CN102236795A - Method for forecasting wind speed in wind power station - Google Patents
Method for forecasting wind speed in wind power station Download PDFInfo
- Publication number
- CN102236795A CN102236795A CN 201110180420 CN201110180420A CN102236795A CN 102236795 A CN102236795 A CN 102236795A CN 201110180420 CN201110180420 CN 201110180420 CN 201110180420 A CN201110180420 A CN 201110180420A CN 102236795 A CN102236795 A CN 102236795A
- Authority
- CN
- China
- Prior art keywords
- wind
- wind speed
- data
- time interval
- predetermined time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for forecasting the wind speed in a wind power station and belongs to the technical field of wind speed forecasting, wherein the method comprises: a step one, reading in future numerical weather prediction data, so as to obtain the wind speed variation trend; a step two, reading in wind measuring history data, so as to obtain similar samples of the wind speed variation trend, and selecting learning samples from the similar samples; a step 3, acquiring the current real-time wind measuring data; and a step 4, forecasting the wind speed according to the selected learning samples and the obtained real-time wind measuring data. The technical problem of poor wind speed forecasting accuracy is solved through the method.
Description
Technical field
The invention belongs to the forecasting wind speed technical field, relate in particular to a kind of Forecasting Methodology of wind farm wind velocity.
Background technology
Wind energy has characteristics such as reserves are huge, renewable, distribution is wide, pollution-free as the important class of regenerative resource, has been subjected to attention more and more widely, and has become novel energy with fastest developing speed.In recent years, the development of global wind generating technology rapidly, Chinese feature Denso machine capacity is since 2005, is doubled growth in continuous 5 years.By the end of the end of the year 2010, the accumulative total installed capacity reaches forty-two million kilowatt, and wind-powered electricity generation installation total volume leaps to the whole world first.Nowadays, the principal mode of Chinese development and utilization wind energy is the wind-power electricity generation that is incorporated into the power networks on a large scale, but since the randomness and the intermittence of the Nature wind speed cause wind-powered electricity generation to be incorporated into the power networks on a large scale electrical network caused very big impact and challenge.If can accurately predict wind farm wind velocity, then help adjusting operation plan, effectively alleviate the adverse effect of wind-powered electricity generation to whole electrical network, reduce Operation of Electric Systems cost and spinning reserve, improve wind-powered electricity generation and penetrate power limit, and help correct formulation electric energy exchange plan under open Power Market, safe operation for electrical network, traffic department adjusts the wind-powered electricity generation operation plan timely and accurately, increases wind-powered electricity generation online ratio, has important practical significance.
Wind speed is subject to factor affecting such as humidity, landform, air pressure, temperature, has very strong randomness, and is bigger to the difficulty of its prediction.At present, wind speed multi-step prediction precision is lower, can not reflect the problem of wind speed future trends.The method of predicting wind speed of wind farm is a lot, and ripe have persistence forecasting method, neural network method, time series analysis method, Kalman filtering method, a wavelet analysis method etc.Conventional forecasting wind speed method is based on a kind of extension Forecasting Methodology of historical wind speed seasonal effect in time series, wherein Chang Yong seasonal effect in time series analytic approach is divided into again: autoregressive model (AR), moving average model (MA), autoregressive moving-average model (ARMA), accumulation formula autoregressive moving-average model (ARIMA), the method of general wind speed multi-step prediction is based on historical data and calculates realization, the multi-step prediction precision is relatively poor, can not reflect the wind speed variation tendency.
Summary of the invention
In order to solve the relatively poor problem of above-mentioned forecasting wind speed precision, the present invention proposes a kind of wind farm wind velocity multistep forecasting method, following numerical weather forecast data that this method is integrated, historical wind data and the real-time wind data of surveying surveyed, on the basis of following the tracks of following wind speed variation tendency, historical wind speed is accurately classified, also for example adopt support vector machine SVM method based on real-time survey wind data, can carry out for example following 4 hours, the 15min time interval, totally 16 the step forecasting wind speeds, can accurately follow the trail of following wind speed variation tendency, can refresh based on the every 15min of actual measurement wind speed again and once forecast the result, the following up-to-date wind speed of 16 steps of the forecast of rolling changes, and has realized following wind speed high precision multi-step prediction.
A kind of wind farm wind velocity multistep forecasting method that the present invention proposes comprises: step 1, read in following numerical weather forecast data, and obtain the wind speed variation tendency; Step 2 is read in and is surveyed the wind historical data, obtains the similar sample of described wind speed variation tendency, and select learning sample from similar sample; Step 3 is obtained current real-time survey wind data; Step 4 according to learning sample of selecting and the real-time survey wind data that obtains, is predicted wind speed.
An aspect according to the proposed method, described following numerical weather forecast data comprise wind speed, wind direction and the temperature data of first predetermined time interval that dopes; Described wind speed variation tendency comprises that wind speed difference and the wind direction between two moment that are divided into first predetermined time interval mutually is poor.
An aspect according to the proposed method, described first predetermined time interval are 1 hour or 15 minutes.
An aspect according to the proposed method, described survey wind historical data comprises historical wind speed, wind direction and the temperature data of second predetermined time interval; Described step 2 specifically comprises: the wind speed difference and the wind direction that calculate between two moment that the distance of surveying the wind historical data is first predetermined time interval are poor, and divide into groups to surveying the wind historical data according to the span of wind speed difference and wind direction difference, every group survey the wind historical data all corresponding scope be X
iWind speed difference and scope be Y
jWind direction poor, i=1 wherein ... m, j=1 ... n is about to survey the wind historical data and is divided into m * n group, forms the wind speed transformation matrices of m * n; Obtain to be positioned at the survey wind historical data of same grouping, form described similar sample with described wind speed variation tendency; From described similar sample, obtain and the many group sample of described wind speed variation tendency difference in predetermined threshold, form described learning sample.
An aspect according to the proposed method, described real-time survey wind data comprises current wind speed and temperature value.
An aspect according to the proposed method, step S4 comprises: described learning sample is learnt, with first constantly wind speed and temperature as input, will with first be separated by the wind speed of second predetermined time interval constantly as output, obtain the relation between the input and output; And according to the relation that is obtained, import described real-time survey wind data, obtain to be divided into mutually next prediction air speed value constantly of second predetermined time interval, and according to the temperature data in two moment that are divided into first predetermined time interval mutually, go out described next temperature value constantly by interpolation calculation, continue to carry out prediction according to described next prediction air speed value and temperature value constantly, to realize the multistep rolling forecast.
An aspect according to the proposed method uses support vector machine SVM algorithm or artificial neural network algorithm that learning sample is learnt, thereby obtains the relation between the input and output.
An aspect according to the proposed method, described second predetermined time interval are less than or equal to the described very first time at interval
Description of drawings
Fig. 1 is the process flow diagram of the method that proposes of the present invention;
Fig. 2 divides synoptic diagram according to the historical data of wind speed difference and wind direction difference;
Fig. 3 is to use support vector machine to obtain the synoptic diagram of input/output relation;
Fig. 4 is the support vector machine method process flow diagram.
Embodiment
Fig. 1 shows the key step of wind farm wind velocity multistep forecasting method proposed by the invention, and as shown in Figure 1, this method comprises: S1, read in following numerical weather forecast data, and obtain the wind speed variation tendency; S2 reads in and surveys the wind historical data, obtains the similar sample of wind speed variation tendency among the S1, and select learning sample from similar sample; S3 obtains current real-time survey wind data; S4 according to the real-time survey wind data that the learning sample of selecting among the S2 and S3 obtain, predicts wind speed.
Below, above-mentioned four steps are specifically described, these specific descriptions only are exemplary, not as the qualification to protection domain of the present invention.
Step S1 reads in following numerical weather forecast data, obtains the wind speed variation tendency.
Concrete, following numerical weather forecast data comprise for example wind speed, wind direction and the temperature data of 70m height (or other height), its time resolution for example is 1h other times resolution such as (or) 15min, promptly, when resolution is 1h, following numerical weather forecast data comprise wind speed, wind direction and the temperature data every prediction in a hour, and when resolution was 15min, following numerical weather forecast data comprised wind speed, wind direction and the temperature data every prediction in 15 minutes.According to provide by the time numerical weather forecast value, it is poor to calculate hour wind speed difference and hour wind direction:
Hour wind speed is poor: Δ V
NWPThe wind speed V in one hour=future
NWP (t+1)-current hour wind speed V
NWP (t)According to Δ V
NWPSize be divided into downtrending, moderate tone and ascendant trend, these three trend correspond respectively to Δ V
NWP＜-0.5 ,-0.5≤Δ V
NWP≤ 0.5, Δ V
NWP＞0.5.The concrete span of subscript " t " can be the moment of any one hour.
Hour wind direction is poor: Δ D
NWPThe wind direction D in one hour=future
NWP (t+1)-current hour wind direction D
NWP (t)If hour wind direction difference Δ D＜-180 (degree), then Δ D=Δ D+360; If Δ D＞180, Δ D=Δ D-360;
So just obtained with fixed time wind speed difference and the wind speed variation tendency represented of wind direction difference at interval.
Step S2 reads in and surveys the wind historical data, obtains the similar sample of wind speed variation tendency among the step S1, and selects learning sample from similar sample.
Concrete, read in wind speed, wind direction, the temperature data of for example 70m height of surveying in the wind historical data (or other height), the historical data temporal resolution can be 15min, that is, survey the wind historical data and can comprise historical wind speed, wind direction and the temperature data of measuring every 15 minutes;
It is poor to calculate hour wind speed difference and hour wind direction of surveying the wind historical data, and press wind speed difference and wind direction difference to surveying the segmentation of wind historical data, for example, can historical data be divided into 3 sections by the wind speed difference, every section becomes 13 sections by the wind direction difference again, constitute one 3 * 13 matrix, as shown in Figure 2, this wind speed transformation matrices further refinement or put thick as 3 * 9 etc. as 5 * 15 etc., here use 3 * 13 only to be exemplary in nature, not as limitation of the present invention.
Concrete segmentation algorithm is:
Wind speed: (1) obtain by the time wind speed, calculate by the time wind speed poor: Δ V=V
T+1-V
t
(2) according to Δ V＜-0.5 ,-0.5≤Δ V≤0.5, Δ V＞0.5 are divided into 3 sections;
Wind direction: (1) obtain by the time wind direction, calculate by the time wind direction poor: Δ D=D
T+1-D
t
(2)ifΔD＜-180，ΔD＝ΔD+360；ifΔD＞180，ΔD＝ΔD-360
(3) according to Δ D＜-55 ,-55≤Δ D＜-45 ,-45≤Δ D＜-35 ,-35≤Δ D＜-25 ,-25≤Δ D＜-15,-15≤Δ D＜-5 ,-5≤Δ D≤5,5＜Δ D≤15,15＜Δ D≤25,25＜Δ D≤35,35＜Δ D≤45,45＜Δ D≤55, Δ D＞55 are divided into 13 sections;
According to Δ V
NWPWith Δ D
NWP(Δ V for example
NWP＜-0.5 ,-15≤Δ D
NWP＜-5), in the historical data of dividing good section, select corresponding section (also select Δ V＜-0.5 ,-corresponding data in this variation range of 15≤Δ D＜-5), this correspondence section in included historical data be the similar sample of historical wind speed variation tendency.Since segmentation to as if survey the wind historical data by the time wind speed difference and by the time wind direction poor, therefore, the time interval of each sample in each section all is 1 hour, and because historical data can comprise historical wind speed, wind direction and the temperature data that recorded every 15 minutes, so each sample all corresponding 4 data that recorded every 15 minutes, for example, if historical data that similar sample is 4:00-5:00, in this sample, comprised 4:00 so, 4:15,4:30 and 4:45 four moment difference corresponding historical data;
Then, it is poor to obtain described similar sample wind speed hourly: Δ V
m=V
T+1-V
t, this wind speed difference calculate before by the time draw during the wind speed difference, calculate Δ V then respectively
mWith Δ V
NWPThe absolute value of difference, and according to the ascending ordering of absolute value, take absolute value and come top many group (generally selecting 20～50 groups) similar samples, and the resolution that obtains to be comprised in this similar sample is the historical data of 15min, wherein, every group of data have comprised wind speed and temperature, form learning sample.
Step S3 obtains current real-time survey wind data, and these data comprise wind speed and temperature value.
Step S4 according to the real-time survey wind data that the learning sample of selecting among the step S2 and step S3 obtain, predicts wind speed.
Concrete, can adopt support vector machine SVM algorithm, at first the learning sample of selecting is learnt (to be input as the one group of wind speed and the temperature of the 15min resolution selected by step S2, be output as the wind speed of next 15min) obtain the relation (as shown in Figure 3) of input and output, import current wind speed that records in real time and temperature value then, predict the wind speed of next 15min, and carry out 4 hours futures, the 15min rolling forecast of totally 16 step wind speed at interval:
At first, adopt step that support vector machine carries out forecasting wind speed as shown in Figure 4:
(1) historical data is carried out normalization and handle, be about to raw data through linear change (general by divided by positive maximal value in this column data or negative minimum value, as to make this column data transform to [1,1] interval) to [1,1] interval, composing training data set.
(2) training data is optimized with different IPs function (linear type, polynomial type, Gauss be fundamental mode and neural type kernel function radially) and different parameters (occurrence of penalty factor C and responsive loss parameter epsilon), generates the training result table of different IPs function and different parameters.
(3) from the training result table, according to training error size (being generally root-mean-square error RMSE), pick out suitable kernel function earlier, select its corresponding optimal parameter (occurrence of penalty factor C and responsive loss parameter epsilon) then.
(4) training dataset is learnt earlier, imported the check that predicts the outcome of one section new data then with the parameter of selection.If dissatisfied to predicated error, returned for (3) step, reselect parameter and learn, if satisfied, carry out next step to predicated error.
(5) the new data set of input carries out forecasting wind speed, carries out error analysis at last.
Rolling forecast is meant every one step of prediction wind speed, and just (temperature of following 15min resolution adopts the interpolation result of data of weather forecast when pursuing, T to the temperature data of trying to achieve with this wind speed predicted and interpolation
n=T
t+ (n-1) * (T
T+1-T
t)/4, n=1,2,3,4,5; T wherein
tBe current hour temperature in the numerical weather forecast, T
T+1Temperature for 1 hour future in the numerical weather forecast), predicts the wind speed of next 15min as input.Every 4 steps of prediction are per 1 hour, and again according to the wind speed variation tendency, once similar sample of screening and learning sample relearn then and predict again.
Certainly, the method for learning to obtain the input and output relation according to learning sample not necessarily adopts support vector machine SVM algorithm (as shown in Figure 3), also can adopt other for example artificial neural network ANN algorithm.
This shows that wind farm wind velocity multistep forecasting method proposed by the invention is integrated to have become following numerical weather forecast data, historically survey wind data and survey the basic data of wind data as forecast in real time; Find out the wind speed variation tendency according to the numerical weather forecast data, adopt the wind speed transformation matrices according to variation tendency, in the historical wind speed data, search similar sample, in similar sample, select most representative data to form learning sample, adopt support vector machine SVM or artificial neural network ANN algorithm, carried out the 15min rolling forecast of totally 16 step wind speed at interval for example following 4 hours.Adopt method of the present invention, can accurately follow the trail of following wind speed variation tendency, can refresh based on for example every 15min of actual measurement wind speed again and once forecast the result, the following up-to-date wind speed of 16 steps of the forecast of rolling changes, and has realized the high precision multi-step prediction of following wind speed.The numerical weather prediction data time resolution that this paper proposed can be the data of 15min or other＜1h; Prediction can be additive methods such as persistence forecasting method, neural network method, time series analysis method, Kalman filtering method, wavelet analysis method, and, already mentioned in front, this wind speed transformation matrices further refinement as 5 * 15, or put thick as 3 * 9 etc., can, wind direction difference division poor according to any wind speed.
Above-mentioned all embodiments that this paper proposed are only for exemplary; only as explanation of the invention and explanation; not as limiting the scope of the invention; those skilled in the art can make according to different actual conditions and changing and adjustment, and these changes and adjustment fall within the scope of protection of the present invention equally.
Claims (8)
1. a predicting wind speed of wind farm method is characterized in that, this method comprises:
Step 1 is read in following numerical weather forecast data, obtains the wind speed variation tendency;
Step 2 is read in and is surveyed the wind historical data, obtains the similar sample of described wind speed variation tendency, and select learning sample from similar sample;
Step 3 is obtained current real-time survey wind data;
Step 4 according to learning sample of selecting and the real-time survey wind data that obtains, is predicted wind speed.
2. according to the method for claim 1, it is characterized in that:
Described following numerical weather forecast data comprise wind speed, wind direction and the temperature data of first predetermined time interval that dopes;
Described wind speed variation tendency comprises that wind speed difference and the wind direction between two moment that are divided into first predetermined time interval mutually is poor.
3. according to the method for claim 2, it is characterized in that:
Described first predetermined time interval is 1 hour or 15 minutes.
4. according to the method for claim 2, it is characterized in that:
Described survey wind historical data comprises historical wind speed, wind direction and the temperature data of second predetermined time interval;
Described step 2 specifically comprises: the wind speed difference and the wind direction that calculate between two moment that the distance of surveying the wind historical data is first predetermined time interval are poor, and divide into groups to surveying the wind historical data according to the span of wind speed difference and wind direction difference, every group survey the wind historical data all corresponding scope be X
iWind speed difference and scope be Y
jWind direction poor, i=1 wherein ... m, j=1 ... n is about to survey the wind historical data and is divided into m * n group, forms the wind speed transformation matrices of m * n;
Obtain to be positioned at the survey wind historical data of same grouping, form described similar sample with described wind speed variation tendency;
From described similar sample, obtain and the many group sample of described wind speed variation tendency difference in predetermined threshold, form described learning sample.
5. method according to claim 4 is characterized in that:
Described real-time survey wind data comprises current wind speed and temperature value.
6. method according to claim 4 is characterized in that:
Step 4 comprises:
Described learning sample is learnt, with first constantly wind speed and temperature as input, will with first be separated by the wind speed of second predetermined time interval constantly as output, obtain the relation between the input and output;
And according to the relation that is obtained, import described real-time survey wind data, obtain to be divided into mutually next prediction air speed value constantly of second predetermined time interval, and according to the temperature data in two moment that are divided into first predetermined time interval mutually, go out described next temperature value constantly by interpolation calculation, continue to carry out prediction according to described next prediction air speed value and temperature value constantly, to realize the multistep rolling forecast.
7. method according to claim 5 is characterized in that:
Use support vector machine SVM algorithm or artificial neural network algorithm that learning sample is learnt, thereby obtain the relation between the input and output.
8. method according to claim 4 is characterized in that:
Described second predetermined time interval is less than or equal to the described very first time at interval.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110180420 CN102236795B (en) | 2011-06-30 | 2011-06-30 | Method for forecasting wind speed in wind power station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110180420 CN102236795B (en) | 2011-06-30 | 2011-06-30 | Method for forecasting wind speed in wind power station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102236795A true CN102236795A (en) | 2011-11-09 |
| CN102236795B CN102236795B (en) | 2013-06-26 |
Family
ID=44887432
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110180420 Active CN102236795B (en) | 2011-06-30 | 2011-06-30 | Method for forecasting wind speed in wind power station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102236795B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102664426A (en) * | 2012-05-10 | 2012-09-12 | 华北电力大学 | Anti-normalization interval correction method for improving air speed prediction precision of SVM (Support Vector Machine) |
| CN102682207A (en) * | 2012-04-28 | 2012-09-19 | 中国科学院电工研究所 | Ultrashort combined predicting method for wind speed of wind power plant |
| 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 |
| CN103793511A (en) * | 2014-02-08 | 2014-05-14 | 中能电力科技开发有限公司 | Method for improving wind speed forecast accuracy |
| CN108152528A (en) * | 2016-12-02 | 2018-06-12 | 北京金风科创风电设备有限公司 | Wind measurement method and device |
| CN108804634A (en) * | 2018-06-01 | 2018-11-13 | 深圳乐信软件技术有限公司 | Data scrolling method, apparatus, headend equipment, background server and medium |
| WO2019006771A1 (en) * | 2017-07-06 | 2019-01-10 | 杨顺伟 | External force interference resistance method and device for unmanned aerial vehicle |
| CN109375292A (en) * | 2018-08-30 | 2019-02-22 | 昆明理工大学 | A kind of prediction of precipitation method based on autoregression integral sliding average and support vector regression |
| CN109685583A (en) * | 2019-01-10 | 2019-04-26 | 博拉网络股份有限公司 | A kind of supply chain needing forecasting method based on big data |
| CN110222899A (en) * | 2019-06-12 | 2019-09-10 | 哈尔滨工业大学 | A kind of cross-country skiing racing track wind speed field prediction technique |
| CN111325376A (en) * | 2018-12-14 | 2020-06-23 | 北京金风科创风电设备有限公司 | Wind speed prediction method and device |
| CN111967653A (en) * | 2020-07-22 | 2020-11-20 | 易天气(北京)科技有限公司 | Method for constructing airport runway wind forecasting model, forecasting method and forecasting system |
| CN112311078A (en) * | 2020-11-04 | 2021-02-02 | 华侨大学 | Solar load adjusting method and device based on information fusion |
| CN114117316A (en) * | 2021-11-18 | 2022-03-01 | 国网山东省电力公司电力科学研究院 | Dynamic wind zone drawing method and system based on multiple data sources |
| CN114662800A (en) * | 2022-05-18 | 2022-06-24 | 浙江浙能能源服务有限公司 | Wind power prediction method and system based on artificial neural network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100080703A1 (en) * | 2008-09-28 | 2010-04-01 | Weiguo Chen | System and method for wind condition estimation |
| CN101793907A (en) * | 2010-02-05 | 2010-08-04 | 浙江大学 | Short-term wind speed forecasting method of wind farm |
| CN102063641A (en) * | 2010-10-14 | 2011-05-18 | 北京大学 | Method for forecasting wind speed of high speed railway line |
-
2011
- 2011-06-30 CN CN 201110180420 patent/CN102236795B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100080703A1 (en) * | 2008-09-28 | 2010-04-01 | Weiguo Chen | System and method for wind condition estimation |
| CN101793907A (en) * | 2010-02-05 | 2010-08-04 | 浙江大学 | Short-term wind speed forecasting method of wind farm |
| CN102063641A (en) * | 2010-10-14 | 2011-05-18 | 北京大学 | Method for forecasting wind speed of high speed railway line |
Non-Patent Citations (4)
| Title |
|---|
| 《中国电机工程学报》 20050615 杨秀媛等 风电场风速和发电功率预测研究 1-5 第25卷, 第11期 * |
| 《太阳能学报》 20110428 彭怀午等 基于组合预测方法的风电场短期风速预测 543-547 第32卷, 第4期 * |
| 《电网与清洁能源》 20090325 陶玉飞等 风电场风速预测模型研究 53-56 第25卷, 第3期 * |
| 《电网与清洁能源》 20090725 彭怀午等 基于SVM方法的风电场短期风速预测 48-52 第25卷, 第7期 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102682207A (en) * | 2012-04-28 | 2012-09-19 | 中国科学院电工研究所 | Ultrashort combined predicting method for wind speed of wind power plant |
| CN102664426A (en) * | 2012-05-10 | 2012-09-12 | 华北电力大学 | Anti-normalization interval correction method for improving air speed prediction precision of SVM (Support Vector Machine) |
| 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 |
| CN103020743B (en) * | 2012-12-27 | 2016-01-13 | 中国科学院电工研究所 | Wind energy turbine set ultra-short term wind speed forecasting method |
| CN103793511A (en) * | 2014-02-08 | 2014-05-14 | 中能电力科技开发有限公司 | Method for improving wind speed forecast accuracy |
| CN103793511B (en) * | 2014-02-08 | 2017-02-01 | 中能电力科技开发有限公司 | Method for improving wind speed forecast accuracy |
| CN108152528A (en) * | 2016-12-02 | 2018-06-12 | 北京金风科创风电设备有限公司 | Wind measurement method and device |
| CN108152528B (en) * | 2016-12-02 | 2020-02-14 | 北京金风科创风电设备有限公司 | Wind speed measuring method and device |
| WO2019006771A1 (en) * | 2017-07-06 | 2019-01-10 | 杨顺伟 | External force interference resistance method and device for unmanned aerial vehicle |
| CN108804634A (en) * | 2018-06-01 | 2018-11-13 | 深圳乐信软件技术有限公司 | Data scrolling method, apparatus, headend equipment, background server and medium |
| CN108804634B (en) * | 2018-06-01 | 2021-10-29 | 深圳乐信软件技术有限公司 | Data rolling method and device, front-end equipment, background server and medium |
| CN109375292A (en) * | 2018-08-30 | 2019-02-22 | 昆明理工大学 | A kind of prediction of precipitation method based on autoregression integral sliding average and support vector regression |
| CN111325376A (en) * | 2018-12-14 | 2020-06-23 | 北京金风科创风电设备有限公司 | Wind speed prediction method and device |
| CN109685583A (en) * | 2019-01-10 | 2019-04-26 | 博拉网络股份有限公司 | A kind of supply chain needing forecasting method based on big data |
| CN110222899A (en) * | 2019-06-12 | 2019-09-10 | 哈尔滨工业大学 | A kind of cross-country skiing racing track wind speed field prediction technique |
| CN111967653A (en) * | 2020-07-22 | 2020-11-20 | 易天气(北京)科技有限公司 | Method for constructing airport runway wind forecasting model, forecasting method and forecasting system |
| CN112311078A (en) * | 2020-11-04 | 2021-02-02 | 华侨大学 | Solar load adjusting method and device based on information fusion |
| CN112311078B (en) * | 2020-11-04 | 2022-03-22 | 华侨大学 | Solar load adjusting method and device based on information fusion |
| CN114117316A (en) * | 2021-11-18 | 2022-03-01 | 国网山东省电力公司电力科学研究院 | Dynamic wind zone drawing method and system based on multiple data sources |
| CN114662800A (en) * | 2022-05-18 | 2022-06-24 | 浙江浙能能源服务有限公司 | Wind power prediction method and system based on artificial neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102236795B (en) | 2013-06-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102236795B (en) | Method for forecasting wind speed in wind power station | |
| CN102269124B (en) | Ultra-short term wind power station generated power forecasting system | |
| CN106505631B (en) | Intelligent wind power wind power prediction system | |
| CN102184337B (en) | Dynamic combination analysis method of new energy generating capacity influenced by meteorological information | |
| CN102298707B (en) | Wind power prediction method based on continuous time slice clustering and support vector machine (SVM) modeling | |
| CN107609697B (en) | A kind of Wind power forecasting method | |
| CN103400009B (en) | Based on the wind energy turbine set Dynamic Equivalence of the semi-supervised spectral clustering of division level | |
| Melzi et al. | Hourly solar irradiance forecasting based on machine learning models | |
| CN102880810A (en) | Wind power prediction method based on time sequence and neural network method | |
| CN109934395B (en) | Multi-hydropower-region short-term power load prediction method based on time-sharing and regional meteorological data | |
| CN102479347B (en) | Method and system for forecasting short-term wind speed of wind farm based on data driving | |
| WO2024051524A1 (en) | Joint prediction method and apparatus for hydraulic, wind and photovoltaic generation power | |
| CN109978242A (en) | The photovoltaic power generation cluster power forecasting method and device of scale are risen based on statistics | |
| CN109800898A (en) | A kind of intelligence short-term load forecasting method and system | |
| CN102938562B (en) | Prediction method of total wind electricity power in area | |
| CN104077632A (en) | Wind power field power prediction method based on deep neural network | |
| CN103345585A (en) | Wind power prediction correction method and system based on support vector machine | |
| CN102738792B (en) | Wind power predicting method | |
| CN112508437A (en) | Method for correcting time sequence similarity error of wind speed forecast in micro-climate environment | |
| CN115829126A (en) | Photovoltaic power generation power prediction method based on multi-view self-adaptive feature fusion | |
| CN102495230B (en) | Correction method used for prediction of wind speed | |
| Yang et al. | Short-term wind power prediction based on multi-parameters similarity wind process matching and weighed-voting-based deep learning model selection | |
| CN113505909B (en) | Error compensation method for short-term wind power trend prediction | |
| CN108694475B (en) | Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model | |
| CN113554203B (en) | Wind power prediction method and device based on high-dimensional meshing and LightGBM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C56 | Change in the name or address of the patentee |
Owner name: INNER MONGOLIA ELECTRIC POWER SURVEY DESIGN INSTIT Free format text: FORMER NAME: INNER MONGOLIA ELECTRICAL POWER INVESTIGATE AND DESIGN INSTITUTE |
|
| CP01 | Change in the name or title of a patent holder |
Address after: 010020 the Inner Mongolia Autonomous Region Hohhot City Xilin Road No. 209 Patentee after: INNER MONGOLIA ELECTRIC POWER SURVEY & DESIGN INSTITUTE CO., LTD. Address before: 010020 the Inner Mongolia Autonomous Region Hohhot City Xilin Road No. 209 Patentee before: Inner Mongolia Electric Power Survey Design Institute |