CN106971032A - Wind speed forecasting method based on nonparametric probability and numerical weather forecast - Google Patents
Wind speed forecasting method based on nonparametric probability and numerical weather forecast Download PDFInfo
- Publication number
- CN106971032A CN106971032A CN201710156943.7A CN201710156943A CN106971032A CN 106971032 A CN106971032 A CN 106971032A CN 201710156943 A CN201710156943 A CN 201710156943A CN 106971032 A CN106971032 A CN 106971032A
- Authority
- CN
- China
- Prior art keywords
- wind speed
- forecasting
- weather forecast
- nonparametric probability
- numerical weather
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
Abstract
The invention discloses a kind of wind speed forecasting method based on nonparametric probability and numerical weather forecast, first, wind speed is predicted using Chebyshev neural network models, then the error of wind speed initial predicted result is estimated by nonparametric probability method, so as to correct the error produced by initial predicted method;Further directed to " correcting " problem occurred in aforementioned process at wind speed mutation flex point by mistake, wind speed phase error corrections model is established with reference to numerical weather forecast method (NWP), to judge in advance at the time of wind speed mutation flex point, so as to introduce forecasting wind speed field by the way that nonparametric probability and numerical weather forecast are combined, the prediction deviation of the initial wind speed value of neutral net can be effectively reduced, precision of prediction of the target wind farm for wind speed is greatly improved.
Description
Technical field
The present invention is based on nonparametric probability is theoretical and numerical weather forecast is theoretical, by building based on non-
The wind speed correction model of parameter Density Estimator, builds wind speed phase with reference to numerical weather forecast method (NWP) on this basis and misses
This model, is finally used for the short-term wind speed prediction of the actual wind power plant in area by poor correction model.
Background technology
Research currently for wind speed forecasting method is broadly divided into physical method and the major class of statistical method two.
Physical method mainly considers some physical quantitys, around weather data (wind speed, wind direction, air pressure etc.), wind power plant
The technical parameter (wheel hub height, penetrating coefficient etc.) of information (contour, roughness, barrier etc.) and Wind turbines.Its purpose
It is to find the wind speed optimal estimation value that wind-powered machine unit hub is highly located, then using model output statistics (Model output
Statistic, MOS) to reduce predicated error, the power curve calculating finally according to wind power plant obtains the output work of wind power plant
Rate.The modeling process of physical model it is relative complex, it is necessary to wind power plant location carry out physical modeling, including wind field landform,
Surface vegetation and roughness, peripheral obstacle etc.;Will also to the hub height of blower fan in itself, power curve, machine driving and
Control strategy etc. is modeled, and the wind speed of future time instance is estimated with this.Its advantage is not need substantial amounts of historical data, only
Need to set up forecast model using real-time data and be predicted, newly-established wind power plant is typically carried out pre- from physical method
Survey.
Statistical method is then on the basis of Probability analysis is carried out to historical data, to meteorological data in following a period of time
(wind speed, wind direction, air pressure etc.) is predicted, so as to be converted to Power Output for Wind Power Field.The simplest statistics mould that foreign countries use
Type is persistence models, and the principle of the forecast model of the wind power assumes that subsequent time wind energy conversion system power output etc.
In the wind energy conversion system power output of last moment, the increase persistence wind power prediction models of yardstick is pre- over time
Survey precision rapid decrease, therefore usually as the forecast model of benchmark to come other pre- for persistence wind power prediction models
The precision for surveying model is compared.Other conventional statistical models also include time series models, artificial neural network (ANN) and
The methods such as SVMs (SVM).However, in above-mentioned statistical prediction methods, the phase error produced by forecasting wind speed remains unchanged
It can not avoid.
The content of the invention
In view of the problem of forecasting wind speed precision and neutral net are present, the present invention proposes a kind of close based on nonparametric kernel
Degree estimation and the forecasting wind speed modification method of numerical weather forecast (Numerical weather prediction, NWP).Utilize
Nonparametric probability method is estimated to the wind speed initial predicted resultant error obtained by Chebyshev neural network predictions
Meter, and the error produced by Chebyshev neural net prediction methods is corrected, this method is in estimation error and without any
Artificial hypothesis link, only from data set off in search error law itself, be considered as a kind of process of error self-correction;With
Afterwards, introduce numerical weather forecast accurately to be corrected to predicting the outcome, effectively increase the precision of prediction of wind speed.How this is utilized
Kind of model be reducing initial forecasting wind speed offset issue the invention solves the problems that key issue.
It is problem to solve above-mentioned technology, it is proposed by the present invention a kind of based on nonparametric probability and Numerical Weather
The wind speed forecasting method of forecast, comprises the following steps:
Step 1: setting data set record is with per hour for the air speed data at interval, for the wind speed of wind power plant
Historical data using Chebyshev neural network models to r days per hour wind speed be predicted, obtain 24 × r forecasting wind speed
The Chebyshev neural network prediction values of point;
Step 2: s is set as from current nearest forecasting wind speed number of days, using from current farthest r-s days per hour
The difference of actual wind speed value and the wind speed per hour of r days predicted by the neutral net that step one is obtained, obtains forecasting wind speed inclined
Difference, obtains the forecasting wind speed biased sequence being made up of the individual wind speed deviations of 24 × (r-s);
Step 3: the forecasting wind speed biased sequence obtained using random run-length testing method to step 2 carries out stationarity inspection
Test, be such as non-stationary series, then circulation execution calculus of finite differences carrys out tranquilization sequence, until the sequence passes through stationary test;
Step 4: using N-W nonparametric probability methods pairEstimated, norm of nonparametric kernel density expression formula
It is as follows:
In formula (1), f () is referred to as kernel function;K is the dimension of forecasting wind speed deviation sample, bjiFor j-th of i-th sample
The smoothing factor of forecasting wind speed deviation variables;Kernel function f () uses standard gaussian kernel function, and dimension k is missed using final prediction
Poor method determination, smoothing factor bjiDetermined by cross-validation method;
Using the forecasting wind speed biased sequence after the tranquilization obtained in step 3, sample estimates X is set upk,i=[y1i,
y2i,…yki]T, i=1,2 ..., n determine nonparametric probability sample dimension k, so as to set up using final predicated error method
Nonparametric probability sample, the number of nonparametric probability sample is n=24 × (r-s)-(k-1);
Step 5: using the nonparametric probability sample obtained in step 4, being determined using cross-validation method each
Nonparametric probability sample Xk,i=[y1i,y2i,…yki]T, i=1,2 ..., n smoothing factor bjiNumerical value, and using public
Formula (1) is modified to forecasting wind speed deviation in s days, so as to obtain the forecasting wind speed estimation of deviation value at 24 × s moment
Step 6: the numerical weather forecast with reference to wind power plant judges the position of wind speed mutation flex point in s days, it is prominent for wind speed
The wind speed value for becoming flex point takes the Chebyshev neural network prediction values that step one is obtained, for non-wind speed mutation flex point
The forecasting wind speed estimation of deviation value that wind speed value is obtained using step 5Carry out wind speed value amendment.
In wind speed forecasting method of the invention based on nonparametric probability and numerical weather forecast, s is not more than 10%
R.
Compared with prior art, the beneficial effects of the invention are as follows:
It is shown experimentally that, the forecasting wind speed based on nonparametric probability and numerical weather forecast is applied in the present invention
Modification method is compared with application Chebyshev neutral nets, nonparametric probability correction model, NWP forecast models, to wind
Its effect of electric field short-term wind speed prediction is more preferable, reduces predicated error.From analysis, NWP precision of prediction is to forecasting wind speed
Correction effect plays vital effect, and NWP can directly cause comprehensive correction method to the error prediction of wind speed mutation flex point
In the failure of some future positions, predicated error increase is even resulted in.
Brief description of the drawings
Fig. 1 is the schematic diagram that nonparametric probability sample is set up in the present invention;
Fig. 2 is the air speed value and reality that Chebyshev neutral nets and the prediction of nonparametric probability correction model are obtained
The comparison diagram of border air speed value;
Fig. 3 is the comparison diagram of wind speed numerical weather forecast value and actual wind speed value;
Fig. 4 is the wind speed mutation flex point comparison diagram of wind speed numerical weather forecast value and actual wind speed value;
Fig. 5 is Chebyshev neural network models, nonparametric probability error correction, nonparametric probability
With the comparison diagram of the air speed value and actual wind speed value of numerical weather forecast comprehensive modification.
Embodiment
Technical solution of the present invention is described in further detail with specific embodiment below in conjunction with the accompanying drawings, described is specific
Only the present invention is explained for embodiment, is not intended to limit the invention.
A kind of wind speed forecasting method based on nonparametric probability and numerical weather forecast proposed by the present invention is set
It is first, wind speed to be predicted using Chebyshev neural network models, then estimated by norm of nonparametric kernel density to count thinking
Meter method is estimated the error of wind speed initial predicted result, so as to correct the error produced by initial predicted method;Enter one
Step is directed to " correcting by mistake " problem occurred in aforementioned process at wind speed mutation flex point, is set up with reference to numerical weather forecast method (NWP)
Wind speed phase error corrections model, to judge in advance at the time of wind speed mutation flex point, so that by the way that norm of nonparametric kernel density is estimated
Meter and numerical weather forecast, which combine, is incorporated into forecasting wind speed field, can effectively reduce the initial wind speed value of neutral net
Prediction deviation, greatly improve precision of prediction of the target wind farm for wind speed.
The of the invention wind speed forecasting method based on nonparametric probability and numerical weather forecast is comprised the following steps that:
Step 1: setting data set record is with per hour for the air speed data at interval, for the wind speed of wind power plant
Historical data using Chebyshev neural network models to r days per hour wind speed be predicted, obtain 24 × r forecasting wind speed
The Chebyshev neural network prediction values of point.For the actual measurement wind in 1 day April in 2011 to 100 days between July 18 of wind power plant
Speed is test data, and first by April 1st, 2011 to May 11, the air speed value of totally 40 days sets up forecasting wind speed
Chebyshev neural network models, choose r=60, using the Chebyshev neural network models of foundation to May 12 to July
18 the wind speed per hour of totally 60 days be predicted, obtain the Chebyshev neural network prediction values of 24 × r forecasting wind speed point.
Step 2: s is set as from current nearest forecasting wind speed number of days, and s is not more than 10% r, and s=3 is taken here, profit
With the r=predicted from the current farthest r-s=57 days value of actual wind speed per hour and the neutral net obtained by step one
The difference of the wind speed per hour of 60 days, obtains forecasting wind speed deviation, obtains the wind speed being made up of the individual wind speed deviations of 24 × (r-s)
Prediction deviation sequence;
Step 3: the forecasting wind speed biased sequence obtained using random run-length testing method to step 2 carries out stationarity inspection
Test, be such as non-stationary series, then circulation execution calculus of finite differences carrys out tranquilization sequence, until the sequence passes through stationary test;
Step 4: using N-W nonparametric probability methods pairEstimated, norm of nonparametric kernel density expression formula
It is as follows:
Wherein, f () is referred to as kernel function;K is the dimension of forecasting wind speed deviation sample, bjiFor i-th of sample, j-th of wind speed
The smoothing factor of prediction deviation variable.Kernel function f () uses standard gaussian kernel function, and dimension k utilizes final predicated error method
It is determined that, smoothing factor bjiDetermined by cross-validation method.
Sample estimates X is set up using the forecasting wind speed biased sequence after the tranquilization obtained in step 3k,i=[y1i,
y2i,…yki]T, i=1,2 ..., n determine nonparametric probability sample dimension k=3 using final predicated error method, so that
Nonparametric probability sample, method for building up are set up as shown in figure 1, the number of nonparametric probability sample is n=24
×(r-s)-(k-1);
Step 5: using the nonparametric probability sample obtained in step 4, being determined using cross-validation method each
Sample Xk,i=[y1i,y2i,…yki]T, i=1,2 ..., n smoothing factor bjiNumerical value so that using formula (1) to July 16
Forecasting wind speed drift correction in s=3 days on the 18th July, so as to obtain the forecasting wind speed estimation of deviation value at 24 × s momentThe characteristics of to illustrate nonparametric probability method amendment initial predicted wind speed, Fig. 2 provides this method wind speed deviation and repaiied
Plus effect and its contrast effect with Chebyshev neural net prediction method results.As shown in Figure 2, norm of nonparametric kernel density is estimated
Meter method to the correction effect of wind speed deviation can in terms of two explanation:First, persistently it is monotonically changed in wind speed and wind speed tends to
The linear change stage, including when wind speed continued smooth rises, steadily decline and wind speed change small, the method is for forecasting wind speed
The correction effect of deviation is more obvious.2-5 hours, the 35- of wind speed continued smooth decline such as the decline of wind speed continued smooth
38 hours and wind speed change small 50-53 hours etc.;Second, in wind speed variation tendency unstable stage, especially wind
Fast curve is mutated the subsequent time of flex point, the such as the 25th, 27,33,37 hours, there is " amendment by mistake " problem, revised prediction
As a result it is very undesirable, overall correction effect is greatly affected, even more so that global error evaluation index is inferior to before amendment.Produce
The root of this problem is the hysteresis quality of the simple common defects, i.e. phase being predicted using numerical value.
Step 6: the numerical weather forecast with reference to wind power plant judges wind speed mutation in July 16 to 18 days, s=3 days July
The position of flex point, the wind speed value for wind speed mutation flex point takes the Chebyshev neural network prediction values that step one is obtained,
The forecasting wind speed estimation of deviation value that wind speed value for non-wind speed mutation flex point is obtained using step 5Carry out wind speed
Predicted value amendment.Numerical weather forecast is not very good for the prediction effect of wind speed, if but not considering wind speed here
The numerical bias of prediction, only focuses on the Phase Prediction effect in wind speed mutation flex point, and numerical weather forecast then has preferably pre-
Survey effect.If Fig. 3 is that in predicted in test sample 72 hours, time point and the number of mutation flex point occurs in actual wind speed
It is worth the time point compares figure that mutation flex point occurs in weather forecast wind speed, if prediction time is wind speed mutation flex point, with the table of numerical value 1
Show, otherwise represented with numerical value 0.Fig. 4 is wind speed mutation flex point compares figure.As seen from Figure 4, actual wind speed occur altogether 27 dash forward
Become flex point, 38 mutation flex points occurs altogether in numerical weather forecast wind speed.Wherein numerical weather forecast success prediction goes out 21 mutation
Flex point, i.e. the subsequent time point in this 20 points, if not using it to carry out wind speed amendment, nonparametric probability method meeting
Undesirable amendment is made, and causes error to increase.Also 7 actual mutation flex points are not predicted out, under this 7 points
One moment, although Nonparametric Estimation is made that undesirable amendment, but correction value is still used.There are 11 numerical value
Weather forecast mutation flex point is error prediction, i.e. the subsequent time point in this 11 points, even if nonparametric probability method
Initial prediction can be effectively corrected, can not be also used, although in this 11 point predictions is wrong, but can't be produced negative
The prediction effect in face.For needing July 16 to 72 hours of July 18 of prediction herein, according to numerical weather forecast
38 wind speed flex points are predicted, this 38 points are not corrected, and directly use initial prediction as final wind speed value,
Remaining 34 points are modified using non-parametric estmation method.
The wind speed deflection forecast value obtained below to above-mentioned example carries out error analysis according to error criterion formula.
In short-term wind speed forecasting, the forecasting wind speed modification method based on nonparametric probability and numerical weather forecast
For Chebyshev neutral nets, nonparametric probability, numerical weather prediction wind speed method, show substantially
Good prediction effect.
Assessment prediction effect is, it is necessary to appropriate evaluation index.Conventional error assessment index has mean absolute error, square
Root error etc..In forecasting wind speed, only using a kind of error assessment index can not comprehensively assessment prediction effect quality.Therefore
Using three kinds of predicated error indexs:1) mean absolute error (Mean absolute error, MAE);2) root-mean-square error
(Root mean square error,RMSE);3) mean absolute percentage error (Mean absolute percentage
Error, MAPE), prediction effect is weighed jointly.Their calculation formula is as follows:
Wherein,With x 'kPrediction of wind speed value and actual wind speed value are represented respectively;P represents the number of prediction of wind speed.
Due to be in practice likely to occur wind speed for 0 or close to 0 situation, therefore when using MAPE, even if very little
Prediction deviation may also cause error criterion close to infinity;And when wind speed is larger, even if larger prediction deviation, error
Index also can very little, so as to cause that precision of prediction can not be objectively responded.To overcome MAPE this limitation, history is introduced maximum
Wind speed x 'max, by the actual wind speed value x ' in MAPEkWith x 'maxInstead of, and revised MAPE is represented with MAPEm, now
MAPEm expression formula is as follows:
According to above-mentioned analysis, using Chebyshev neutral nets, nonparametric probability method, norm of nonparametric kernel density
Estimation is combined method (forecasting wind speed modification method i.e. of the present invention) with numerical weather forecast, to July 16 to wind speed on July 18
It is predicted and corrects, final effect is as shown in Figure 5.The predicated error of wind speed before and after amendment is calculated according to formula (2)-(5),
As a result it is as shown in table 1.
Table 1 corrects front and rear forecasting wind speed error
By Fig. 5 and table 1 as can be seen that can be predicted using wind speed forecasting method proposed by the present invention at quite a few
Point is modified to predicting the outcome for neutral net, and three class error criterions all make moderate progress.In this method, numerical weather forecast
Play an important role, in final result, a part of wind speed mutation flex point does not have due to the inaccurate of numerical weather forecast
It is judged out, causes undesirable amendment to be used by mistake;Some not mutated flex point is missed by numerical weather forecast
Sentence, cause preferable amendment not to be utilized, these all have impact on final correction effect.How the turning of numerical weather forecast is improved
Point prediction ability, then be the emphasis studied from now on.
Although above in conjunction with accompanying drawing, invention has been described, and the invention is not limited in above-mentioned specific implementation
Mode, above-mentioned embodiment is only schematical, rather than restricted, and one of ordinary skill in the art is at this
Under the enlightenment of invention, without deviating from the spirit of the invention, many variations can also be made, these belong to the present invention's
Within protection.
Claims (2)
1. a kind of wind speed forecasting method based on nonparametric probability and numerical weather forecast, comprises the following steps:
Step 1: setting data set record is with per hour for the air speed data at interval, for the wind speed history of wind power plant
Data using Chebyshev neural network models to r days per hour wind speed be predicted, obtain 24 × r forecasting wind speed point
Chebyshev neural network prediction values;
Step 2: setting s from current nearest forecasting wind speed number of days, to utilize the reality per hour from currently farthest r-s days
The difference of air speed value and the wind speed per hour of r days predicted by the neutral net that step one is obtained, obtains forecasting wind speed deviation,
Obtain the forecasting wind speed biased sequence being made up of the individual wind speed deviations of 24 × (r-s);
Step 3: the forecasting wind speed biased sequence obtained using random run-length testing method to step 2 carries out stationary test,
Such as it is non-stationary series, then circulation execution calculus of finite differences carrys out tranquilization sequence, until the sequence passes through stationary test;
Step 4: using N-W nonparametric probability methods pairEstimated, norm of nonparametric kernel density expression formula is such as
Under:
In formula (1), f () is referred to as kernel function;K is the dimension of forecasting wind speed deviation sample, bjiFor i-th of sample, j-th of wind speed
The smoothing factor of prediction deviation variable;Kernel function f () uses standard gaussian kernel function, and dimension k utilizes final predicated error method
It is determined that, smoothing factor bjiDetermined by cross-validation method;
Using the forecasting wind speed biased sequence after the tranquilization obtained in step 3, sample estimates X is set upk,i=[y1i,y2i,…
yki]T, i=1,2 ..., n determine nonparametric probability sample dimension k using final predicated error method, so as to set up non-ginseng
Number Density Estimator sample, the number of nonparametric probability sample is n=24 × (r-s)-(k-1);
Step 5: using the nonparametric probability sample obtained in step 4, each non-ginseng is determined using cross-validation method
Number Density Estimator sample i smoothing factor bjiNumerical value, and s days interior forecasting wind speed deviations are modified using formula (1),
So as to obtain the forecasting wind speed estimation of deviation value at 24 × s moment
Step 6: the numerical weather forecast with reference to wind power plant judges the position of wind speed mutation flex point in s days, turned for wind speed mutation
The wind speed value of point takes the Chebyshev neural network prediction values that step one is obtained, for the wind speed of non-wind speed mutation flex point
The forecasting wind speed estimation of deviation value that predicted value is obtained using step 5Carry out wind speed value amendment.
2. the wind speed forecasting method based on nonparametric probability and numerical weather forecast according to claim 1, it is special
Levy and be, s is not more than 10% r.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710156943.7A CN106971032A (en) | 2017-03-16 | 2017-03-16 | Wind speed forecasting method based on nonparametric probability and numerical weather forecast |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710156943.7A CN106971032A (en) | 2017-03-16 | 2017-03-16 | Wind speed forecasting method based on nonparametric probability and numerical weather forecast |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106971032A true CN106971032A (en) | 2017-07-21 |
Family
ID=59329671
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710156943.7A Pending CN106971032A (en) | 2017-03-16 | 2017-03-16 | Wind speed forecasting method based on nonparametric probability and numerical weather forecast |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106971032A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108009689A (en) * | 2017-12-22 | 2018-05-08 | 北京金风慧能技术有限公司 | The training of wind speed correction model, wind speed modification method, device and wind power generating set |
CN111917568A (en) * | 2020-05-19 | 2020-11-10 | 宁波大学 | Method for detecting mis-configuration of mobile internet network element equipment |
CN112288164A (en) * | 2020-10-29 | 2021-01-29 | 四川大学 | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast |
CN115293393A (en) * | 2022-04-11 | 2022-11-04 | 北京城市气象研究院 | Near-ground wind speed prediction method combining turbulence physical model and historical data optimization |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104485665A (en) * | 2014-12-17 | 2015-04-01 | 河海大学 | Dynamic probabilistic power flow (PPF) calculating method considering wind speed predication error temporal-spatial coherence |
CN106407627A (en) * | 2016-11-23 | 2017-02-15 | 西南石油大学 | Wind speed probability distribution modeling method and system |
-
2017
- 2017-03-16 CN CN201710156943.7A patent/CN106971032A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104485665A (en) * | 2014-12-17 | 2015-04-01 | 河海大学 | Dynamic probabilistic power flow (PPF) calculating method considering wind speed predication error temporal-spatial coherence |
CN106407627A (en) * | 2016-11-23 | 2017-02-15 | 西南石油大学 | Wind speed probability distribution modeling method and system |
Non-Patent Citations (3)
Title |
---|
周介圭: "风电场短期风速预测技术方法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
岳捷等: "基于数值天气预报统计降尺度的风电场短期风速预测方法", 《2013电力行业信息化年会论文集》 * |
徐玉琴等: "基于非参数核密度估计的风速短期相依模型", 《太阳能学报》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108009689A (en) * | 2017-12-22 | 2018-05-08 | 北京金风慧能技术有限公司 | The training of wind speed correction model, wind speed modification method, device and wind power generating set |
CN111917568A (en) * | 2020-05-19 | 2020-11-10 | 宁波大学 | Method for detecting mis-configuration of mobile internet network element equipment |
CN111917568B (en) * | 2020-05-19 | 2022-10-04 | 宁波大学 | Method for detecting mis-configuration of mobile internet network element equipment |
CN112288164A (en) * | 2020-10-29 | 2021-01-29 | 四川大学 | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast |
CN112288164B (en) * | 2020-10-29 | 2023-04-07 | 四川大学 | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast |
CN115293393A (en) * | 2022-04-11 | 2022-11-04 | 北京城市气象研究院 | Near-ground wind speed prediction method combining turbulence physical model and historical data optimization |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106971032A (en) | Wind speed forecasting method based on nonparametric probability and numerical weather forecast | |
CN110761947B (en) | Yaw calibration method and system for wind turbine generator | |
EP3396571A1 (en) | Method and device for calculating power generation capacity of wind farm | |
US11713747B2 (en) | Wind turbine control method | |
Haque et al. | A hybrid intelligent model for deterministic and quantile regression approach for probabilistic wind power forecasting | |
CN102880810B (en) | Wind power prediction method based on time sequence and neural network method | |
JP5723362B2 (en) | Solar radiation intensity prediction system and photovoltaic power generation output prediction system | |
CN108428017B (en) | Wind power interval prediction method based on nuclear extreme learning machine quantile regression | |
WO2020199875A1 (en) | Wind profile data prediction method | |
CN103246805B (en) | A kind of method of estimation for becoming stoppage in transit probability under disaster caused by a windstorm weather during overhead transmission line | |
CN104933483A (en) | Wind power forecasting method dividing based on weather process | |
CN111079343B (en) | Wind turbine generator effective wind speed estimation method based on width learning | |
CN113361766A (en) | Multi-mode rainfall estimation method integrating machine learning | |
US20230258161A1 (en) | A method and an apparatus for computer-implemented prediction of power production of one or more wind turbines in a wind farm | |
CN108985490A (en) | Wind-driven generator power curve modification method based on the fitting of sliding window characteristic | |
Baheri et al. | Altitude optimization of airborne wind energy systems: A Bayesian optimization approach | |
Barque et al. | Improving wind power prediction with retraining machine learning algorithms | |
CN113205210A (en) | Wind speed and power prediction method, system, equipment and storage medium for wind power plant with complex terrain | |
CN112508437A (en) | Method for correcting time sequence similarity error of wind speed forecast in micro-climate environment | |
Teferra et al. | Fuzzy-based prediction of solar PV and wind power generation for microgrid modeling using particle swarm optimization | |
JP7068531B1 (en) | Information processing equipment, information processing methods, information processing systems and computer programs | |
Potter et al. | Short-term wind forecasting techniques for power generation | |
Guo et al. | A data-driven real-time irrigation control method based on model predictive control | |
CN115860248A (en) | Wind speed numerical assimilation short-term prediction method and system | |
CN111242371B (en) | Photovoltaic power generation short-term prediction correction method based on non-iterative multi-model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170721 |