CN109063915A - Short-term wind speed forecasting method, device, equipment, system and storage medium - Google Patents
Short-term wind speed forecasting method, device, equipment, system and storage medium Download PDFInfo
- Publication number
- CN109063915A CN109063915A CN201810909547.1A CN201810909547A CN109063915A CN 109063915 A CN109063915 A CN 109063915A CN 201810909547 A CN201810909547 A CN 201810909547A CN 109063915 A CN109063915 A CN 109063915A
- Authority
- CN
- China
- Prior art keywords
- wind speed
- particle
- short
- fitness
- neural network
- 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
- 238000013277 forecasting method Methods 0.000 title claims abstract description 30
- 239000002245 particle Substances 0.000 claims abstract description 114
- 238000013528 artificial neural network Methods 0.000 claims abstract description 61
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 35
- 238000005457 optimization Methods 0.000 claims abstract description 35
- 238000000576 coating method Methods 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 35
- 238000012549 training Methods 0.000 claims description 29
- 230000008569 process Effects 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000005856 abnormality Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 2
- 230000006978 adaptation Effects 0.000 claims 2
- 210000005036 nerve Anatomy 0.000 claims 1
- 230000007547 defect Effects 0.000 abstract description 7
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 238000012216 screening Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 8
- 239000008186 active pharmaceutical agent Substances 0.000 description 5
- 238000000354 decomposition reaction Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000005611 electricity Effects 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000009472 formulation Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000004218 nerve net Anatomy 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000001151 other effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000009987 spinning Methods 0.000 description 1
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Primary Health Care (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of short-term wind speed forecasting methods, parameter optimization is carried out to neural network using crossover algorithm in length and breadth, by being iterated longitudinal direction and lateral cross calculating to wind speed particle, filter out the highest particle of fitness, fitness embodies the gap of target output and reality output, fitness is higher, the predictive ability of neural network is stronger, the highest particle of the fitness filtered out can embody the optimal solution of trained wind speed group entirety, screening, which obtains the highest particle of fitness, not only may be implemented to seek optimal solution to whole, avoid the defect of neural network local optimum, and the generalization ability that neural network can be improved is adjusted to the parameter of neural network by the highest Fe coatings of fitness, network parameter is more excellent, the precision of prediction can be greatly improved, and then utilization of the raising to wind energy.The invention also discloses a kind of short-term wind speed forecasting device, equipment, system and a kind of readable storage medium storing program for executing, have above-mentioned beneficial effect.
Description
Technical field
The present invention relates to wind energy field, in particular to a kind of short-term wind speed forecasting method, device, equipment, system and one kind
Readable storage medium storing program for executing.
Background technique
In recent years, it is used due to the exhaustive exploitation to non-renewable resources and excessively, society is anxious to global climate situation
The worry that play deteriorates increasingly increases.Many countries and cities have begun from traditional energy and turn to renewable energy life in the world
It produces.In all renewable energy, wind energy has become growth point of greatest concern and with fastest developing speed in the world.
Since wind-powered electricity generation has the characteristics that randomness, intermittence and fluctuation, cause large-scale wind power integration to electric system
Stability bring and seriously affect, such as increase the operating cost and spinning reserve of electric system, the tremendous growth of wind energy is wind
Traditional power grid can be incorporated to and bring stern challenge.Accurate wind speed or wind power prediction are to raising power system stability operation
Economic benefit is of great significance, and can also improve the ability that wind power plant is participated in market competition.
Forecasting wind speed can be divided into short-term, medium and long term prediction, wherein short-term wind speed forecasting is the electric system containing wind-powered electricity generation
The important evidence of economic load dispatching.Common short-term wind speed forecasting model includes time series models, artificial intelligence model and mixing
The neural network model of the routine such as model.
However, the parameter of current conventional neural network model is easily trapped into local optimum, that is, be easy to be trapped in one it is limited
Seek optimal solution in space, and large-scale optimized parameter can not be sought, the model trained is unstable, prediction result precision
It is lower.
Therefore, the defect of nerve net local optimum how is avoided, the Accurate Prediction to short-term wind speed is realized, is this field skill
Art personnel's technical issues that need to address.
Summary of the invention
The object of the present invention is to provide a kind of short-term wind speed forecasting method, this method is using crossover algorithm in length and breadth to the mind
Parameter optimization is carried out through network, the precision of short-term wind speed forecasting can be can be improved to avoid the defect of neural network local optimum；
It is a further object of the present invention to provide a kind of short-term wind speed forecasting device, equipment, system and a kind of readable storage medium storing program for executing, have upper
State beneficial effect.
The present invention provides a kind of short-term wind speed forecasting method, comprising:
The wind speed historical data of acquisition is pre-processed, wind series are obtained；
The wind series are input to prediction model, obtain prediction result；Wherein, the prediction model is by length and breadth
Crossover algorithm carries out the neural network after parameter optimization；
Carrying out parameter optimization to the neural network by crossover algorithm in length and breadth includes:
Lateral cross and crossed longitudinally, generation are carried out to particle in training wind speed group according to the output of neural network
For particle；
Fitness evaluation is carried out to particle in the trained wind speed group and the filial generation particle, obtains the suitable of each particle
Response；
Filter out the highest particle of fitness；
The parameter of the neural network is set according to the highest Fe coatings of fitness；Wherein, the parameter includes each layer
Weight and threshold value.
Optionally, described be grouped to each particle carries out lateral cross and crossed longitudinally includes:
Each particle is grouped and carries out lateral cross, obtains first child particle；
It is crossed longitudinally to first child particle grouping progress, obtain second filial generation particle；
Lateral cross and it is crossed longitudinally alternately, when the number of iterations reaches default maximum number of iterations, iteration ends,
Obtain all filial generation particles generated；Wherein, the maximum number of iterations is calculated according to the parameter of the trained wind speed group
It arrives.
Optionally, before the air speed data being input to prediction model further include:
It is decomposed based on changeable mode and the wind series is reconstructed according to wind speed frequency, obtain several mode wind speed sequences
Column；
The wind series are then input to prediction model specifically:
Several mode wind series are sequentially input to prediction model, the corresponding prediction of each mode air speed data is obtained
Value；
It is superimposed the corresponding predicted value of each mode air speed data, using the prediction summation of generation as prediction result.
Optionally, the wind speed historical data of described pair of acquisition, which pre-process, includes:
Obtain wind speed historical data；
The Incomplete Point and rejecting abnormalities point in the wind speed historical data are corrected, wind series are obtained.
Optionally, the short-term wind speed forecasting method further include:
Prediction result judge is carried out to the prediction result, obtains prediction error value；
Wherein, the prediction result judge includes: to calculate mean absolute error, calculate standard error and calculate average exhausted
To percentage error.
Optionally, the short-term wind speed forecasting method further include:
When the prediction error value is more than corresponding threshold value, parameter is carried out to current predictive model by crossover algorithm in length and breadth
Optimization.
The present invention discloses a kind of short-term wind speed forecasting device
Pretreatment unit obtains wind series for pre-processing to the wind speed historical data of acquisition；
Predicting unit obtains prediction result for the wind series to be input to prediction model；Wherein, the prediction
Model is that parameter optimization unit carries out the neural network after parameter optimization by crossover algorithm in length and breadth；
Wherein, the parameter optimization unit includes:
Calculated crosswise subelement, for carrying out lateral cross to particle in training wind speed group according to the output of neural network
And it is crossed longitudinally, generate filial generation particle；
Fitness evaluation subelement, for adapting to particle in the trained wind speed group and the filial generation particle
Degree evaluation, obtains the fitness of each particle；
Subelement is screened, for filtering out the highest particle of fitness；
Parameter setting subelement, for the parameter of the neural network to be arranged according to the highest Fe coatings of fitness；Its
In, the parameter includes each layer weight and threshold value.
The present invention discloses a kind of short-term wind speed forecasting equipment, comprising:
Memory, for storing computer program；
Processor, the step of short-term wind speed forecasting method is realized when for executing the computer program.
The present invention discloses a kind of short-term wind speed forecasting system, comprising:
Air speed data acquires equipment and obtains wind speed historical data, and the wind speed is gone through for carrying out real time video collection
History data are sent to short-term wind speed forecasting equipment；
The short-term wind speed forecasting equipment obtains wind series for pre-processing to the wind speed historical data of acquisition；
The wind series are input to prediction model, obtain prediction result；Wherein, the prediction model is to pass through crossover algorithm in length and breadth
Neural network after carrying out parameter optimization；Carrying out parameter optimization to the neural network by crossover algorithm in length and breadth includes: basis
The output of neural network carries out lateral cross and crossed longitudinally, generation filial generation particle to particle in training wind speed group；To institute
It states particle and the filial generation particle in trained wind speed group and carries out fitness evaluation, obtain the fitness of each particle；It filters out
The highest particle of fitness；The parameter of the neural network is set according to the highest Fe coatings of fitness；Wherein, the parameter
Including each layer weight and threshold value.
The present invention discloses a kind of readable storage medium storing program for executing, and program is stored on the readable storage medium storing program for executing, and described program is located
The step of reason device realizes the short-term wind speed forecasting method when executing.
In order to solve the above technical problems, the present invention provides a kind of short-term wind speed forecasting method, using crossover algorithm pair in length and breadth
Neural network carries out parameter optimization, by being iterated longitudinal direction and lateral cross calculating to wind speed particle, filters out fitness
Highest particle, fitness embody the gap of target output and reality output, and fitness is higher, and the predictive ability of neural network is got over
By force, the highest particle of the fitness filtered out can embody the optimal solution of trained wind speed group entirety, and screening obtains fitness most
High particle not only may be implemented to seek optimal solution to whole, avoid the defect of neural network local optimum, and by suitable
The highest Fe coatings of response are adjusted the generalization ability that neural network can be improved, network parameter to the parameter of neural network
It is more excellent, the air speed data of acquisition is predicted by the prediction model that this kind of method training obtains, prediction can be greatly improved
Precision, and then improve utilization to wind energy.
It is disclosed in another embodiment of the present invention and is also based on changeable mode before air speed data is input to prediction model
Wind series are reconstructed according to wind speed frequency for decomposition, obtain several mode wind series；Then wind series are input to pre-
Survey model specifically: several mode wind series are sequentially input to prediction model, it is corresponding pre- to obtain each mode air speed data
Measured value；It is superimposed the corresponding predicted value of each mode air speed data, using the prediction summation of generation as prediction result.Due to wind series
With non-stationary and nonlinear complex characteristics, original wind speed is resolved by a series of moulds by changeable mode decomposition technique
State, then predicted using prediction model, the raising forecasting wind speed precision of the big degree of energy.
The invention also discloses a kind of short-term wind speed forecasting device, equipment, system and a kind of readable storage medium storing program for executing, have upper
Beneficial effect is stated, details are not described herein.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is the flow chart of short-term wind speed forecasting method provided in an embodiment of the present invention；
Fig. 2 is the process stream provided in an embodiment of the present invention for carrying out parameter optimization to neural network by crossover algorithm in length and breadth
Cheng Tu；
Fig. 3 is that prediction result provided in an embodiment of the present invention and actual wind speed compare schematic diagram；
Fig. 4 is the structural block diagram of short-term wind speed forecasting device provided in an embodiment of the present invention；
Fig. 5 is the structural block diagram of short-term wind speed forecasting equipment provided in an embodiment of the present invention；
Fig. 6 is the structural schematic diagram of short-term wind speed forecasting equipment provided in an embodiment of the present invention；
Fig. 7 is the structural block diagram of short-term wind speed forecasting system provided in an embodiment of the present invention.
Specific embodiment
Core of the invention is to provide a kind of short-term wind speed forecasting method, and this method is using crossover algorithm in length and breadth to nerve net
Network carries out parameter optimization, can be improved the precision of short-term wind speed forecasting to avoid the defect of neural network local optimum；This hair
Bright another core is to provide a kind of short-term wind speed forecasting device, system and a kind of readable storage medium storing program for executing, has above-mentioned beneficial to effect
Fruit.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Referring to FIG. 1, Fig. 1 is the flow chart of short-term wind speed forecasting method provided in an embodiment of the present invention；This method can be with
Include:
Step s100, the wind speed historical data of acquisition is pre-processed, obtains wind series.
Pretreated process is carried out specifically without limitation to the wind speed historical data of acquisition, may include to wind speed history number
It, can be according to prediction model according to carrying out Incomplete Point amendment, abnormity point elimination, in chronological order cut, arranged to data
Data processing need voluntarily to select the pretreated type carry out respective handling to the wind speed historical data of acquisition.Wherein, excellent
Selection of land goes through the wind speed of acquisition to avoid prediction model to lead to the reduction of whole precision of prediction to the prediction of invalid data as far as possible
History data carry out pretreatment and are specifically as follows: obtaining wind speed historical data；Correct the Incomplete Point in wind speed historical data and rejecting
Abnormal point obtains wind series.Avoid the influence of a small number of Incomplete Points and abnormal point to overall performance.
Step s200, wind series are input to prediction model, obtain prediction result.
Prediction model is the neural network carried out after parameter optimization by crossover algorithm in length and breadth, to the specific net of prediction model
The number of network topographical form and each layer neuron without limitation, can be determined according to given training sample.
After the network topology structure and each layer neuron number that determine prediction model, according to the training sample of acquisition to pre-
It surveys model to be trained, allows to differentiate the hiding attribute in air speed data, air speed data is precisely predicted.
The process flow diagram flow chart of parameter optimization is carried out as shown in Fig. 2, specifically can be with to neural network by crossover algorithm in length and breadth
The following steps are included:
Step s110: lateral cross and longitudinal friendship are carried out to particle in training wind speed group according to the output of neural network
Fork generates filial generation particle.
Lateral cross (lateral cross probability usually takes 1) is the crossover operation that counts in two particles, and two particle is
It is randomly generated with one-dimensional, the filial generation that crossover operation obtains can be stored in matrix MShcThe inside.It is crossed longitudinally be all particles not
Count intersection with one kind for carrying out between dimension, and bidimensional be random combine together, the solution generated after intersection can be stored in
Matrix MSvcIn.
Specifically, lateral cross operation can be carried out according to (formula 1) and (formula 2).
MShc(i, d)=r1×X(i,d)+(1-r1)×X(j,d)+c1× (X (i, d)-X (j, d)) (formula 1)
MShc(j, d)=r2×X(j,d)+(1-r2)×X(i,d)+c2× (X (j, d)-X (i, d)) (formula 2)
Wherein, i, j ∈ N (1, M), d ∈ N (1, D).
In formula, r1、r2It is the random number between [0,1]；c1、c2It is the random number between [- 1,1]；M is the model of population
It encloses；D is the dimension of variable；The d that X (i, d), X (j, d) respectively indicate parent particle X (i) and X (j) is tieed up；MShc(i,d)、MShc
(j, d) respectively indicates X (i, d) and X (j, d) and generates filial generation in d by lateral cross.
Specifically, crossed longitudinally operation can be carried out according to (formula 3).
MSvc(i,d1)=rX (i, d1)+(1-r)·X(i,d2) (formula 3)
Wherein, (1, M) i ∈ N, d1,d2∈ N (1, D), r ∈ [0,1].
MS in formulavc(i,d1) be particle X (i) d1Peacekeeping d2Tie up the filial generation by generating after crossed longitudinally operation.
The crossed longitudinally probability P of crossover algorithm in length and breadthv, the parameters such as population scale M specifically can be according to the training sample of acquisition
This voluntarily determines that specific value is arranged without limitation.
Lateral operation or longitudinal operation are carried out to particle, generating filial generation particle can be regarded as an iteration, at this to repeatedly
Generation number without limitation, can carry out iteration twice, four iteration etc., wherein preferably, can be according to training wind speed group
Parameter is iterated the setting of number, for example, 4 iteration can be set when number of particles is few in training wind speed group, generates
Four filial generation particles；When number of particles is more in training wind speed group, 6 iteration can be set, to seek more preferably particle ginseng
Number.It is only introduced for training wind speed group number of particles herein, other parameters can refer to above-mentioned introduction.
Specifically, each particle is grouped and carries out lateral cross and crossed longitudinally can specifically include following steps:
Step 1: each particle is grouped and carries out lateral cross, obtains first child particle；
Step 2: it is crossed longitudinally to the progress of first child particle, obtain second filial generation particle；
Step 3: lateral cross and it is crossed longitudinally alternately, when the number of iterations reaches default maximum number of iterations, repeatedly
In generation, terminates, and obtains all filial generation particles of generation；Wherein, maximum number of iterations is calculated according to the parameter of training wind speed group
It arrives.
Step s120: fitness evaluation is carried out to particle in training wind speed group and filial generation particle, obtains each particle
Fitness.
Fitness embodies the gap of target output and reality output, and fitness is higher, and the predictive ability of neural network is stronger
It by successive ignition, generates and intersects filial generation particle, Fe coatings are different, carry out fitness to different ions, filter out
The highest particle of fitness can embody the optimal solution of trained wind speed group entirety, and screening obtains the highest particle of fitness not only
It may be implemented to seek optimal solution to whole, avoid the defect of neural network local optimum, and pass through the highest grain of fitness
Subparameter is adjusted the generalization ability that neural network can be improved to the parameter of neural network, and network parameter is more excellent.
The process for calculating particle fitness can refer to the prior art, specifically, for example can use (formula 4) and is adapted to
Degree evaluation.
Fitness
Wherein, ptIndicate the reality output of neural network,Indicate the target output of neural network, N indicates training sample
Number.
Step s130: the highest particle of fitness is filtered out.
Step s140: according to the parameter of the highest Fe coatings setting neural network of fitness；Wherein, parameter includes each layer
Weight and threshold value.
It can refer to the prior art according to the process that Fe coatings carry out the setting of neural network parameter, this will not be repeated here.
Based on the above-mentioned technical proposal, short-term wind speed forecasting method provided by the present embodiment, using crossover algorithm pair in length and breadth
Neural network carries out parameter optimization, by being iterated longitudinal direction and lateral cross calculating to wind speed particle, filters out fitness
Highest particle, fitness embody the gap of target output and reality output, and fitness is higher, and the predictive ability of neural network is got over
By force, the highest particle of the fitness filtered out can embody the optimal solution of trained wind speed group entirety, and screening obtains fitness most
High particle not only may be implemented to seek optimal solution to whole, avoid the defect of neural network local optimum, and by suitable
The highest Fe coatings of response are adjusted the generalization ability that neural network can be improved, network parameter to the parameter of neural network
It is more excellent, the air speed data of acquisition is predicted by the prediction model that this kind of method training obtains, prediction can be greatly improved
Precision, and then improve utilization to wind energy, can be widely applied to the scientific research of electricity market and electric system related fields
And engineer application.
It include different frequencies in the air speed data of acquisition since wind series have non-stationary and nonlinear complex characteristics
The air speed data of rate, for the precision for improving forecasting wind speed, it is preferable that can be decomposed wind series based on changeable mode according to wind
Fast frequency is reconstructed, and obtains several mode wind series；Wind series are then input to prediction model specifically: by several moulds
State wind series are sequentially input to prediction model, obtain the corresponding predicted value of each mode air speed data；It is superimposed each mode wind speed number
According to corresponding predicted value, using the prediction summation of generation as prediction result.
Original wind speed is resolved into a series of mode by changeable mode decomposition technique, then is carried out in advance using prediction model
It surveys, the raising forecasting wind speed precision of the big degree of energy.
To deepen the understanding to short-term wind speed forecasting method provided by the invention, the present embodiment is based on changeable mode with one kind
It is introduced for the short-term wind speed forecasting process of decomposition, mainly may include the training process and the prediction of model of model
Journey.Specifically, comprising the following steps:
S1, it obtains wind speed historical data and data is pre-processed；
In step sl, historical data includes 700 air speed datas.
S2, it decomposes adaptively to resolve into historical wind speed data using changeable mode and a series of possesses specific sparse category
The discrete mode of property.
Specifically, step S2 includes following sub-step:
S21: each mode function u is obtained using Hilbert transform for original input signal f (t)k(t) parsing
Signal, and obtain monotropic frequency spectrumWherein, t indicates t moment, and k indicates k-th of mode, and j indicates imaginary unit,
σ (t) indicates k-th of mode in the centre frequency of t moment.
S22: centre frequency is estimated into the mixing-of the frequency spectrum of each mode and each mode analytic signalOn the basis of
It is modulated to corresponding Base BandWherein wkIndicate the angular frequency of k-th of mode.
S23: by square L of the above demodulated signal gradient2Norm, estimation there emerged a the bandwidth of the signal of mode, controlled
Changeable mode resolution problem is as follows:
Wherein { uk}={ u1..., uK, { wk}={ w1,…,wK}；K=1,2,3 ... K,Expression asks local derviation, f (t) to t
Indicate input signal.
Wherein specific step is as follows for changeable mode resolution problem:
Step S23.1: introducing quadratic penalty function item a and Lagrange multiplier operator λ (t), can be by above formula restricted problem
Unconstrained problem is converted to, extension Lagrangian formulation is formd, such as formula:
Step S23.2: initiation parameterAnd n.
Wherein, { uk}={ u1..., uKIndicate k mode function,Indicate the initial value of this k mode function, { wk}=
{w1,…,wKIndicate k-th of centre frequency,Indicate the initial value of this k centre frequency,It is Lagrange multiplier operator
Initial value, n are the number of iterations.
Step S23.3: the above changeable mode resolution problem is solved using alternately multiplier direction method, by alternately updatingAnd λn+1Seek the saddle point of extension Lagrangian formulation.
Wherein, ukAnd wkRespectively by formulaWithIt carries out
It updates, λ is usedIt is updated.
Step S23.4: for given discrimination precision e > 0, ifThen stop iteration.Obtain one
A component U1。
Step S23.5: step S23.3 and S23.4 are repeated and is achieved with remaining component U2、U3、...、Un。
S3, selection training sample establish the prediction model that crossover algorithm in length and breadth optimizes the robust extreme learning machine that peels off；
S4, all subsequences are all made of with the prediction model progress that crossover algorithm in length and breadth optimizes the robust extreme learning machine that peels off
Single-step Prediction:
In step s 4, select training sample for preceding 600 historical wind speed datas.Crossover algorithm optimization in length and breadth is established to peel off
The prediction model of robust extreme learning machine:
The given training sample of S41, basis, determines the neuron number of neural network topology structure and each layer, and determine
The crossed longitudinally probability P of crossover algorithm in length and breadthv, population scale M, maximum number of iterations Tmaxgen；
S42, the particle to be optimized is encoded, in the solution space of coding, and initial population X=[X is randomly generated1,
X2,...,XM]T；
S43, fitness evaluation is carried out to particle each in group using following formula:
Wherein, ptIndicate the reality output of neural network,Indicate the target output of neural network, N indicates training sample
Number.
S44, lateral cross operation is carried out according to the following formula, lateral cross (lateral cross probability usually takes 1) is in two grains
Count crossover operation in son, and two particle is randomly generated with one-dimensional, and the filial generation that crossover operation obtains is stored in matrix MShc
The inside, then the adaptive value of all particles in the matrix is calculated, by obtained adaptive value and parent population X (i.e. DSvc, the first generation removes
It compares outside), selects the better particle of fitness and be retained in DShcIn.
MShc(i, d)=r1×X(i,d)+(1-r1)×X(j,d)+c1×(X(i,d)-X(j,d))
MShc(j, d)=r2×X(j,d)+(1-r2)×X(i,d)+c2×(X(j,d)-X(i,d))
I, j ∈ N (1, M), d ∈ N (1, D)
In formula, r1、r2It is the random number between [0,1]；c1、c2It is the random number between [- 1,1]；M is the model of population
It encloses；D is the dimension of variable；The d that X (i, d), X (j, d) respectively indicate parent particle X (i) and X (j) is tieed up；MShc(i,d)、MShc
(j, d) respectively indicates X (i, d) and X (j, d) and generates filial generation in d by lateral cross.
S45, crossed longitudinally operation is carried out according to the following formula, crossed longitudinally is a kind of calculation carried out between all particle different dimensionals
Number intersect, and bidimensional be random combine together, the solution generated after intersection is stored in matrix MSvcIn, then calculate MSvcOften
The adaptive value of a particle, with its parent population X (i.e. DShc) be compared, select more excellent particle to be retained in DSvcIn.
MSvc(i,d1)=rX (i, d1)+(1-r)·X(i,d2)
I ∈ N (1, M), d1,d2∈ N (1, D), r ∈ [0,1]
MS in formulavc(i,d1) be particle X (i) d1Peacekeeping d2Tie up the filial generation by generating after crossed longitudinally operation.
S46, judge whether current iteration number k is greater than Tmaxgen, if not, going to step S44 iteration again.If
It is to terminate optimizing, then iteration ends, and by DSvcOne group of best solution of middle fitness is set as weight corresponding to neural network
And threshold value, it obtains carrying out the prediction model after parameter optimization, the prediction of practical short-term wind speed can be carried out based on the prediction model.
The above process is a kind of Model Parameter Optimization process provided in this embodiment, will be utilized below excellent in above-described embodiment
Prediction model after change carries out actual wind speed prediction.It can specifically include following steps:
1, the wind speed historical data of acquisition predict is pre-processed, obtains wind series.
2, it is decomposed using changeable mode and wind series is decomposed into a series of useful specific sparse attributes (wind speed frequency)
Subsequence.
3, the prediction model after each subsequence to be sequentially input to the carry out parameter optimization exported to step S46, it is pre- to obtain this
Survey the predicted value of the subsequence of model output.
4, it is superimposed the predicted value of all subsequences, obtains actual prediction result.
Furthermore it is preferred that further prediction result judge can also be carried out to prediction result after obtaining prediction result, obtain
To prediction error value.It specifically, without limitation to the specific method of prediction result judge, may include: to calculate average absolute to miss
Difference calculates standard error and calculates mean absolute percentage error.For example, predicted value and corresponding actual wind speed value are obtained,
Using the standard error being calculated as prediction error value, to be judged according to prediction error value estimated performance.
It is as follows for based on crossover algorithm in length and breadth prediction model (CSO-ORELM), decomposed based on changeable mode
The prediction of crossing prediction model (VMD-CSO-ORELM) and the existing robust extreme learning machine model (ORELM) that peels off in length and breadth
Effect is judged.By calculating mean absolute error (MAE) respectively to above-mentioned three kinds of models, standard error (RMSE) and
MAPE (absolute percent error) carries out error and compares analysis, and error comparison is as shown in table 1.
Table 1
MAE, RMSE and MAPE are got over hour, and prediction effect is better.By checkout result in upper table as it can be seen that CSO-ORELM mould
Type compares ORELM prediction model, and precision of prediction is improved, and illustrates the short-term wind provided by the invention based on crossover algorithm in length and breadth
Fast prediction technique can greatly promote precision of prediction.
And in three kinds of models, highest precision of prediction is VMD-CSO-ORELM model, is illustrated based on crossover algorithm in length and breadth
On the basis of Optimized model parameter, decomposes to resolve into historical wind speed data using changeable mode and a series of possess specific sparse category
Property discrete mode carry out prediction can be with Optimization Prediction effect.
The present embodiment only judges prediction effect for carrying out MAE, RMSE and MAPE and calculating, and certainly, may be used also
With otherwise, for example building prediction result and actual wind speed compare schematic diagram, as shown in Figure 3 etc., other effects are judged
Details are not described herein for mode.
When prediction error value is small, illustrate that precision of prediction is higher；If predict that error is larger, illustrate structure in prediction model
Or parameter may also not be very perfect.To ensure estimated performance, it is preferable that can be more than corresponding threshold value in prediction error value
When, parameter optimization is carried out to current predictive model by crossover algorithm in length and breadth.
Short-term wind speed forecasting device provided by the invention is introduced below, referring to FIG. 4, Fig. 4 is that the present invention is implemented
The structural block diagram for the short-term wind speed forecasting device that example provides；The apparatus may include: pretreatment unit 400 and predicting unit
401。
Wherein, pretreatment unit 400 is mainly used for pre-processing the wind speed historical data of acquisition, obtains wind speed sequence
Column；
Predicting unit 401 is mainly used for wind series being input to prediction model, obtains prediction result；Wherein, mould is predicted
Type is that parameter optimization unit carries out the neural network after parameter optimization by crossover algorithm in length and breadth；
Wherein, parameter optimization unit specifically includes that
Calculated crosswise subelement is mainly used for carrying out laterally particle in training wind speed group according to the output of neural network
Intersect and crossed longitudinally, generates filial generation particle；
Fitness evaluation subelement is mainly used for commenting particle in training wind speed group and filial generation particle progress fitness
Valence obtains the fitness of each particle；
Subelement is screened, is mainly used for filtering out the highest particle of fitness；
Parameter setting subelement is mainly used for the parameter according to the highest Fe coatings setting neural network of fitness；Its
In, parameter includes each layer weight and threshold value.
Preferably, calculated crosswise subelement can specifically include:
Lateral cross subelement is mainly used for being grouped each particle progress lateral cross, obtains first child particle；
Crossed longitudinally subelement, be mainly used for first child particle be grouped carry out it is crossed longitudinally, obtain second filial generation grain
Son；
Particle obtain subelement, be mainly used for lateral cross and it is crossed longitudinally alternately, when the number of iterations reaches default
When maximum number of iterations, iteration ends obtain all filial generation particles of generation；Wherein, maximum number of iterations is according to training wind speed
The parameter of group is calculated.
Preferably, short-term wind speed forecasting device provided in this embodiment may further include: Mode Decomposition unit, mainly
Wind series are reconstructed according to wind speed frequency for being decomposed based on changeable mode, obtain several mode wind series.
Then predicting unit is specifically used for:
Several mode wind series are sequentially input to prediction model, the corresponding predicted value of each mode air speed data is obtained；
It is superimposed the corresponding predicted value of each mode air speed data, using the prediction summation of generation as prediction result.
Preferably, pretreatment unit may further include:
Data acquisition subelement is mainly used for obtaining wind speed historical data；
Revise subelemen obtains wind series for correcting the Incomplete Point in wind speed historical data and rejecting abnormalities point.
Preferably, short-term wind speed forecasting device provided in this embodiment may further include: unit be judged, for pre-
It surveys result and carries out prediction result judge, obtain prediction error value；Wherein, prediction result judge include: calculate mean absolute error,
It calculates standard error and calculates mean absolute percentage error.
Preferably, short-term wind speed forecasting device provided in this embodiment may further include: when prediction error value is more than
When corresponding threshold value, parameter optimization is carried out to current predictive model by crossover algorithm in length and breadth.
It should be noted that each unit in short-term wind speed forecasting device in the specific embodiment of the invention, work
The corresponding specific embodiment of short-term wind speed forecasting method is please referred to as process, details are not described herein.
Short-term wind speed forecasting equipment provided by the invention is introduced below, specifically to Jie of short-term wind speed forecasting equipment
Continue the step of can refer to above-mentioned short-term wind speed forecasting method, and Fig. 5 is short-term wind speed forecasting equipment provided in an embodiment of the present invention
Structural block diagram；The equipment may include:
Memory 500, for storing computer program；
Processor 501, when for executing computer program the step of realization short-term wind speed forecasting method.
Referring to FIG. 6, the structural schematic diagram of short-term wind speed forecasting equipment provided in an embodiment of the present invention, which can
Bigger difference is generated because configuration or performance are different, may include one or more processors (central
Processing units, CPU) 322 (for example, one or more processors) and memory 332, one or more
Store the storage medium 330 (such as one or more mass memory units) of application program 342 or data 344.Wherein, it deposits
Reservoir 332 and storage medium 330 can be of short duration storage or persistent storage.The program for being stored in storage medium 330 may include
One or more modules (diagram does not mark), each module may include to the series of instructions behaviour in data processing equipment
Make.Further, central processing unit 322 can be set to communicate with storage medium 330, executes and deposits on pre- measurement equipment 301
Series of instructions operation in storage media 330.
Pre- measurement equipment 301 can also include one or more power supplys 326, one or more wired or wireless nets
Network interface 350, one or more input/output interfaces 358, and/or, one or more operating systems 341, such as
Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc..
Step in short-term wind speed forecasting method described in above figure 1 can be real by the structure of short-term wind speed forecasting equipment
It is existing.
Short-term wind speed forecasting system provided in an embodiment of the present invention is introduced below, short-term wind speed described below is pre-
Examining system can correspond to each other reference with above-described short-term wind speed forecasting equipment.
Fig. 7 is the structural block diagram of short-term wind speed forecasting system provided in an embodiment of the present invention；The system may include: wind speed
Data acquisition equipment 700 and short-term wind speed forecasting equipment 701.
Air speed data acquisition equipment 700 is mainly used for carrying out real time video collection, obtains wind speed historical data, and by wind speed
Historical data is sent to short-term wind speed forecasting equipment；
Short-term wind speed forecasting equipment 701 is mainly used for pre-processing the wind speed historical data of acquisition, obtains wind speed sequence
Column；Wind series are input to prediction model, obtain prediction result；Wherein, prediction model is to be carried out by crossover algorithm in length and breadth
Neural network after parameter optimization；Carrying out parameter optimization to neural network by crossover algorithm in length and breadth includes: according to neural network
Output lateral cross and crossed longitudinally is carried out to particle in training wind speed group, generate filial generation particle；To training wind speed group
Particle and filial generation particle carry out fitness evaluation in body, obtain the fitness of each particle；Filter out the highest particle of fitness；
According to the parameter of the highest Fe coatings setting neural network of fitness；Wherein, parameter includes each layer weight and threshold value.
Readable storage medium storing program for executing provided in an embodiment of the present invention is introduced below, readable storage medium storing program for executing described below with
Above-described short-term wind speed forecasting method can correspond to each other reference.
A kind of readable storage medium storing program for executing disclosed by the invention, is stored thereon with program, realizes when program is executed by processor short
The step of phase wind speed forecasting method.
It is apparent to those skilled in the art that for convenience and simplicity of description, the device of foregoing description,
The specific work process of equipment, storage medium and unit, can refer to corresponding processes in the foregoing method embodiment, herein no longer
It repeats.
In several embodiments provided by the present invention, it should be understood that disclosed device, system, storage medium and
Method may be implemented in other ways.For example, apparatus embodiments described above are merely indicative, for example, single
Member division, only a kind of logical function partition, there may be another division manner in actual implementation, for example, multiple units or
Component can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point is shown
The mutual coupling, direct-coupling or communication connection shown or discussed can be through some interfaces, between device or unit
Coupling or communication connection are connect, can be electrical property, mechanical or other forms.
Unit may or may not be physically separated as illustrated by the separation member, shown as a unit
Component may or may not be physical unit, it can and it is in one place, or may be distributed over multiple networks
On unit.It can some or all of the units may be selected to achieve the purpose of the solution of this embodiment according to the actual needs.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
It, can if integrated unit is realized in the form of SFU software functional unit and when sold or used as an independent product
To be stored in a mobile terminal.Based on this understanding, technical solution of the present invention is substantially in other words to the prior art
The all or part of the part to contribute or the technical solution can be embodied in the form of software products, which deposits
It stores up in one storage medium, including some instructions are used so that a mobile terminal (can be mobile phone or tablet computer
Deng) execute all or part of the steps of each embodiment method of the present invention.And storage medium above-mentioned includes: USB flash disk, moves firmly
Disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM),
The various media that can store program code such as magnetic or disk.
Each embodiment is described in a progressive manner in specification, the highlights of each of the examples are with other realities
The difference of example is applied, the same or similar parts in each embodiment may refer to each other.For device disclosed in embodiment
Speech, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is referring to method part illustration
?.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, it can be realized with the combination of electronic hardware, terminal or the two, in order to clearly demonstrate hardware and software
Interchangeability generally describes each exemplary composition and step according to function in the above description.These functions are studied carefully
Unexpectedly it is implemented in hardware or software, the specific application and design constraint depending on technical solution.Professional technique people
Member can use different methods to achieve the described function each specific application, but this realization is it is not considered that super
The scope of the present invention out.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Short-term wind speed forecasting method provided by the present invention, device, equipment, system and readable storage medium storing program for executing are carried out above
It is discussed in detail.Used herein a specific example illustrates the principle and implementation of the invention, above embodiments
Explanation be merely used to help understand method and its core concept of the invention.It should be pointed out that for the common of the art
, without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for technical staff, these
Improvement and modification are also fallen within the protection scope of the claims of the present invention.
Claims (10)
1. a kind of short-term wind speed forecasting method characterized by comprising
The wind speed historical data of acquisition is pre-processed, wind series are obtained；
The wind series are input to prediction model, obtain prediction result；Wherein, the prediction model is by intersecting in length and breadth
Algorithm carries out the neural network after parameter optimization；
Carrying out parameter optimization to the neural network by crossover algorithm in length and breadth includes:
Lateral cross and crossed longitudinally, generation filial generation grain are carried out to particle in training wind speed group according to the output of neural network
Son；
Fitness evaluation is carried out to particle in the trained wind speed group and the filial generation particle, obtains the adaptation of each particle
Degree；
Filter out the highest particle of fitness；
The parameter of the neural network is set according to the highest Fe coatings of fitness；Wherein, the parameter includes each layer weight
And threshold value.
2. short-term wind speed forecasting method as described in claim 1, which is characterized in that described be grouped to each particle carries out lateral friendship
Fork and crossed longitudinally include:
Each particle is grouped and carries out lateral cross, obtains first child particle；
It is crossed longitudinally to first child particle grouping progress, obtain second filial generation particle；
Lateral cross and it is crossed longitudinally alternately, when the number of iterations reaches default maximum number of iterations, iteration ends, obtain
All filial generation particles generated；Wherein, the maximum number of iterations is calculated according to the parameter of the trained wind speed group.
3. short-term wind speed forecasting method as described in claim 1, which is characterized in that the air speed data is input to prediction mould
Before type further include:
It is decomposed based on changeable mode and the wind series is reconstructed according to wind speed frequency, obtain several mode wind series；
The wind series are then input to prediction model specifically:
Several mode wind series are sequentially input to prediction model, the corresponding predicted value of each mode air speed data is obtained；
It is superimposed the corresponding predicted value of each mode air speed data, using the prediction summation of generation as prediction result.
4. short-term wind speed forecasting method as described in claim 1, which is characterized in that described pair acquisition wind speed historical data into
Row pre-processes
Obtain wind speed historical data；
The Incomplete Point and rejecting abnormalities point in the wind speed historical data are corrected, wind series are obtained.
5. short-term wind speed forecasting method as described in claim 1, which is characterized in that further include:
Prediction result judge is carried out to the prediction result, obtains prediction error value；
Wherein, the prediction result judge includes: to calculate mean absolute error, calculate standard error and calculate average absolute hundred
Divide ratio error.
6. short-term wind speed forecasting method as claimed in claim 5, which is characterized in that further include:
When the prediction error value is more than corresponding threshold value, it is excellent that parameter is carried out to current predictive model by crossover algorithm in length and breadth
Change.
7. a kind of short-term wind speed forecasting device characterized by comprising
Pretreatment unit obtains wind series for pre-processing to the wind speed historical data of acquisition；
Predicting unit obtains prediction result for the wind series to be input to prediction model；Wherein, the prediction model
The neural network after parameter optimization is carried out by crossover algorithm in length and breadth for parameter optimization unit；
Wherein, the parameter optimization unit includes:
Calculated crosswise subelement, for according to the output of neural network to particle in training wind speed group carry out lateral cross and
It is crossed longitudinally, generate filial generation particle；
Fitness evaluation subelement is commented for carrying out fitness to particle in the trained wind speed group and the filial generation particle
Valence obtains the fitness of each particle；
Subelement is screened, for filtering out the highest particle of fitness；
Parameter setting subelement, for the parameter of the neural network to be arranged according to the highest Fe coatings of fitness；Wherein, institute
Stating parameter includes each layer weight and threshold value.
8. a kind of short-term wind speed forecasting equipment characterized by comprising
Memory, for storing computer program；
Processor realizes the short-term wind speed forecasting side as described in any one of claim 1 to 6 when for executing the computer program
The step of method.
9. a kind of short-term wind speed forecasting system characterized by comprising
Air speed data acquires equipment, for carrying out real time video collection, obtains wind speed historical data, and by the wind speed history number
According to being sent to short-term wind speed forecasting equipment；
The short-term wind speed forecasting equipment obtains wind series for pre-processing to the wind speed historical data of acquisition；By institute
It states wind series and is input to prediction model, obtain prediction result；Wherein, the prediction model is to be carried out by crossover algorithm in length and breadth
Neural network after parameter optimization；Carrying out parameter optimization to the neural network by crossover algorithm in length and breadth includes: according to nerve
The output of network carries out lateral cross and crossed longitudinally, generation filial generation particle to particle in training wind speed group；To the instruction
Practice particle and the filial generation particle in wind speed group and carry out fitness evaluation, obtains the fitness of each particle；Filter out adaptation
Spend highest particle；The parameter of the neural network is set according to the highest Fe coatings of fitness；Wherein, the parameter includes
Each layer weight and threshold value.
10. a kind of readable storage medium storing program for executing, which is characterized in that be stored with program on the readable storage medium storing program for executing, described program is located
It manages and is realized when device executes as described in any one of claim 1 to 6 the step of short-term wind speed forecasting method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810909547.1A CN109063915A (en) | 2018-08-10 | 2018-08-10 | Short-term wind speed forecasting method, device, equipment, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810909547.1A CN109063915A (en) | 2018-08-10 | 2018-08-10 | Short-term wind speed forecasting method, device, equipment, system and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109063915A true CN109063915A (en) | 2018-12-21 |
Family
ID=64683284
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810909547.1A Pending CN109063915A (en) | 2018-08-10 | 2018-08-10 | Short-term wind speed forecasting method, device, equipment, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109063915A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109687515A (en) * | 2018-12-28 | 2019-04-26 | 广东工业大学 | A kind of the power generation amount control method and relevant apparatus of wind power plant |
| CN111697621A (en) * | 2020-06-23 | 2020-09-22 | 石家庄铁道大学 | Short-term wind power prediction method based on EWT-PDBN combination |
| CN113689030A (en) * | 2021-08-03 | 2021-11-23 | 广东工业大学 | Short-term wind power prediction method based on bidirectional attention and secondary optimization |
| CN114169252A (en) * | 2021-12-27 | 2022-03-11 | 广东工业大学 | Short-term region wind power prediction method for dynamically selecting representative wind power plant |
| CN116108989A (en) * | 2023-01-13 | 2023-05-12 | 华润电力技术研究院有限公司 | Wind power ultra-short-term power prediction method, system, storage medium and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0949884A (en) * | 1995-08-04 | 1997-02-18 | Aasu Uezaa:Kk | Predicting method for local weather using neural network |
| CN103020743A (en) * | 2012-12-27 | 2013-04-03 | 中国科学院电工研究所 | Ultra-short-term wind speed forecasting method for wind power plant |
| CN104408529A (en) * | 2014-11-21 | 2015-03-11 | 广东工业大学 | Short-term load predicting method of power grid |
| CN106485365A (en) * | 2016-10-25 | 2017-03-08 | 广东工业大学 | A kind of Load Prediction In Power Systems method and device |
| CN107392397A (en) * | 2017-08-25 | 2017-11-24 | 广东工业大学 | A kind of short-term wind speed forecasting method, apparatus and system |
-
2018
- 2018-08-10 CN CN201810909547.1A patent/CN109063915A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0949884A (en) * | 1995-08-04 | 1997-02-18 | Aasu Uezaa:Kk | Predicting method for local weather using neural network |
| CN103020743A (en) * | 2012-12-27 | 2013-04-03 | 中国科学院电工研究所 | Ultra-short-term wind speed forecasting method for wind power plant |
| CN104408529A (en) * | 2014-11-21 | 2015-03-11 | 广东工业大学 | Short-term load predicting method of power grid |
| CN106485365A (en) * | 2016-10-25 | 2017-03-08 | 广东工业大学 | A kind of Load Prediction In Power Systems method and device |
| CN107392397A (en) * | 2017-08-25 | 2017-11-24 | 广东工业大学 | A kind of short-term wind speed forecasting method, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| 中国矿业大学（北京）: "《中央高校基本科研业务费项目研究成果论文集2016下》", 31 January 2017, 煤炭工业出版社 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109687515A (en) * | 2018-12-28 | 2019-04-26 | 广东工业大学 | A kind of the power generation amount control method and relevant apparatus of wind power plant |
| CN111697621A (en) * | 2020-06-23 | 2020-09-22 | 石家庄铁道大学 | Short-term wind power prediction method based on EWT-PDBN combination |
| CN111697621B (en) * | 2020-06-23 | 2021-08-24 | 石家庄铁道大学 | Short-term wind power prediction method based on EWT-PDBN combination |
| CN113689030A (en) * | 2021-08-03 | 2021-11-23 | 广东工业大学 | Short-term wind power prediction method based on bidirectional attention and secondary optimization |
| CN113689030B (en) * | 2021-08-03 | 2022-05-27 | 广东工业大学 | Short-term wind power prediction method based on bidirectional attention and secondary optimization |
| CN114169252A (en) * | 2021-12-27 | 2022-03-11 | 广东工业大学 | Short-term region wind power prediction method for dynamically selecting representative wind power plant |
| CN114169252B (en) * | 2021-12-27 | 2022-11-29 | 广东工业大学 | Short-term region wind power prediction method for dynamically selecting representative wind power plant |
| CN116108989A (en) * | 2023-01-13 | 2023-05-12 | 华润电力技术研究院有限公司 | Wind power ultra-short-term power prediction method, system, storage medium and device |
| CN116108989B (en) * | 2023-01-13 | 2024-02-02 | 华润电力技术研究院有限公司 | Wind power ultra-short-term power prediction method, system, storage medium and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109063915A (en) | Short-term wind speed forecasting method, device, equipment, system and storage medium | |
| CN106357456B (en) | A kind of prediction technique and device of network flow | |
| CN108764471A (en) | The neural network cross-layer pruning method of feature based redundancy analysis | |
| CN109344921A (en) | A kind of image-recognizing method based on deep neural network model, device and equipment | |
| CN110334580A (en) | The equipment fault classification method of changeable weight combination based on integrated increment | |
| CN106610867B (en) | On-chip network task scheduling method and device | |
| CN109635946A (en) | A kind of combined depth neural network and the clustering method constrained in pairs | |
| CN109614975A (en) | A kind of figure embedding grammar, device and storage medium | |
| CN109344969B (en) | Neural network system, training method thereof, and computer-readable medium | |
| CN113312874B (en) | Overall wiring method based on improved deep reinforcement learning | |
| CN110442143A (en) | A kind of unmanned plane situation data clustering method based on combination multiple target dove group's optimization | |
| CN112016691A (en) | Construction method and device of quantum line | |
| CN116316591A (en) | Short-term photovoltaic power prediction method and system based on hybrid bidirectional gating cycle | |
| CN110826692B (en) | Automatic model compression method, device, equipment and storage medium | |
| CN114358216B (en) | Quantum clustering method based on machine learning framework and related device | |
| CN112800115A (en) | Data processing method and data processing device | |
| CN108960511B (en) | Wind speed prediction method and related device | |
| Samplawski et al. | Towards objection detection under iot resource constraints: Combining partitioning, slicing and compression | |
| CN116306780B (en) | Dynamic graph link generation method | |
| Yang et al. | Host load prediction based on PSR and EA-GMDH for cloud computing system | |
| CN106503793A (en) | A kind of neural network short-term wind speed forecasting method based on improvement difference algorithm | |
| CN116342420A (en) | Method and system for enhancing mixed degraded image | |
| CN116403397A (en) | Traffic prediction method based on deep learning | |
| CN108898227A (en) | Learning rate calculation method and device, disaggregated model calculation method and device | |
| CN115936926A (en) | SMOTE-GBDT-based unbalanced electricity stealing data classification method and device, computer equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181221 |
|
| RJ01 | Rejection of invention patent application after publication |