CN112418553A - Offshore wind power control method based on VMD-CNN network - Google Patents
Offshore wind power control method based on VMD-CNN network Download PDFInfo
- Publication number
- CN112418553A CN112418553A CN202011416488.8A CN202011416488A CN112418553A CN 112418553 A CN112418553 A CN 112418553A CN 202011416488 A CN202011416488 A CN 202011416488A CN 112418553 A CN112418553 A CN 112418553A
- Authority
- CN
- China
- Prior art keywords
- data
- speed
- vmd
- wind
- wind speed
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 22
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 5
- 150000001875 compounds Chemical class 0.000 claims description 3
- 230000003190 augmentative effect Effects 0.000 claims description 2
- 238000013527 convolutional neural network Methods 0.000 description 23
- 238000004364 calculation method Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 238000011161 development Methods 0.000 description 6
- 230000006872 improvement Effects 0.000 description 4
- 230000001965 increasing effect Effects 0.000 description 3
- 238000010248 power generation Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 239000002028 Biomass Substances 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06N3/045—Combinations of 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
-
- 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
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Wind Motors (AREA)
- Control Of Eletrric Generators (AREA)
Abstract
The invention discloses an offshore wind power control method based on a VMD-CNN network, which comprises the following steps: s1: decomposing historical wind speed data through a VMD technology; s2: the wind speed data and the DFIG active power, the generator rotating speed and the pitch angle data at the corresponding moment are combined for data preprocessing; s3: establishing a CNN network model, and training the CNN network model by using historical data; s4: obtaining a wind speed prediction result by utilizing the trained network model; s5: and detecting a wind speed prediction result, calculating the reference rotating speed of the fan rotor according to the wind speed prediction value, and performing different control operations on the DFIG according to the reference rotating speed of the fan rotor.
Description
Technical Field
The invention belongs to the technical field of offshore wind power generation, and particularly relates to an offshore wind power control method based on a VMD-CNN network.
Background
In retrospect over the past 20 years, the increasing deterioration of global environment is an important reason for hindering economic development, and the rational exploitation of renewable energy sources would be an effective countermeasure. In this regard, countries around the world have increased support for renewable energy research, and have been planned to deal with the global problem of "energy crisis". Renewable energy sources include solar energy, hydroenergy, wind energy, biomass energy, wave energy, tidal energy, ocean thermal energy, geothermal energy, and the like. Among various renewable energy sources, wind energy is widely favored by countries all over the world by virtue of a plurality of advantages, the wind power technology is the most mature, and the development and utilization scale is the largest.
With the continuous progress and maturity of the installation and manufacturing technology of the offshore wind turbine, the single machine capacity of the offshore wind turbine is continuously improved, the scale of the offshore wind farm is enlarged, and the algorithm calculation amount of the whole wind farm is increased. Because the control algorithm of the traditional offshore wind power generation set has large calculation amount, an intelligent algorithm with small calculation amount and high prediction precision is an important support for keeping the high-speed development of offshore wind power.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, effectively solves the problem of large calculation amount of the traditional maximum power tracking algorithm by adopting a deep learning algorithm, reduces the dependence on wind speed, and provides the offshore wind power control method based on the VMD-CNN network by fully considering the running state of the offshore wind power system to accurately predict the wind speed at the next moment.
In order to achieve the purpose, the invention adopts the following technical scheme:
an offshore wind power control method based on a VMD-CNN network comprises the following steps:
s1: decomposing historical wind speed data through a VMD technology;
s2: the wind speed data and the DFIG active power, the generator rotating speed and the pitch angle data at the corresponding moment are combined for data preprocessing;
s3: establishing a CNN network model, and training the CNN network model by using historical data;
s4: obtaining a wind speed prediction result by utilizing the trained network model;
s5: and checking a wind speed prediction result, calculating the reference rotating speed of the fan rotor according to the wind speed prediction value, and performing different control operations on the DFIG according to the reference rotating speed of the fan rotor.
In a further improvement of the present invention, in step S1, the historical wind speed data is decomposed by VMD technique, and the formula is described as follows:
in the formula, K is the modal number of the wind speed sequence to be decomposed, { u (K) }, { ω (K) } respectively corresponds to the K-th modal component and the center frequency after decomposition, δ (t) is a dirac function, and x is a convolution operator.
The invention is further improved, the Lagrange expression is enlarged to solve the above formula, and the description formula is as follows:
in the formula, λ is Lagrange multiplier, and α is secondary penalty factor.
In a further development of the invention, the data preprocessing in step S2 includes the combination of wind speed data with DFIG active power, generator speed and pitch angle data at their respective moments and data normalization.
The invention further improves the method, 4 groups of eigen-mode functions acquired by decomposing the wind speed sequence and DFIG active power, generator rotating speed and pitch angle data at corresponding moments are combined and grouped according to the principle that 8 time sequences form one group, the first seven data of each group are extracted and divided into a plurality of groups of 7 x 7 matrixes, and the eighth data of each group is used as a wind speed result corresponding to the eighth data.
According to the further improvement of the invention, each group of data is subjected to normalization operation.
In a further refinement of the present invention, step S3 includes building a 6-layer CNN model and training the network.
In a further improvement of the present invention, the 6-layer CNN model is composed of 2 convolutional layers, 2 pooling layers and 2 fully-connected layers, wherein the sizes of the convolutional cores of the 2 convolutional layers are all selected to be 2 × 2, and the step sizes of the pooling layers are all selected to be 2.
In a further development of the invention, step S5 includes calculating a fan rotor reference speed and performing different control operations on the DFIG based on the fan rotor reference speed.
In a further improvement of the present invention, the calculation of the reference rotation speed of the fan rotor is described by the formula:
in the formula (I), the compound is shown in the specification,reference speed for the fan rotor, K gear ratio, R turbine radius, λoptFor optimum tip speed ratio, VstimIs an estimate of wind speed.
The invention further improves that the control operation is mainly as follows: if the rotating speed of the rotor is lower than the rated rotating speed, adopting a mode one, calculating the corresponding optimal rotating speed of the rotor and taking the pitch angle as zero; if the rotating speed of the rotor is higher than the rated rotating speed, a mode II is adopted, and the wind energy utilization coefficient of the wind driven generator is changed by adjusting the pitch angle, so that the output power is stabilized at the rated power to ensure the running safety of the fan.
The invention has the beneficial effects that: the method overcomes the defects in the prior art, effectively solves the problem of large calculation amount of the traditional maximum power tracking algorithm by adopting a deep learning algorithm, and reduces the dependence on wind speed; the DSP is adopted to realize the function of the mode switching system controller, so that the hardware development cost is effectively reduced, and the approach precision of the prediction model to the output power of the nonlinear wind driven generator is improved. Compared with the prior art, the wind speed at the next moment is predicted in advance according to the state of the DFIG at the current moment and the wind speed at the past moment, so that the DFIG can react to the wind speed at the next moment more quickly, and the power generation efficiency is improved.
Drawings
FIG. 1 is a schematic diagram of the overall model structure in the present invention.
FIG. 2 is a flow chart of the VMD algorithm in the present invention.
FIG. 3 is a schematic diagram of the VMD-CNN model structure in the present invention.
Fig. 4 is a structure diagram of CNN in the present invention.
Fig. 5 is a schematic diagram of the operation mode switching in the present invention.
Fig. 6 is a block diagram of a mode switching system implementation in the present invention.
Detailed Description
For the purpose of enhancing the understanding of the present invention, the present invention will be described in further detail with reference to the accompanying drawings and examples, which are provided for the purpose of illustration only and are not intended to limit the scope of the present invention.
Relation between wind speed and output power of offshore wind farm
Wind is a main factor influencing the output power of an offshore wind farm, and the wind speed can change constantly within a certain range and has randomness. The wind power generated by the wind generator can be written as:
Pm=Cp(λ,β)ρAv3/2
in the formula, PmIs the mechanical power (W, C) of the offshore wind turbinepIs the power coefficient of the fan, and ρ is the air density (kg/m)3) A isSwept area of fan (m)2) V is the wind speed (m/s) and λ is the tip speed ratio.
VMD decomposition of wind speed
The Variable Modal Decomposition (VMD) is a self-adaptive and completely non-recursive modal variation and signal processing method, and the technology has the advantages of determining the number of modal decompositions, and the self-adaptability of the technology is represented by determining the number of modal decompositions of a given sequence according to the actual situation, and the subsequent searching and solving processes can be self-adaptively matched with the optimal central frequency and limited bandwidth of each modal, and can realize the effective separation of inherent modal components (IMF), the frequency domain division of signals, further obtain the effective decomposition components of given signals, and finally obtain the optimal solution of the variation problem. The flow chart of the VMD algorithm is shown in FIG. 2. Because the wind speed sequence has the characteristics of nonlinearity and instability, the error of the final wind speed prediction result is larger by directly taking the original wind speed sequence as input. Therefore, the inherent modal component of the historical wind speed sequence is obtained by adopting the VMD technology, and the description formula is as follows:
in the formula, K is the number of modes of the wind speed sequence to be decomposed (in the present invention, K is taken to be 4), { u (K) }, { ω (K) } respectively correspond to the K-th mode component and the center frequency after decomposition, δ (t) is a dirac function, and δ (t) is a convolution operator.
Further, the above formula is solved by augmenting Lagrange's expression, which describes the formula as:
in the formula, λ is Lagrange multiplier, and α is secondary penalty factor.
Short-term wind speed prediction model based on VMD-CNN network
And selecting historical wind power data as a sample of the prediction model. The structure diagram of the VMD-CNN prediction model is shown in FIG. 3. In the present invention, the prediction model is divided into two modules: a VMD-based wind speed decomposition module and a CNN-based prediction module. Firstly, decomposing an original wind speed sequence into a series of relatively stable and stable inherent mode functions by adopting a VMD (virtual model decomposition); secondly, combining 4 groups of eigen-mode functions acquired by decomposing a wind speed sequence with DFIG active power, generator rotating speed and pitch angle data at corresponding moments, grouping according to the principle that 8 time sequences form one group, extracting and dividing the first seven data of each group into a plurality of groups of 7 x 7 matrixes, and taking the eighth data of each group as a wind speed result corresponding to the eighth data; then, carrying out normalization processing on the data; and finally, feature extraction and wind speed prediction are realized through CNN.
The structure of the convolutional neural network is shown in fig. 4. The input of the convolutional neural network is a normalized 7 multiplied by 7 time sequence matrix formed by combining 4 groups of eigen mode functions obtained through VMD decomposition and DFIG active power, generator rotating speed and pitch angle at corresponding moment, and the output is predicted wind speed V of the next 1 time sequencestim. The established network model has 6 hyper-parameters: the number of CNN convolutional layers and the size of convolutional cores, the number of pooling layers and step length, and the dimension of output layers. After a plurality of training, the parameters are selected as follows: the 6-layer CNN model consists of 2 convolutional layers, 2 pooling layers and 2 full-connection layers, wherein the sizes of convolutional cores of the 2 convolutional layers are all selected to be 2 multiplied by 2, and the step lengths of the pooling layers are all selected to be 2. Since the present prediction task is to predict the wind speed of the next step from the historical data, the output dimension is set to 1. In order to avoid the over-fitting problem during network training and ensure that the generalization capability of the network is good enough, a Dropout layer with a set value of 0.2 is added after the fully connected layer 1. The whole CNN training process is as follows: generating 8 characteristic graphs by the 7 multiplied by 7 historical wind speed matrix subjected to normalization processing through the convolutional layer 1; then, the maximum pooling layer 1 reduces the size of 8 feature maps from 7 × 7 to 4 × 4, enters the convolutional layer 2 and generates 16 feature maps; then max pooling layer 2 will have 16 featuresThe graph is reduced from 4 x 4 to 2 x 2 and passes through the fully connected layer 1; after being screened by the Dropout layer, the wind speed enters the full-connection layer 2 to obtain a predicted value of the wind speed, and the predicted value is compared with a true value, and the weight of each layer is updated through a random gradient descent method.
Working mode selection at predicted wind speed
According to the method, the reference rotating speed of the fan rotor is calculated according to the wind speed prediction result obtained by the short-term wind speed prediction model, and the working mode of the DFIG is switched according to the reference rotating speed of the fan rotor. Fig. 5 is a schematic diagram of the operation mode switching. The calculation formula of the reference rotating speed of the fan rotor is as follows:
in the formula (I), the compound is shown in the specification,reference speed for the fan rotor, K gear ratio, R turbine radius, λoptFor optimum tip speed ratio, VstimAnd the wind speed is predicted value.
The two DFIG working modes mainly comprise: if the rotating speed of the rotor is lower than the rated rotating speed, adopting a mode one, calculating the corresponding optimal rotating speed of the rotor and taking the pitch angle as zero; if the rotating speed of the rotor is higher than the rated rotating speed, a mode II is adopted, and the wind energy utilization coefficient of the wind driven generator is changed by adjusting the pitch angle, so that the output power is stabilized at the rated power to ensure the running of the fan.
The mode switching system controller is realized by adopting a digital signal processor DSP with the model number of TMS320F2812, and FIG. 6 is a structural diagram for realizing the mode switching system of the invention, and the constant power output of the wind turbine generator set is realized by adjusting the rotating speed and the pitch angle of a rotor. Wind speed estimation value VstimAnd the DI port of the F28125DSP is connected, the fan rotor reference rotating speed is calculated through a fan rotor reference rotating speed calculation formula and is used as an input signal of the mode switching controller. If the rotating speed of the rotor is lower than the rated rotating speed, adopting a mode I to refer the rotating speed of the fan rotor to the reference rotating speedWith the current rotor speed omegarComparing to obtain rotor rotation speed deviation, and using the rotor rotation speed deviation as the input of a PI controller to obtain an electromagnetic torque reference value Tem_cmdReference value T of electromagnetic torqueem_cmdThe output port of the fan is connected with a DO port of the F2812DSP and is used as the input of a variable rotating speed mechanism, so that the rotating speed of a rotor of a wind power system is controlled, and the output power of the fan is maximized; if the rotating speed of the rotor is higher than the rated rotating speed, adopting a mode II to drive the rated rotating speed omega of the fan rotorratedWith the current rotor speed omegarAnd comparing to obtain the rotor rotation speed deviation, using the rotor rotation speed deviation as the input of a PI controller to obtain a pitch angle reference value beta, connecting the pitch angle reference value beta with a DO port of an F2812DSP, and using the pitch angle reference value beta as the input of a pitch change mechanism, thereby controlling the pitch angle of the wind power system and enabling the output power of the fan to be constant.
Case analysis study
The accuracy of the short-term wind speed prediction model based on the VMD-CNN network is verified by adopting data of a certain offshore wind farm in 2014. 52068 pieces of wind speed data are taken from all data in 2014 of the offshore wind farm, the average wind speed, the output power of the wind turbine generator, the rotating speed of the generator and the pitch angle are taken once every 1 minute, the data are divided into 6508 groups according to a time sequence by taking 8 time sequences as one group, and a VMD-CNN prediction model is constructed on a Matlab platform. The set of data is randomly drawn 6308 out of the set of data as the training data set and 200 out of the set of data as the test data set.
The present invention will use the mean absolute error MAE and the root mean square error RMSE and study the prediction accuracy. It is defined as follows:
in the formula, N is the number of predicted values, YiFor the standardized actual observed value at the time,i is a predicted normalized value in the same period, and i is a sequence number.
In summary, the process steps of the present invention are:
1: decomposing historical wind speed data through a VMD technology;
2: preprocessing data aiming at different wind speed data;
3: establishing a VMD-CNN network prediction model;
4: training a VMD-CNN network prediction model by using historical data;
5: obtaining a wind speed prediction result by utilizing the trained network model;
6: and checking the wind speed prediction result and carrying out different control operations on the offshore wind farm according to the corresponding measures of the wind speed prediction result.
The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (9)
1. An offshore wind power control method based on a VMD-CNN network is characterized by comprising the following steps:
s1: decomposing historical wind speed data through a VMD technology;
s2: the wind speed data and the DFIG active power, the generator rotating speed and the pitch angle data at the corresponding moment are combined for data preprocessing;
s3: establishing a CNN network model, and training the CNN network model by using historical data;
s4: obtaining a wind speed prediction result by utilizing the trained network model;
s5: and checking a wind speed prediction result, calculating the reference rotating speed of the fan rotor according to the wind speed prediction value, and performing different control operations on the DFIG according to the reference rotating speed of the fan rotor.
2. The VMD-CNN network based offshore wind power control method of claim 1, wherein the step S1 is implemented by decomposing the historical wind speed data by VMD technology, and the formula is as follows:
in the above formula, K is the modal number of the wind speed sequence to be decomposed, { u (K) }, { ω (K) } respectively corresponds to the K-th modal component and the center frequency after decomposition, δ (t) is a dirac function, and is a convolution operator;
solving the above formula by augmenting Lagrange's expression, which describes the formula as:
in the formula, λ is Lagrange multiplier, and α is secondary penalty factor.
3. The VMD-CNN network based offshore wind power control method of claim 1, wherein the data preprocessing in step S2 includes data normalization and data combination of wind speed data and DFIG active power, generator speed and pitch angle data at corresponding time instants.
4. The VMD-CNN network-based offshore wind power control method according to claim 3, wherein 4 sets of eigen-mode functions obtained by decomposing the wind speed sequence are combined with the DFIG active power, the generator speed and the pitch angle data at the corresponding time and grouped according to the principle that 8 time sequences form one set, the first seven data of each set are extracted and divided into a plurality of 7 x 7 matrixes, and the eighth data of each set is used as the corresponding wind speed result.
5. The VMD-CNN network-based offshore wind power control method of claim 4, wherein each set of data is normalized.
6. The VMD-CNN network-based offshore wind power control method of claim 1, wherein the step S3 includes establishing a 6-layer CNN model and training of the network, the 6-layer CNN model is composed of 2 convolutional layers, 2 pooling layers and 2 fully-connected layers, wherein the convolutional kernel size of 2 convolutional layers is selected to be 2 x 2, and the pooling layer step size is selected to be 2.
7. The VMD-CNN network-based offshore wind power control method of claim 1, wherein the step S5 comprises calculating a wind turbine rotor reference speed and performing different control operations on the DFIG according to the wind turbine rotor reference speed.
8. The VMD-CNN network-based offshore wind power control method according to claim 7, wherein the wind turbine rotor reference speed is calculated by the formula:
9. The VMD-CNN network based offshore wind power control method of claim 8, wherein the control operation is: if the rotating speed of the rotor is lower than the rated rotating speed, adopting a mode one, calculating the corresponding optimal rotating speed of the rotor and taking the pitch angle as zero; if the rotating speed of the rotor is higher than the rated rotating speed, a mode II is adopted, and the wind energy utilization coefficient of the wind driven generator is changed by adjusting the pitch angle, so that the output power is stabilized at the rated power to ensure the running safety of the fan.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011416488.8A CN112418553A (en) | 2020-12-07 | 2020-12-07 | Offshore wind power control method based on VMD-CNN network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011416488.8A CN112418553A (en) | 2020-12-07 | 2020-12-07 | Offshore wind power control method based on VMD-CNN network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112418553A true CN112418553A (en) | 2021-02-26 |
Family
ID=74775186
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011416488.8A Pending CN112418553A (en) | 2020-12-07 | 2020-12-07 | Offshore wind power control method based on VMD-CNN network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112418553A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113392961A (en) * | 2021-08-17 | 2021-09-14 | 中国海洋大学 | Method for extracting mesoscale eddy track stable sequence and predicting cyclic neural network |
CN113482853A (en) * | 2021-08-06 | 2021-10-08 | 贵州大学 | Yaw control method, system, electronic equipment and storage medium |
CN113988394A (en) * | 2021-10-21 | 2022-01-28 | 中国电建集团华东勘测设计研究院有限公司 | Wind power ultra-short-term power prediction method based on gram matrix and convolutional neural network |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140306451A1 (en) * | 2013-04-12 | 2014-10-16 | King Fahd University Of Petroleum And Minerals | Adaptive pitch control system for wind generators |
CN110083940A (en) * | 2019-04-28 | 2019-08-02 | 东华大学 | A kind of short-term wind speed forecasting method based on SSA-HMD-CNNSVM model |
CN110363360A (en) * | 2019-07-24 | 2019-10-22 | 广东工业大学 | A kind of short-term wind power forecast method, device and equipment |
CN110966144A (en) * | 2019-12-04 | 2020-04-07 | 浙江大学 | BLS-based intelligent constant power control method for wind turbine generator |
CN111091236A (en) * | 2019-11-27 | 2020-05-01 | 长春吉电能源科技有限公司 | Multi-classification deep learning short-term wind power prediction method classified according to pitch angles |
CN111608868A (en) * | 2020-05-27 | 2020-09-01 | 上海海事大学 | Maximum power tracking adaptive robust control system and method for wind power generation system |
-
2020
- 2020-12-07 CN CN202011416488.8A patent/CN112418553A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140306451A1 (en) * | 2013-04-12 | 2014-10-16 | King Fahd University Of Petroleum And Minerals | Adaptive pitch control system for wind generators |
CN110083940A (en) * | 2019-04-28 | 2019-08-02 | 东华大学 | A kind of short-term wind speed forecasting method based on SSA-HMD-CNNSVM model |
CN110363360A (en) * | 2019-07-24 | 2019-10-22 | 广东工业大学 | A kind of short-term wind power forecast method, device and equipment |
CN111091236A (en) * | 2019-11-27 | 2020-05-01 | 长春吉电能源科技有限公司 | Multi-classification deep learning short-term wind power prediction method classified according to pitch angles |
CN110966144A (en) * | 2019-12-04 | 2020-04-07 | 浙江大学 | BLS-based intelligent constant power control method for wind turbine generator |
CN111608868A (en) * | 2020-05-27 | 2020-09-01 | 上海海事大学 | Maximum power tracking adaptive robust control system and method for wind power generation system |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113482853A (en) * | 2021-08-06 | 2021-10-08 | 贵州大学 | Yaw control method, system, electronic equipment and storage medium |
CN113482853B (en) * | 2021-08-06 | 2023-02-24 | 贵州大学 | Yaw control method, system, electronic equipment and storage medium |
CN113392961A (en) * | 2021-08-17 | 2021-09-14 | 中国海洋大学 | Method for extracting mesoscale eddy track stable sequence and predicting cyclic neural network |
CN113392961B (en) * | 2021-08-17 | 2021-11-09 | 中国海洋大学 | Method for extracting mesoscale eddy track stable sequence and predicting cyclic neural network |
CN113988394A (en) * | 2021-10-21 | 2022-01-28 | 中国电建集团华东勘测设计研究院有限公司 | Wind power ultra-short-term power prediction method based on gram matrix and convolutional neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112418553A (en) | Offshore wind power control method based on VMD-CNN network | |
Zhang et al. | Wind speed prediction of IPSO-BP neural network based on lorenz disturbance | |
CN109670213B (en) | Multi-machine equivalent modeling method and device for wind power plant | |
CN111709109B (en) | Photovoltaic absorption capacity calculation method and device considering source-load time sequence correlation | |
CN102831475B (en) | Sine normalization method for power forecast model of wind power plant | |
CN111832812A (en) | Wind power short-term prediction method based on deep learning | |
CN109209768A (en) | A kind of constant output control method of large scale wind power machine | |
CN108468622B (en) | Wind turbines blade root load estimation method based on extreme learning machine | |
CN110889535A (en) | Multi-point wind speed prediction method in wind power plant based on convolution cyclic neural network | |
CN109376426B (en) | Wind power grid-connected power scheduling method and device | |
CN110866633A (en) | Micro-grid ultra-short term load prediction method based on SVR support vector regression | |
CN112183841A (en) | Optimized dispatching method of micro-grid containing electric automobile based on simulated annealing algorithm | |
CN110909994A (en) | Small hydropower station power generation amount prediction method based on big data drive | |
CN113555908A (en) | Energy storage optimization configuration method for intelligent power distribution network | |
CN117132132A (en) | Photovoltaic power generation power prediction method based on meteorological data | |
CN109802440B (en) | Offshore wind farm equivalence method, system and device based on wake effect factor | |
CN116865248A (en) | Ultra-short-term wind power prediction method and electronic equipment | |
CN114429233A (en) | Cluster algorithm-based station-level optimal power control method and system | |
Yin et al. | Deep learning-accelerated optimization algorithm for controller parameters optimization of doubly-fed induction generators | |
CN113095596B (en) | Photovoltaic power prediction method based on multi-stage Gate-SA-TCN | |
CN116306236A (en) | Wind power plant dynamic equivalent modeling method and system based on GRU-LSTM-FC combined network | |
Lin et al. | Clustering wind turbines for a large wind farm using spectral clustering approach based on diffusion mapping theory | |
CN115271254A (en) | Short-term wind power prediction method for optimizing extreme learning machine based on gull algorithm | |
Yang et al. | Research on energy regulation and optimal operation strategy of multi-energy ship power station based on improved particle swarm algorithm | |
Xiaosheng et al. | A deep learning approach for wind power prediction based on stacked denoising auto encoders optimized by bat algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |