, device, equipment and storage medium |
-
2023
- 2023-05-04 CN CN202310492612.6A patent/CN116378897B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110094297A (en) * | 2018-01-31 | 2019-08-06 | 北京金风科创风电设备有限公司 | Control method and control system of wind generating set based on sectors |
| EP3792484A1 (en) * | 2019-09-16 | 2021-03-17 | Siemens Gamesa Renewable Energy A/S | Wind turbine yaw offset control based on reinforcement learning |
| US20210246881A1 (en) * | 2020-02-11 | 2021-08-12 | Inventus Holdings, Llc | Wind turbine drivetrain wear detection using azimuth variation clustering |
| CN114021346A (en) * | 2021-11-04 | 2022-02-08 | 中南大学 | Wind power plant distributed operation optimization method based on wake directed graph |
| CN114692397A (en) * | 2022-03-16 | 2022-07-01 | 中山大学 | Cluster distributed capture method based on multi-mechanism combination strategy |
| CN114899876A (en) * | 2022-07-12 | 2022-08-12 | 华北电力大学 | Wind power plant operation control method, device, equipment and storage medium |
| CN115333168A (en) * | 2022-09-20 | 2022-11-11 | 浙江大学 | Offshore wind farm field level control strategy based on distributed rolling optimization |
| CN116025512A (en) * | 2023-03-02 | 2023-04-28 | 华北电力大学 | Wind driven generator group control method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116378897B (en) | 2023-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108805434B (en) | Improved NSGA-III-based multi-objective optimization scheduling method for cascade hydropower station | |
| CN112459965B (en) | Yaw optimization control method, device, equipment and medium considering wind power plant wake flow | |
| Shokrzadeh et al. | Wind turbine power curve modeling using advanced parametric and nonparametric methods | |
| CN104632521B (en) | A kind of wind power optimization system and method based on drift correction | |
| CN109670213B (en) | Multi-machine equivalent modeling method and device for wind power plant | |
| CN113591359B (en) | Wind turbine generator set cut-in/cut-out wind speed adjusting and optimizing method, system and equipment medium | |
| CN108022019B (en) | Wind power plant scheduling method and system based on wind turbine generator classification | |
| Lu et al. | Study on the optimal hydropower generation of Zhelin reservoir | |
| CN106208046A (en) | A kind of tidal current energy generating field unit layout method considering power generation settings cost | |
| Gualtieri | Comparative analysis and improvement of grid-based wind farm layout optimization | |
| CN109800931A (en) | Wind power plant generated energy loss measurement method and system based on blower SCADA data | |
| CN111091236A (en) | Multi-classification deep learning short-term wind power prediction method classified according to pitch angles | |
| CN113139326A (en) | Wind power plant yaw cooperative control optimization method, electronic device and storage medium | |
| Yu et al. | Reinforcement learning-based multi-objective differential evolution for wind farm layout optimization | |
| CN116378897B (en) | Wind farm yaw angle control method and device | |
| CN110397553A (en) | A kind of wind power plant wake flow management method and system being not based on model | |
| Yao et al. | New design of a wind farm frequency control considering output uncertainty and fatigue suppression | |
| Sun et al. | Wind farm layout and hub height optimization with a novel wake model | |
| CN116581827A (en) | Wind farm unit start-stop strategy under load limiting condition considering wake flow influence | |
| CN116658359A (en) | Real-time collaborative yaw control method for wind farm | |
| Su et al. | A coordinative optimization method of active power and fatigue distribution in onshore wind farms | |
| CN116221021A (en) | Wind farm collaborative yaw control method based on multilayer artificial intelligence system | |
| Li et al. | Discrete complex-valued code pathfinder algorithm for wind farm layout optimization problem | |
| CN115995840A (en) | New energy system peak regulation optimization method and device based on coordination of wind, light and energy storage | |
| CN109742752B (en) | Wind power plant power generation optimization method considering real-time electricity price change |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |