CN103473621A - Wind power station short-term power prediction method - Google Patents
Wind power station short-term power prediction method Download PDFInfo
- Publication number
- CN103473621A CN103473621A CN2013104564738A CN201310456473A CN103473621A CN 103473621 A CN103473621 A CN 103473621A CN 2013104564738 A CN2013104564738 A CN 2013104564738A CN 201310456473 A CN201310456473 A CN 201310456473A CN 103473621 A CN103473621 A CN 103473621A
- Authority
- CN
- China
- Prior art keywords
- power
- wind speed
- wind
- blower fan
- separate unit
- 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
Images
Landscapes
- Wind Motors (AREA)
Abstract
The invention provides a wind power station short-term power prediction method. The wind power station short-term power prediction method comprises the steps that 1 a BP neural network is used for establishing the historical data statistics relationship between forecasting factors of a mesoscale numerical value weather forecast and the actually-measured wind speed, at the position of the hub height, of a single fan to generate a wind speed statistics downscaling model of the single fan; 2 according to the forecasting factors of the mesoscale numerical value weather forecast of a wind power station in the following 48 hours and the statistics downscaling model of the single fan in the step 1, the predicted wind speed, at the position of the hub height, of the single fan is generated; 3 according to historical data of the actually-measured wind speed of the single fan and the actually-measured power of the single can, a speed-power characteristic curve of the single fan is fit and the short-term predicted power of the fan is obtained through the combination of the predicted wind speed, at the position of the hub height, of the single fan. Therefore, by means of the wind power station short-term power prediction method, the uncertainty caused by the fact that the mesoscale resolution ratio is insufficient is reduced, and the short-term powder prediction precision of the wind power station is obviously improved.
Description
Technical field
The present invention relates to a kind of method for forecasting short-term power in wind power station.
Background technology
In recent years, along with China energy policy is adjusted, grid connected wind power installed capacity rapid growth, large-scale wind power is concentrated and grid-connected electric power netting safe running is brought to impact.Improve the predictability of output of wind electric field, can effectively reduce the impact that wind-powered electricity generation causes electrical network, alleviate dispatching of power netwoks peak regulation pressure, this is for taking full advantage of wind energy resources, and further improving grid connected wind power installation ratio has positive effect.According to the domestic wind farm power prediction technical manual of having issued and implemented, wind energy turbine set must every day the time is uploaded exert oneself prediction curve accept the precision of prediction examination of 0～24h next day to power dispatching station according to the rules.In order accurately to reflect air motion state next day, must use the input data of mesoscale numerical weather forecast (NWP) pattern output as wind energy turbine set short term power prognoses system, so mesoscale numerical weather prediction model prediction output accuracy has determined the precision of wind energy turbine set short term power prediction to a great extent.Yet, mesoscale model atmospheric physics procedure parameter scheme can not effectively be simulated time grid yardstick (being less than 1km) atmospheric physics process, there are error in synoptic process and the truth of its description, and this error can increase along with the growth of pattern integral operation time.Therefore, grid resolution is not enough describes inaccurate meeting with the atmospheric physics procedure parameter scheme relevant with resolution and mesoscale model is predicted the outcome exist uncertain.Directly apply to power prediction and can bring larger uncertainty, must be fallen the yardstick pre-service to it.
Two kinds of methods of the main employing of yardstick research are fallen in centering yardstick numerical weather forecast output at present: 1, use physical model to solve the impact of wind energy turbine set Local factor on air-flow.This method calculation cost is less, but physical model structure and implementation procedure are comparatively complicated, and precision improves limited.
2, adopt power to fall two time scales approach, for example Fluid Mechanics Computation (CFD) is simulated wind energy turbine set interior flow field evolution process, this method can obtain comparatively accurately wind speed profile, but need to use the CFD method to solve the Navier-Stokes equation when setting up prediction of wind speed Query Database or direct prediction of wind speed, Project Realization complexity and calculation cost are huge, high to hardware requirement.
Summary of the invention
In view of this, fundamental purpose of the present invention is, a kind of method for forecasting short-term power in wind power station is provided, fall yardstick according to statistics and solved the larger problem of mesoscale numerical weather forecast air speed error, effectively reduced the uncertainty that the mesoscale lack of resolution is brought, greatly improve the mesoscale numerical weather forecast and fallen dimension calculation efficiency, significantly improved wind energy turbine set short term power precision of prediction.
Described method for forecasting short-term power in wind power station comprises step:
A, utilize the BP neural network, set up the predictor of mesoscale numerical weather forecast and the historical data statistical relationship between separate unit axial fan hub height actual measurement wind speed, Scale Model falls in the wind speed statistics that generates the separate unit blower fan;
B, according to the predictor of the following 48 hours mesoscale numerical weather forecasts in wind energy turbine set zone, and in steps A, Scale Model falls in separate unit blower fan statistics, generates the prediction of wind speed of separate unit blower fan position hub height;
C, according to the actual measurement wind speed of separate unit blower fan and the historical data of measured power, the wind speed-power characteristic of matching separate unit blower fan, in integrating step B, the prediction of wind speed of separate unit blower fan position hub height obtains this blower fan short-term forecasting power.
By upper, set up statistics by historical data and fall Scale Model, mesoscale pattern count value weather forecast wind speed is added up and fallen yardstick, draw the every Fans of wind energy turbine set position hub height prediction of wind speed, in conjunction with wind speed-power characteristic, every Fans is exerted oneself and predicted, realize the short-term forecasting of exerting oneself of wind energy turbine set integral body.Above-mentioned Forecasting Methodology has effectively reduced the uncertainty that the mesoscale lack of resolution is brought, and has significantly improved wind energy turbine set short term power precision of prediction.
Optionally, the predictor of described mesoscale numerical weather forecast at least comprises: wind speed, wind direction, pressure and the relative humidity of 500hPa geopotential unit, 850hPa geopotential unit, axial fan hub height.
By upper, because above-mentioned predictor can be predicted comparatively accurately, and between different predictor, be weak relevant or irrelevant, therefore the input quantity using above-mentioned predictor as the BP neural network, can be trained the BP neural network comparatively accurately.
Optionally, in steps A, the predictor of centering yardstick numerical weather forecast is carried out normalized.
Optionally, described step C comprises:
C1, collect each blower fan actual measurement air speed data, and the blower fan active power data of time match with it;
C2, wind speed setting step-length, divide the wind speed interval according to the wind speed step-length, the active power value in each wind speed interval rejected to bad point and process;
C3, to record the highest active power value of probability of occurrence in each wind speed interval be this interval power features value;
C4, gather the power features value in all wind speed interval, utilize multistage Gaussian function to carry out curve fitting to it, obtain the analytical function of matching wind speed-power characteristic.
By upper, set up the wind speed of separate unit blower fan-power characteristic analytical function according to historical data, and fall in conjunction with the statistics of separate unit blower fan position hub height the short-term forecasting power that the yardstick wind speed obtains blower fan, significantly improved the wind farm power prediction precision of prediction.
Further, described step C2 comprises:
C21, the active power value in each wind speed interval is sorted from small to large;
C22, the power upper limit of determining active power in each wind speed interval and power lower limit;
C23, active power value in each wind speed interval is rejected lower than the power lower limit with higher than the data of power upper limit and corresponding actual measurement air speed data.
By upper, by power upper limit being set and the power lower limit is rejected the abnormal data in historical data, thereby make wind speed-power characteristic more truly reflect the actual set performance.
Further, also comprise step D: wind-powered electricity generation unit control end, according to the output power of each separate unit blower fan of predicting in step C, is regulated the online power of every Fans.
By upper, realize that wind-powered electricity generation unit control end is controlled the higher several Fans of output power and stopped power online output when the wind-powered electricity generation unit runs into limit and exerts oneself state, transferring is that electric energy is stored by power transfer, utilizes to greatest extent thus natural resources.
Optionally, the historical forecast data of described numerical weather forecast predictor, be no less than 6 calendar months with the time span of the separate unit axial fan hub height of mesoscale numerical weather forecast predictor time match actual measurement wind speed and active power data.
By upper, only by the historical data of half a year, just can set up the wind energy turbine set Short-term Forecasting Model, improving mesoscale NWP and short term power precision of prediction simultaneously, greatly reduced operand, effectively reduce hardware cost.
Optionally, after described step C, also comprise: the predicted power summation-wind energy turbine set integrated plant electricity consumption amount of losing of calculating full factory short-term forecasting power=all blower fans.
By upper, owing to producing electricity consumption in the running of wind generating set process, lose, for realizing more accurate power prediction, the above-mentioned integrated plant electricity consumption amount of losing need be counted.
The accompanying drawing explanation
The process flow diagram that Fig. 1 is method for forecasting short-term power in wind power station of the present invention;
The principle schematic that Fig. 2 is BP neural network of the present invention;
The former factory atmosphere speed that Fig. 3 is the separate unit blower fan-power characteristic schematic diagram;
Former factory atmosphere speed-power characteristic that Fig. 4 is separate unit blower fan of the present invention and matching wind speed-power characteristic schematic diagram.
Embodiment
Method for forecasting short-term power in wind power station provided by the present invention, by the mesoscale model prediction is exported and is fallen the yardstick computing, draw the short-term forecasting wind speed of separate unit blower fan position hub height, and then realize the short term power of every Fans is predicted, further realize the short-term forecasting that wind energy turbine set integral body is exerted oneself.Above-mentioned Forecasting Methodology has effectively reduced the uncertainty that the mesoscale lack of resolution is brought, and has significantly improved the wind farm power prediction precision of prediction.
The wind farm power prediction method of as shown in Figure 1, based on statistics, falling yardstick comprises:
Step 10: collect each blower fan actual measurement air speed data, the historical forecast data of the zone of the blower fan active power data of time match, and wind energy turbine set with it mesoscale numerical weather forecast predictor.
The selection of mesoscale numerical weather forecast predictor is that link important in the two time scales approach process falls in applied statistics, and the selection of predictor has determined the Forecast characteristic of the local meteorological condition of wind energy turbine set to a great extent.Choose the predictor that wind speed is had to appreciable impact, reduce its quantity, can effectively reduce the complexity that Scale Model falls in statistics, reduce the model calculated amount, avoid introducing extra interfere information.Choose predictor and follow following methods:
1, predictor must be able to be predicted more exactly by the mesoscale numerical weather prediction model, and has significant Nonlinear Statistical relation between predictor and wind energy turbine set domestic site meteorological element, and this statistical relationship is stable and effective;
2, predictor must be able to reflect important Mesoscale physical change process;
3, between different predictor, be weak relevant or irrelevant.
Based on above-mentioned some, selected predictor is respectively: wind speed, wind direction, pressure and the relative humidity of the 500hPa geopotential unit that the mesoscale numerical weather forecast provides, 850hPa geopotential unit, axial fan hub height.
In the wind-powered electricity generation unit, the hub positions place of every Fans is provided with air velocity transducer, but the actual measurement wind speed of this Fans present position hub height of this air velocity transducer Real-time Obtaining.Described historical data is at least half a year, preferably, is 1 year.
Step 20: use the statistics of BP neural network separate unit blower fan to fall Scale Model.
Use BP neural network statistics to fall Scale Model, specifically, be the statistical relationship between mesoscale numerical weather forecast predictor and separate unit blower fan position hub height wind speed, fall Scale Model using this statistical relationship as the statistics of separate unit blower fan and be applied to produce this blower fan position hub height prediction of wind speed.
The BP neural network is a kind of Multi-layered Feedforward Networks by the training of error backpropagation algorithm, can approach any Nonlinear Mapping with arbitrary accuracy, and its topological structure is comprised of input layer, hidden layer and output layer, and this model is output as:
in formula, the t that y (t) is model output is separate unit axial fan hub height forecasting wind speed value constantly; F is transport function, and in the present embodiment, transport function is the tangent hyperbolic function; w
jfor connecting the weight coefficient of hidden layer and output layer; x
i(t) be mode input t i predictor value constantly; v
ijfor connecting the weight coefficient of input layer and hidden layer; N is the input layer dimension; The dimension that m is hidden layer; θ
jfor the hidden layer threshold value; θ
0for the output layer threshold value.
Take and set up separate unit blower fan statistics and fall Scale Model and illustrate as example, the mesoscale numerical weather prediction model predictor that modeling data is year June in January, 2012 to 2012 and blower fan actual measurement air speed data, data time resolution is 15min, data length is chosen 0～24h next day, using wind speed, wind direction, pressure and the relative humidity of the 500hPa geopotential unit of pattern output, 850hPa geopotential unit, axial fan hub height as the model training input quantity, axial fan hub height wind speed measured value, as the training output quantity, is set up the BP neural network model of three layers.This BP neural network input layer number is 6; The hidden layer neuron number is preferentially determined through test; Network output layer neuron number is 1.During training BP neural network, the input and output layer data is carried out to normalized, adopts following method to carry out normalized here:
the pressure of take describes as example, and X means to predict pressure values, X
minmean default minimum pressure value, X
maxmean default maximum pressure value,
mean normalized pressure values.
Analyze through screening, determine that the hidden layer neuron number is at 19 o'clock, training sample error minimum, now each weight coefficient matrix and threshold matrix are also determined thereupon.BP neural network model after weighted value and threshold matrix parameter identification falls the scale prediction model as the statistics of blower fan.
Step 30: fall according to separate unit blower fan statistics the wind energy turbine set zone short-term forecasting wind speed that Scale Model and mesoscale numerical weather forecast provide, generate separate unit blower fan position hub height short-term forecasting wind speed.
After completing separate unit blower fan statistics and falling the yardstick modeling, Scale Model, just the short-term forecasting wind speed of exportable separate unit blower fan position hub height are fallen in mesoscale numerical weather forecast predictor (500hPa geopotential unit, 850hPa geopotential unit, axial fan hub height wind speed, wind direction, pressure and relative humidity) the input statistics of following 24 or 48 hours.
Two time scales approach falls in mesoscale numerical weather forecast statistics, use above-mentioned 6 predictor just can calculate the forecasting wind speed value of separate unit blower fan position hub height, calculated amount is little and computing velocity is fast, can meet the requirement of wind energy turbine set short term power predictive engine fully, the method is fallen the yardstick technology than existing power, greatly improves counting yield.
Step 40: matching wind speed-power characteristic of setting up the separate unit blower fan.
Prediction of wind speed is converted into to predicted power and must sets up blower fan wind speed – power mapping relations under actual condition, Fig. 3 reflects that former factory atmosphere speed-power characteristic that blower fan producer provides can not describe the input/output relation of blower fan under actual condition well, and the loose point correspondence of the wind speed-power of the blower fan that is incorporated into the power networks is distributed in a wider zone, in order to hold on the whole the relation of wind speed and power, must be processed wind speed-power scatter diagram.In the present embodiment, the startup wind speed of blower fan is 3m/s, and cut-out wind speed is 25m/s, take 0.1m/s as the interval step-length of wind speed, and the wind speed interval is [3-3.1], [3.1-3.2], and [3.2-3.3] ..., [24.8-24.9], [24.9-25], totally 41 wind speed intervals.
[3-3.1] wind speed of take is interval is the example explanation, at first needs that misoperation data point in wind speed and power coordinate system is rejected to bad point and processes.Because the blower fan group is shut down maintenance; blower fan group operation exception; the reasons such as air velocity transducer is malfunctioning; make in wind speed and power coordinate system and contain a large amount of misoperation data point (being scattered in the data point of the discrete Fig. 3 of being distributed in wind speed-power characteristic edge), these abnormity point can have a strong impact on blower fan actual wind speed-power characteristic fitting effect usually.The process of rejecting bad point comprises: all active power values to [3-3.1] wind speed interval sort from small to large, calculate the mean value of active power, power typical value as this wind speed interval, choose come back a certain performance number as power upper limit, come a certain performance number of front as the power lower limit.The upper and lower bound of rejecting bad point in each wind speed interval is revised, the active power value of predicting in each wind speed interval is rejected lower than lower limit with higher than the data point of the upper limit, to complete the rejecting bad point.For example, choose and be positioned at 99%(in the wind speed interval and be 100% to the maximum, minimum is 1%) the active power value located is as power upper limit, is positioned at 1%(and is 100% to the maximum, minimum is 1%) the active power value located is as the power lower limit.All do aforesaid operations for each wind speed interval, determine power upper limit and power lower limit in each wind speed interval.
Secondly, the performance number in different wind speed interval is carried out to statistical calculation, choose active power value that occurrence number is maximum as the power features value, and the power features value all wind speed interval in is connected, foundation matching wind speed-power characteristic as shown in Figure 4.
Finally, the power features value in all wind speed interval is carried out to multistage Gauss curve fitting, result obtains a smooth continuous matching wind speed-power characteristic, and this curve analytical expression P (v) is as follows:
P(v)=a1*exp(-((v-b1)/c1)^2)+a2*exp(-((v-b2)/c2)^2)+a3*exp(-((v-b3)/c3)^2)+a4*exp(-((v-b4)/c4)^2)
a1=1470(1359,1581)、b1=29.55(27.79,31.32)、c1=10.54(-0.1442,21.23)；
a2=445.1(297.2,593)、b2=10.23(10.09,10.37)、c2=2.286(1.929,2.644)；
a3=-2.993e+006(-1.611e+013,1.611e+013)、b3=16.29(-1160,1193)、c3=5.815(-2379,2391)；
a4=2.994e+006(-1.611e+013,1.611e+013)、b4=16.29(-1159,1192)、c4=5.816(-2378,2390)。
In formula, v means air speed value, ai, and bi, ci means the normal value coefficient of multistage Gauss curve fitting curve, the root-mean-square error of matched curve: 15.07.
By the above matched curve equation of short-term forecasting wind speed substitution of separate unit blower fan position hub height, just can obtain the short-term forecasting power of this blower fan.Adopt said method to be analyzed wind energy turbine set wind turbine service data, can obtain its matching wind speed-power characteristic, the Gauss's analytical expression that statistics is fallen to this curve of separate unit blower fan position hub height prediction of wind speed substitution that two time scales approach obtains just can obtain the blower fan predicted power.
Step 50: the wind energy turbine set short term power is predicted.
Gauss's analytical expression that this curve of blower fan position hub height prediction of wind speed substitution that two time scales approach obtains is fallen in statistics just can obtain the short-term forecasting power of separate unit blower fan
Using the input value of prediction of wind speed value matching wind speed-power characteristic in step 40 of separate unit blower fan position hub height in step 30, draw the short-term forecasting power of this separate unit blower fan.And the power sum that the output power of described wind-powered electricity generation unit is produced by all blower fans is calculated and obtain.Further, owing to producing electricity consumption in the running of wind generating set process, lose, therefore the predicted power-electricity consumption of wind-powered electricity generation unit output power=all blower fans is lost.
Preferably, also comprise that step 60(is not shown), wind-powered electricity generation unit control end, according to the output power of each separate unit blower fan of predicting in step 50, is regulated its online power.For instance, when the wind-powered electricity generation unit runs into limit while exerting oneself state, wind-powered electricity generation unit control end is controlled the higher several Fans of output power and is stopped power online output, then is that electric energy is stored by power transfer.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (8)
1. a method for forecasting short-term power in wind power station, is characterized in that, comprises step:
A, utilize the BP neural network, set up the predictor of mesoscale numerical weather forecast and the historical data statistical relationship between separate unit axial fan hub height actual measurement wind speed, Scale Model falls in the wind speed statistics that generates the separate unit blower fan;
B, according to the predictor of the following 48 hours mesoscale numerical weather forecasts in wind energy turbine set zone, and in steps A, Scale Model falls in separate unit blower fan statistics, generates the prediction of wind speed of separate unit blower fan position hub height;
C, according to the actual measurement wind speed of separate unit blower fan and the historical data of measured power, the wind speed-power characteristic of matching separate unit blower fan, in integrating step B, the prediction of wind speed of separate unit blower fan position hub height obtains this blower fan short-term forecasting power.
2. Forecasting Methodology according to claim 1, is characterized in that, the predictor of described mesoscale numerical weather forecast at least comprises: wind speed, wind direction, pressure and the relative humidity of 500hPa geopotential unit, 850hPa geopotential unit, axial fan hub height.
3. power forecasting method according to claim 2, is characterized in that, in steps A, the predictor of centering yardstick numerical weather forecast is carried out normalized.
4. power forecasting method according to claim 1, is characterized in that, described step C comprises:
C1, collect each blower fan actual measurement air speed data, and the blower fan active power data of time match with it;
C2, wind speed setting step-length, divide the wind speed interval according to the wind speed step-length, the active power value in each wind speed interval rejected to bad point and process;
C3, to record the highest active power value of probability of occurrence in each wind speed interval be this interval power features value;
C4, gather the power features value in all wind speed interval, utilize multistage Gaussian function to carry out curve fitting to it, obtain the analytical function of matching wind speed-power characteristic.
5. power forecasting method according to claim 4, is characterized in that, described step C2 comprises:
C21, the active power value in each wind speed interval is sorted from small to large;
C22, the power upper limit of determining active power in each wind speed interval and power lower limit;
C23, active power value in each wind speed interval is rejected lower than the power lower limit with higher than the data of power upper limit and corresponding actual measurement air speed data.
6. power forecasting method according to claim 1, is characterized in that, also comprises step D: wind-powered electricity generation unit control end, according to the output power of each separate unit blower fan of predicting in step C, is regulated the online power of every Fans.
7. power forecasting method according to claim 1, it is characterized in that, the historical forecast data of described numerical weather forecast predictor, with the time span of the separate unit axial fan hub height of mesoscale numerical weather forecast predictor time match actual measurement wind speed and active power data, be no less than 6 calendar months.
8. power forecasting method according to claim 1, is characterized in that, after described step C, also comprises: the predicted power summation-wind energy turbine set integrated plant electricity consumption amount of losing of calculating full factory short-term forecasting power=all blower fans.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013104564738A CN103473621A (en) | 2013-09-29 | 2013-09-29 | Wind power station short-term power prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013104564738A CN103473621A (en) | 2013-09-29 | 2013-09-29 | Wind power station short-term power prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103473621A true CN103473621A (en) | 2013-12-25 |
Family
ID=49798464
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013104564738A Pending CN103473621A (en) | 2013-09-29 | 2013-09-29 | Wind power station short-term power prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103473621A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103996079A (en) * | 2014-05-29 | 2014-08-20 | 东南大学 | Wind power weighting predication method based on conditional probability |
| CN105205544A (en) * | 2014-06-24 | 2015-12-30 | 华北电力大学（保定） | Wind power prediction method based on bi-random theory |
| CN105279576A (en) * | 2015-10-23 | 2016-01-27 | 中能电力科技开发有限公司 | Wind speed forecasting method |
| CN105512766A (en) * | 2015-12-11 | 2016-04-20 | 中能电力科技开发有限公司 | Wind power plant power predication method |
| CN105680485A (en) * | 2014-11-18 | 2016-06-15 | 国家电网公司 | Smooth output method of wind power plant |
| CN106650977A (en) * | 2015-10-29 | 2017-05-10 | 中能电力科技开发有限公司 | Short-term power prediction method used for newly-built wind farm |
| CN109154281A (en) * | 2016-05-23 | 2019-01-04 | 通用电气公司 | System and method for predicting the power output of wind field |
| CN109190845A (en) * | 2018-09-29 | 2019-01-11 | 南京信息工程大学 | A kind of two stages dynamic optimization short-term wind power forecast method |
| CN110442836A (en) * | 2019-07-23 | 2019-11-12 | 国能日新科技股份有限公司 | Short-term wind speed forecasting method of wind farm and system |
| CN111414573A (en) * | 2020-04-28 | 2020-07-14 | 北京拾易技术有限公司 | Input and output data processing method and system in engineering application field |
| CN112561149A (en) * | 2020-12-10 | 2021-03-26 | 江苏冰象智气信息技术有限公司 | Wind power prediction method based on power downscaling data fusion and statistical model |
| CN112836870A (en) * | 2021-01-25 | 2021-05-25 | 华北电力大学 | Method, device, equipment and computer readable storage medium for predicting wind power |
| CN113205210A (en) * | 2021-04-29 | 2021-08-03 | 西安热工研究院有限公司 | Wind speed and power prediction method, system, equipment and storage medium for wind power plant with complex terrain |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102102626A (en) * | 2011-01-30 | 2011-06-22 | 华北电力大学 | Method for forecasting short-term power in wind power station |
| CN102588210A (en) * | 2011-12-21 | 2012-07-18 | 中能电力科技开发有限公司 | Filtering method for preprocessing fitting data of power curve |
| US20120185414A1 (en) * | 2010-12-15 | 2012-07-19 | Vaisala, Inc. | Systems and methods for wind forecasting and grid management |
| CN102938562A (en) * | 2012-07-19 | 2013-02-20 | 中国电力科学研究院 | Prediction method of total wind electricity power in area |
| CN102945509A (en) * | 2012-10-18 | 2013-02-27 | 中能电力科技开发有限公司 | Wind electric field power prediction method |
| CN103001249A (en) * | 2012-11-15 | 2013-03-27 | 中国电力科学研究院 | Method for predicting short-term power of wind farm on the basis of BP (back propagation) neural network |
-
2013
- 2013-09-29 CN CN2013104564738A patent/CN103473621A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120185414A1 (en) * | 2010-12-15 | 2012-07-19 | Vaisala, Inc. | Systems and methods for wind forecasting and grid management |
| CN102102626A (en) * | 2011-01-30 | 2011-06-22 | 华北电力大学 | Method for forecasting short-term power in wind power station |
| CN102588210A (en) * | 2011-12-21 | 2012-07-18 | 中能电力科技开发有限公司 | Filtering method for preprocessing fitting data of power curve |
| CN102938562A (en) * | 2012-07-19 | 2013-02-20 | 中国电力科学研究院 | Prediction method of total wind electricity power in area |
| CN102945509A (en) * | 2012-10-18 | 2013-02-27 | 中能电力科技开发有限公司 | Wind electric field power prediction method |
| CN103001249A (en) * | 2012-11-15 | 2013-03-27 | 中国电力科学研究院 | Method for predicting short-term power of wind farm on the basis of BP (back propagation) neural network |
Non-Patent Citations (2)
| Title |
|---|
| 蔡祯祺: ""基于数值天气预报NWP修正的BP神经网络风电功率短期预测研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, no. 7, 15 July 2010 (2010-07-15) * |
| 蔡祯祺: ""基于数值天气预报NWP修正的BP神经网络风电功率短期预测研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》, no. 7, 15 July 2012 (2012-07-15) * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103996079A (en) * | 2014-05-29 | 2014-08-20 | 东南大学 | Wind power weighting predication method based on conditional probability |
| CN103996079B (en) * | 2014-05-29 | 2017-02-15 | 东南大学 | Wind power weighting predication method based on conditional probability |
| CN105205544A (en) * | 2014-06-24 | 2015-12-30 | 华北电力大学（保定） | Wind power prediction method based on bi-random theory |
| CN105205544B (en) * | 2014-06-24 | 2018-07-24 | 华北电力大学（保定） | A kind of wind power forecasting method based on dual random theory |
| CN105680485A (en) * | 2014-11-18 | 2016-06-15 | 国家电网公司 | Smooth output method of wind power plant |
| CN105680485B (en) * | 2014-11-18 | 2017-10-13 | 国家电网公司 | A kind of wind power plant is smoothly exerted oneself method |
| CN105279576A (en) * | 2015-10-23 | 2016-01-27 | 中能电力科技开发有限公司 | Wind speed forecasting method |
| CN106650977A (en) * | 2015-10-29 | 2017-05-10 | 中能电力科技开发有限公司 | Short-term power prediction method used for newly-built wind farm |
| CN105512766A (en) * | 2015-12-11 | 2016-04-20 | 中能电力科技开发有限公司 | Wind power plant power predication method |
| US11242842B2 (en) | 2016-05-23 | 2022-02-08 | General Electric Company | System and method for forecasting power output of a wind farm |
| CN109154281B (en) * | 2016-05-23 | 2021-09-24 | 通用电气公司 | System and method for predicting power output of a wind farm |
| CN109154281A (en) * | 2016-05-23 | 2019-01-04 | 通用电气公司 | System and method for predicting the power output of wind field |
| CN109190845A (en) * | 2018-09-29 | 2019-01-11 | 南京信息工程大学 | A kind of two stages dynamic optimization short-term wind power forecast method |
| CN109190845B (en) * | 2018-09-29 | 2022-05-31 | 南京信息工程大学 | Two-stage dynamic optimization short-term wind power prediction method |
| CN110442836A (en) * | 2019-07-23 | 2019-11-12 | 国能日新科技股份有限公司 | Short-term wind speed forecasting method of wind farm and system |
| CN110442836B (en) * | 2019-07-23 | 2023-05-12 | 国能日新科技股份有限公司 | Wind power plant short-term wind speed prediction method and system |
| CN111414573B (en) * | 2020-04-28 | 2020-12-01 | 北京拾易技术有限公司 | Input and output data processing method and system in engineering application field |
| CN111414573A (en) * | 2020-04-28 | 2020-07-14 | 北京拾易技术有限公司 | Input and output data processing method and system in engineering application field |
| CN112561149A (en) * | 2020-12-10 | 2021-03-26 | 江苏冰象智气信息技术有限公司 | Wind power prediction method based on power downscaling data fusion and statistical model |
| CN112561149B (en) * | 2020-12-10 | 2024-08-09 | 江苏冰象智气信息技术有限公司 | Wind power prediction method based on dynamic downscaling data fusion and statistical model |
| CN112836870A (en) * | 2021-01-25 | 2021-05-25 | 华北电力大学 | Method, device, equipment and computer readable storage medium for predicting wind power |
| CN112836870B (en) * | 2021-01-25 | 2023-12-12 | 华北电力大学 | Method, apparatus, device and computer readable storage medium for predicting wind power |
| CN113205210A (en) * | 2021-04-29 | 2021-08-03 | 西安热工研究院有限公司 | Wind speed and power prediction method, system, equipment and storage medium for wind power plant with complex terrain |
| CN113205210B (en) * | 2021-04-29 | 2023-07-21 | 西安热工研究院有限公司 | Wind power plant wind speed and power prediction method, system and equipment for complex terrain and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103473621A (en) | Wind power station short-term power prediction method | |
| CN103489046A (en) | Method for predicting wind power plant short-term power | |
| CN107766990B (en) | Method for predicting power generation power of photovoltaic power station | |
| CN102184337B (en) | Dynamic combination analysis method of new energy generating capacity influenced by meteorological information | |
| CN103020462B (en) | Take into account the wind energy turbine set probability output power calculation algorithms of complicated wake effect model | |
| CN113496311A (en) | Photovoltaic power station generated power prediction method and system | |
| Yang et al. | Day-ahead forecasting of photovoltaic output power with similar cloud space fusion based on incomplete historical data mining | |
| CN105512766A (en) | Wind power plant power predication method | |
| CN105069521A (en) | Photovoltaic power plant output power prediction method based on weighted FCM clustering algorithm | |
| CN103268366A (en) | Combined wind power prediction method suitable for distributed wind power plant | |
| CN104978608A (en) | Wind power prediction apparatus and prediction method | |
| CN111488896B (en) | Distribution line time-varying fault probability calculation method based on multi-source data mining | |
| CN110264002B (en) | Wind power plant micro-siting scheme evaluation method based on cluster analysis | |
| CN102147839A (en) | Method for forecasting photovoltaic power generation quantity | |
| CN114792156A (en) | Photovoltaic output power prediction method and system based on curve characteristic index clustering | |
| CN106650977A (en) | Short-term power prediction method used for newly-built wind farm | |
| CN111626473A (en) | Two-stage photovoltaic power prediction method considering error correction | |
| WO2024087876A1 (en) | Wind power forecasting error determination method, and processor and storage medium | |
| CN117272850B (en) | Elastic space analysis method for safe operation scheduling of power distribution network | |
| CN105825002A (en) | Method for modeling dynamic equivalence of wind power farm based on dynamic grey-relevancy analysis method | |
| CN115759467A (en) | Time-division integrated learning photovoltaic prediction method for error correction | |
| CN112580899A (en) | Medium-and-long-term wind power generation prediction method and system fused with machine learning model | |
| CN111008725B (en) | Meteorological factor fluctuation feature extraction method for short-term wind power prediction | |
| CN112633565A (en) | Photovoltaic power aggregation interval prediction method | |
| CN109858668B (en) | Coordination prediction method for power load region in thunder and lightning climate |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into 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: 20131225 |