Summary of the invention
For addressing the above problem, the invention provides a kind of modification method that is used to predict wind speed.
The invention provides a kind of modification method that is used to predict wind speed, it is characterized in that, this method may further comprise the steps:
A. according to the powertrace of wind-powered electricity generation unit and the power error value scope of the said wind-powered electricity generation unit of predesignating, confirm first modified value and second modified value of prediction wind speed;
B. when predicting air speed value, said prediction air speed value is confirmed as first modified value less than first modified value;
C. when predicting air speed value, said prediction air speed value is confirmed as second modified value greater than second modified value.
Forecast wind speed through to different wind speed sections carries out different corrections, has improved the accurate qualified rate of forecast wind speed, and has reduced the error of forecast wind speed.
Wherein, before steps A, this method also comprises:
Through the operational data of wind-powered electricity generation unit, obtain the powertrace of said wind-powered electricity generation unit.
The data that record during real work through the wind-powered electricity generation unit, thus the powertrace that obtains is more accurate.
Wherein, steps A comprises:
According to the power error value of the wind-powered electricity generation unit of predesignating, obtain the error upper limit curve and the error lower limit curve of said powertrace;
Error upper limit curve and error lower limit curve according to said powertrace obtain predicting air speed error upper limit curve and error lower limit curve;
According to said prediction air speed error upper limit curve and error lower limit curve, confirm first modified value and second modified value of prediction wind speed.
Power error scope through system's regulation; Derive the forecast error range that wind speed allowed; According to the wind speed of this error range and each wind speed section and the characteristics of power, formulate modified value, thus can be according to the wind speed of wind-powered electricity generation unit and the actual relationship of power; To the correction of forecast wind speed, make the forecast wind speed more accurate.
Wherein, this device comprises:
The modified value acquisition module is used for confirming first modified value and second modified value of prediction wind speed according to the powertrace of wind-powered electricity generation unit and the power error value scope of the said wind-powered electricity generation unit of predesignating;
Prediction wind speed deriving means is used to obtain the prediction wind speed;
Prediction wind speed correcting module is used for the prediction wind speed that is obtained is revised, and when predicting air speed value less than first modified value, said prediction air speed value is confirmed as first modified value; Be used for when predicting air speed value, said prediction air speed value being confirmed as second modified value greater than second modified value.
Forecast wind speed through to different wind speed sections carries out different corrections, has improved the accurate qualified rate of forecast wind speed, and has reduced the error of forecast wind speed.
Wherein, this device also comprises:
The powertrace acquisition module through the operational data of wind-powered electricity generation unit, obtains the powertrace of said wind-powered electricity generation unit.
The data that record during real work through the wind-powered electricity generation unit, thus the powertrace that obtains is more accurate.
Wherein, the modified value acquisition module comprises:
Powertrace error acquisition unit is used for the power error value according to the wind-powered electricity generation unit of predesignating, and obtains the error upper limit curve and the error lower limit curve of said powertrace;
Prediction air speed error acquiring unit, error upper limit curve and error lower limit curve according to said powertrace obtain predicting air speed error upper limit curve and error lower limit curve;
The modified value acquiring unit is used for according to said prediction air speed error upper limit curve and error lower limit curve, confirms first modified value and second modified value of prediction wind speed.
Power error scope through system's regulation; Derive the forecast error range that wind speed allowed; According to the wind speed of this error range and each wind speed section and the characteristics of power, formulate modified value, thus can be according to the wind speed of wind-powered electricity generation unit and the actual relationship of power; To the correction of forecast wind speed, make the forecast wind speed more accurate.
Embodiment
The embodiment of the invention provides a kind of modification method that is used to predict wind speed.To combine accompanying drawing below, the embodiment of the invention will be described in detail.Referring to Fig. 2, this method may further comprise the steps:
S200:, obtain the powertrace of the corresponding different capacity of different wind speed through the operational data of wind-powered electricity generation unit;
In wind energy turbine set central control system database, store the wind speed that records in the anemometer of wind-powered electricity generation unit; And the real power of this unit under different wind speed; Utilize the wind speed and the power data of wind-powered electricity generation unit, can obtain powertrace, referring to the curve in the curve shown in Figure 12.The horizontal ordinate of powertrace is the wind speed that anemometer records, and ordinate is the real power of wind-powered electricity generation unit.
S210:, obtain the error upper limit curve and the error lower limit curve of powertrace according to the power error value of the wind-powered electricity generation unit of predesignating.
Have certain error during the work of wind-powered electricity generation unit; Usually be 20% of unit installed capacity in the root-mean-square error value that this area allowed; Promptly; Allow the error range of the wind-powered electricity generation power of the assembling unit, concerning single predicted value, the absolute error that the error range that is allowed is this predicted value the unit installed capacity 20% in.
According to this error amount and above-mentioned powertrace, the curve of the power error upper limit that can obtain being allowed under the different wind speed and the curve of power error lower limit are referring to curve shown in Figure 1 middle and upper part curve 1 and lower curve 3.For example, when actual wind speed is 5m/s (meter per second), through curve 2; Can obtain real power and be 210KW (kilowatt), through curve 1 and 3, error range is at-90KW to 510KW in theory; Consider real power all be on the occasion of, can obtain the error range that the wind-powered electricity generation unit allows this power thus, for example: 0KW to 510KW; The i.e. error range that all can be used as this power and allowed of performance number in this scope, that is, and effective estimation range of this power; As long as the power of being predicted is in 0KW to 510KW scope, all can be used as to 210KW (kilowatt) effective predicted value of power.
S220: error upper limit curve and error lower limit curve according to powertrace, powertrace obtain predicting air speed error upper limit curve and error lower limit curve.
Error upper limit curve and error lower limit curve through powertrace, powertrace can obtain the real power of actual wind speed, the error power that is allowed, and utilize power then; The error power corresponding air speed that allowed can be obtained, effective prediction wind speed that actual wind speed allows can be obtained, promptly; The predicated error that is allowed; Referring to Fig. 2, curve 4 is prediction air speed error upper limit curve, and curve 5 is prediction air speed error lower limit curve.
Can utilize this two curves, the wind speed of prediction is in the past estimated, be qualified point as long as the predicted value of wind speed dropped between these two curves.
S230:, confirm first modified value and second modified value of prediction wind speed according to prediction air speed error upper limit curve and error lower limit curve.
With the ordinate of prediction air speed error upper limit curve and axis of ordinates intersection point as first modified value, with the ordinate of predicting air speed error lower limit curve and axis of ordinates intersection point as second modified value.Referring to Fig. 3, for example, first modified value is 5m/s, and second modified value is 10m/s.
S240: the wind speed to being predicted is revised.
If the air speed value of being predicted during more than or equal to first modified value and smaller or equal to second modified value, is not revised the wind speed of being predicted.
If the air speed value of being predicted is during less than first modified value, the wind speed of being predicted confirmed as equals first modified value.
If the air speed value of being predicted is during greater than second modified value, the wind speed of being predicted confirmed as equals second modified value.
Through utilizing historical data in the past, can find that this modification method has following advantage:
With low wind speed section prediction air speed value is that 2m/s is an example, like Fig. 3 and Fig. 4.Because the actual wind speed that it is corresponding is not also known; So the actual wind speed of this prediction air speed value might appear at any one position on the corresponding straight line of this predicted value; Those drop on those points that are worth between prediction air speed error upper limit curve and the error lower limit curve and are called qualified point; Drop on the lower left corner referring to Fig. 3 and scribble the zone of straight line and those points that the upper right corner shown in Figure 4 scribbles the zone of straight line, these qualified points are the qualified prediction wind speed of prediction.Those drop on, and the point of those beyond the value is underproof prediction wind speed for not conforming to lattice point between prediction air speed error upper limit curve and the error lower limit curve.
To hang down mention first modified value on the wind speed section prediction air speed value 2m/s after; Consider from qualified number of spots; Because in 20% the air speed error interval (between prediction air speed error upper limit curve and the error lower limit curve); Low wind section zone is that an area constantly becomes big process by the prediction wind speed from low to high, therefore in low wind speed section, will predict carry on the air speed value after, can be on the basis of not reducing original qualified number of spots; To much originally mention in 20% error range, effectively increase the shared number percent of qualified point at the extraneous point of 20% error band.
Consider that from error amount in the ideal case, predicted value is equal to actual value, that is, predicted value is distributed on the oblique line of y=x under the ideal situation, referring to the straight line 6 of Fig. 3 and Fig. 4.The distribution range of predicted value is more near oblique line y=x, and effect is just good more.
And after will predicting that air speed value is carried generally, in low wind speed section, the data point and the distance between this oblique line that distribute in the overwhelming majority scope of oblique line y=x right side all will significantly reduce; Be that absolute error can reduce; Only the data point in seldom a part of zone, the y=x in 20% error range left side has the increase (distance between these data points and this oblique line) of absolute error, though the absolute error of these data points increased; The increase of these data point absolute errors is the increase in allowed band still; Promptly can not surpass 20%, therefore will predict on air speed value is in low wind speed section and carry, reduce the error of wind speed predicted value.
Obviously, through improving, can improve the qualification rate of the wind speed predicted value in the 20% power error scope effectively, and can reduce the error of little wind speed predicted value effectively, and then improve the forecast precision of power in the prediction air speed value in the low wind speed section.
The correcting process of high wind speed section and low wind section principle are similar; Just with the air speed value of being predicted during greater than second modified value; The wind speed of being predicted is forced down second modified value, identical with top analytical approach, utilize historical data in the past; Can prove: through improving in the prediction air speed value in the high wind speed section; Can improve the qualification rate of the wind speed predicted value in the 20% power error scope effectively, and can reduce the error of little wind speed predicted value effectively, and then improve the forecast precision of power.
Can take not repair to the forecasting wind speed value of middle wind speed section, or further this wind speed interval segmented,, take linear the correction, or the Neural Network Based Nonlinear approximating method revised through analysis of history NWP prediction wind velocity distribution.
Fig. 6 predicts that wind speed distribution of results synoptic diagram, Fig. 7 are to utilize modification method of the present invention to the revised synoptic diagram as a result of the prediction wind speed of Fig. 6 for certain typhoon group of motors one day; As can beappreciated from fig. 7; After typhoon group of motors purpose prediction wind speed is used method of the present invention and is revised; The forecast point number percent that drops in the 20% power error wind speed boundary brings up to 92.7083% from 63.5417%, and the root-mean-square error RMSE on this unit same day is decreased to 12.98256% from 31.43502%.
The embodiment of the invention also provides a kind of correcting device that is used to predict wind speed; Referring to Fig. 8, this device comprises prediction wind speed acquisition module 300, modified value acquisition module 310 and the prediction wind speed correcting module 320 that is connected with, modified value acquisition module 310 with prediction wind speed acquisition module 300 respectively.
Prediction wind speed acquisition module 300 is used to obtain the prediction wind speed; Modified value acquisition module 310 is used for confirming first modified value and second modified value of prediction wind speed according to the powertrace of wind-powered electricity generation unit and the power error value scope of the said wind-powered electricity generation unit of predesignating; Prediction wind speed correcting module 320 is used for when predicting air speed value less than first modified value, said prediction air speed value being confirmed as first modified value; When predicting air speed value, said prediction air speed value is confirmed as second modified value greater than second modified value.
This device also comprises: powertrace acquisition module 330 through the operational data of wind-powered electricity generation unit, obtains the powertrace of said wind-powered electricity generation unit.Modified value acquisition module 310 comprises powertrace error acquisition unit, is used for the power error value according to the wind-powered electricity generation unit of predesignating, and obtains the error upper limit curve and the error lower limit curve of said powertrace; Prediction air speed error acquiring unit, error upper limit curve and error lower limit curve according to said powertrace obtain predicting air speed error upper limit curve and error lower limit curve; The modified value acquiring unit is used for according to said prediction air speed error upper limit curve and error lower limit curve, confirms first modified value and second modified value of prediction wind speed.
The embodiment corresponding method embodiment of apparatus of the present invention, this device can be implemented arbitrary step of said method, and it is specifically realized referring to method embodiment.
The above is merely preferred embodiment of the present invention; Not in order to restriction the present invention; For example, can the module of difference in functionality be realized etc. through an integrated chip, all within spirit of the present invention and principle; Any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.