CN108488038B - A kind of Yaw control method of wind power generating set - Google Patents

Abstract

A kind of Yaw control method of wind power generating set, comprising: Step 1: calculating separately wind speed average value and the Mathematics models in preset duration according to the wind speed and direction got, historical wind speed data and history wind direction data are obtained, according to the air speed data and wind direction data of historical wind speed data and history wind direction data prediction subsequent time；Step 2: determining control parameter according to the air speed data of subsequent time, and yaw control is carried out to wind power generating set using control parameter and wind direction data.Compared to traditional Yaw control method, the yaw number of this method increases relative to Traditional control strategy, but the number improved is concentrated mainly on middle high wind speed area, therefore power loss coefficient is substantially reduced.This method can effectively reduce the yaw error in middle high wind speed area, to reduce power loss coefficient (improving the utilization rate of wind energy).

Description

A yaw control method for wind turbines

technical field

The present invention relates to the technical field of wind power generation, in particular to a yaw control method of a wind power generating set.

Background technique

At present, with the exhaustion of traditional fossil fuels and the increasing demand for energy, people are paying more and more attention to the development and utilization of renewable green and clean energy. As one of the green renewable energy power generation methods, wind power generation has attracted the attention of industry and academia in various countries. Wind power generation technology is becoming more and more mature, and the cost of renewable energy is relatively low, so it has broad development prospects.

The yaw regulator is the wind adjustment device of the wind turbine, which makes the axis of the wind wheel of the wind turbine always consistent with the wind direction, and the control accuracy of the regulator has a significant impact on the power generation performance of the wind turbine. Modern large wind turbines operate under the premise of yaw error.

On the one hand, the existence of yaw error will lead to the reduction of wind energy acquisition. According to relevant data, the annual average energy loss caused by yaw error is 2.7%, and when the yaw error is 20°, the annual loss can reach 11%. %. On the other hand, the existence of yaw error will also cause the increase of component load, which will lead to unstable yaw and cause the generator set to oscillate and cause shutdown.

With the increasing size of modern wind turbine blades, the influence of the yaw adjuster has gradually become more prominent. Relevant data show that the failure rate caused by the yaw system accounts for 12.5%, and the downtime caused by the yaw failure accounts for 13.3%. Therefore, it is necessary to conduct in-depth research on the active yaw control device and control strategy of large-scale wind turbines.

SUMMARY OF THE INVENTION

In order to solve the above problems, the present invention provides a yaw control method for a wind turbine, the yaw control method includes:

Step 1. Calculate the average value of the wind speed and the average value of the wind direction within the preset time length according to the obtained wind speed and wind direction, obtain historical wind speed data and historical wind direction data, and predict the wind speed at the next moment according to the historical wind speed data and historical wind direction data. data and wind direction data;

Step 2: Determine control parameters according to the wind speed data at the next moment, and use the control parameters and wind direction data to perform yaw control on the wind turbine.

According to an embodiment of the present invention, in the step 1, the preset duration is 10s, 30s or 60s.

According to an embodiment of the present invention, in the step 1, the step of predicting the wind speed data and wind direction data at the next moment includes:

Decompose the wind vector according to the historical wind speed data and the historical wind direction data to obtain historical wind vector abscissa data and historical wind vector ordinate data;

Utilize the ARMA model to determine the wind vector abscissa data and the wind vector ordinate data at the next moment according to the historical wind vector abscissa data and the historical wind vector ordinate data;

According to the wind vector abscissa data and the wind vector ordinate data at the next moment, the wind speed data and the wind direction data at the next moment are respectively determined.

According to one embodiment of the present invention, the wind vector is decomposed according to the following expression:

in, and respectively represent the wind vector abscissa data and the wind vector ordinate data at time t, represents the wind speed data, Represents the wind direction data at time t.

According to an embodiment of the present invention, the wind speed data at the next moment is determined according to the following expression:

in, represents the wind speed data at time t+1, and Respectively represent the wind vector abscissa data and the wind vector ordinate data at time t+1.

According to an embodiment of the present invention, the wind direction data at the next moment is determined according to the following expression:

in, represents the wind direction data at time t+1, and Respectively represent the wind vector abscissa data and the wind vector ordinate data at time t+1.

According to an embodiment of the present invention, in the step 1, the step of predicting the wind direction data at the next moment includes:

Perform circular variable transformation on the historical wind direction data to obtain the sine and cosine values of the historical wind direction data;

Use the ARMA model to determine the sine and cosine values of the wind direction data at the next moment according to the sine and cosine values of the historical wind direction data, and determine the wind direction data at the next moment according to the sine and cosine values of the wind direction data at the next moment. .

According to an embodiment of the present invention, circular variable transformation is performed on the historical wind direction data according to the following expression:

in, and represent the sine and cosine values of the wind direction data at time t, respectively, Represents the wind direction data at time t.

According to an embodiment of the present invention, the wind direction data at the next moment is determined according to the following expression:

in, represents the wind direction data at time t+1, and Represent the sine and cosine values of the wind direction data at time t+1, respectively.

According to an embodiment of the present invention, in the first step, the ARMA model is used to determine the wind direction data at the next moment according to the historical wind direction data.

According to an embodiment of the present invention, in the first step, the ARMA model is used to determine the wind speed data at the next moment according to the historical wind speed data.

According to an embodiment of the present invention, the step of determining the wind speed data at the next moment includes:

Step a, performing detrending processing on the historical wind speed data to obtain detrending wind speed data;

Step b, according to the autocorrelation function and the partial autocorrelation function of the detrended wind speed data, determine the tail truncation mode;

Step c, based on the smear truncation mode, utilize preset criteria to determine the order of the ARMA model, and determine the automatic regression order, the moving average order and the difference order;

Step d, based on the ARMA model, using the automatic regression order, the moving average order and the difference order to calculate the wind speed data at the next moment according to the detrended wind speed data.

According to an embodiment of the present invention, in the second step, the wind speed interval to which the wind speed data at the next moment belongs is determined, and the control parameter is determined according to the wind speed interval to which the wind speed data belongs.

According to an embodiment of the present invention, in the second step, if the wind speed data at the next moment is less than the preset cut-in wind speed, the wind generator set is controlled to be in a shutdown state.

According to an embodiment of the present invention, in the second step, if the wind direction data at the next moment is greater than or equal to the preset cut-out wind speed, the wind turbine is controlled to yaw to a downwind position and is in a shutdown state.

According to an embodiment of the present invention, in the second step,

If the wind speed data at the next moment is greater than or equal to the preset cut-in wind speed and less than the first preset wind speed threshold, keeping the original control parameters unchanged;

And/or, if the wind speed data at the next moment is greater than or equal to the first preset wind speed threshold and less than the preset cut-out wind speed, the original control parameter is reduced by a specific value to obtain the required control parameter.

According to an embodiment of the present invention, in the second step, several wind speed intervals are included between the first preset wind speed threshold and the preset rated wind speed, wherein, for these wind speed intervals, the larger the wind speed is , the control parameter corresponding to the wind speed interval is smaller.

According to an embodiment of the present invention, in the second step, if the wind speed data at the next moment is greater than or equal to the preset rated wind speed and less than the preset cut-out wind speed, then according to the wind direction data at the next moment, the The wind power generating set performs yaw control so that the yaw error of the wind power generating set is within a preset error range.

Compared with the traditional yaw control method, the yaw number of the yaw control method provided by the present invention is improved compared with the traditional control strategy, but the increased number of times is mainly concentrated in the middle and high wind speed area, so the power loss coefficient is significantly reduced. The partitioned predictive control method provided by the present invention can effectively reduce the yaw error in the middle and high wind speed regions, thereby reducing the power loss coefficient (ie, improving the utilization rate of wind energy).

Other features and advantages of the present invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the description, claims and drawings.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the description of the embodiments or the prior art will be briefly introduced below:

Fig. 1 is a structural schematic diagram of a wind turbine with an active yaw regulator;

Fig. 2 is a schematic diagram of the yaw system driving motor rotating forward so that the wind turbine nacelle is adjusted clockwise;

Fig. 3 is a schematic diagram of the yaw system driving motor rotating forward so that the wind turbine nacelle is adjusted counterclockwise;

Fig. 4 is the realization flow schematic diagram of the existing yaw logic control algorithm;

Figures 5 to 7 show the distribution relationship between the wind speed and the wind direction of a wind farm in the south;

Figures 8 to 10 are schematic diagrams of actual operation results under the traditional yaw control strategy;

FIG. 11 is a schematic flowchart of the implementation of a wind speed independent prediction method according to an embodiment of the present invention;

12 and 13 are schematic diagrams of the autocorrelation function and the partial correlation function of the average value of the wind speed sequence 10s according to an embodiment of the present invention;

FIG. 14 is a schematic flowchart of the implementation of a method for predicting wind speed and direction according to an embodiment of the present invention;

FIG. 15 is a schematic flowchart of the implementation of a method for predicting wind speed and direction according to an embodiment of the present invention;

FIG. 16 shows a schematic diagram of the original wind direction and the average wind direction under different time periods according to an embodiment of the present invention;

FIG. 17 shows a schematic diagram of the original wind speed and the average wind speed under different time periods according to an embodiment of the present invention;

FIG. 18 and FIG. 19 respectively show schematic diagrams of 10s wind direction prediction results and wind speed prediction results obtained by different prediction methods according to an embodiment of the present invention;

20 and 21 respectively show schematic diagrams of the 30s wind direction prediction result and the wind speed prediction result obtained by different prediction methods according to an embodiment of the present invention;

Fig. 22 and Fig. 23 respectively show schematic diagrams of 60s wind direction prediction results and wind speed prediction results obtained by different prediction methods according to an embodiment of the present invention;

FIG. 24 shows a schematic flowchart of the implementation of a yaw control method for a wind turbine according to an embodiment of the present invention;

FIG. 25 shows a graph of ideal operating power of a wind turbine according to an embodiment of the present invention;

Fig. 26 shows a position map of the engine room under different control strategies according to an embodiment of the present invention;

27 to 30 respectively show yaw error distribution diagrams in different wind speed intervals under different control strategies according to an embodiment of the present invention.

Detailed ways

The embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples, so as to fully understand and implement the implementation process of how the present invention applies technical means to solve technical problems and achieve technical effects. It should be noted that, as long as there is no conflict, each embodiment of the present invention and each feature of each embodiment can be combined with each other, and the formed technical solutions all fall within the protection scope of the present invention.

Meanwhile, in the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to those skilled in the art that the present invention may be practiced without the specific details or in the specific manner described herein.

Additionally, the steps shown in the flowcharts of the figures may be performed in a computer system, such as a set of computer-executable instructions, and, although shown in a logical order in the flowcharts, in some cases, may be executed differently The steps shown or described are performed in the order shown herein.

The control of the current yaw system mainly focuses on power control, such as maximum power point tracking (maximum power point tracking, MPPT for short) control. Due to the limitation of measurement technology in the early days, the hill-climbing method was mostly used for yaw control. However, because the MPP of the fan is not only related to the wind direction, but also to the wind speed, it is impossible to accurately locate the MPP, so this method is still controversial in the industry.

With the development of measurement technology, some scholars have proposed yaw control methods and logic control methods that combine PID and fuzzy control. These methods use active yaw control based on wind direction measurement, which is also the yaw commonly used in the industry. Control Method. However, because the measurement of wind direction is always mixed with interference noise and abnormal values, and at the same time, the wind direction is constantly changing, which is different from the future wind direction. Therefore, this active yaw control based on wind direction feedback cannot significantly improve the control performance of the yaw system.

In recent years, some scholars have proposed to use lidar to detect the wind speed and direction 150m in front of the impeller, and based on this, a predictive control of the yaw system has been proposed. This yaw control strategy based on advanced measurement technology can improve wind energy harvest and reduce some extreme wind down loads. However, due to the high cost of this wind measurement technology, it is still in the experimental stage.

The wind direction is also very important for the wind turbine to obtain the maximum power. The yaw control based on wind direction prediction makes it possible to keep the axis of the wind turbine consistent with the wind direction to obtain the maximum power output. Bao et al. proposed a method based on circular regression and Bayesian averaging to correct the bias of the forecast data obtained by the weather forecast model. Ergin Erdem et al. proposed an ARMA-based forecasting method for the combination of wind speed and wind direction. Kalsuner et al. proposed a method to predict wind vectors based on "similar days". Prediction of wind speed and direction are both critical to the rate of wind energy harvesting, and wind speed and direction are two distinct attributes, and there is little research on how to simultaneously predict multiple wind attributes and use predictions for yaw system control. few.

On the basis of the original wind speed and wind direction independent forecasting method based on ARMA, a new wind speed and wind direction forecasting method based on ARMA model is proposed.

Figure 1 is a schematic structural diagram of a wind turbine with an active yaw regulator.

As shown in FIG. 1 , the wind turbine includes: a motor module 101 , a pitch control module 102 , an aerodynamic system module 103 , a frequency converter control module 104 , a yaw control module 105 , and a tower and transmission module. The wind in the air converts the wind energy into mechanical energy through the rotation of the wind turbine blades in the aerodynamic system module 103 to drive the rotor of the generator in the motor module 101 to rotate, and then applies the space vector control technology through the frequency conversion control module 104. The variable frequency and variable voltage are converted into fixed frequency and fixed voltage that the grid can accept.

According to the Bezier theory in aerodynamics, the power P _a that the wind turbine can obtain and output from the wind is:

V _e =V ₀ cos(θ _e )=V ₀ cos(θ _w -θ _np ) (2)

Among them, ρ is the air density, _Ar is the area swept by the wind rotor, C _p is the wind energy utilization coefficient of the wind turbine, _Ve is the effective wind speed, V ₀ is the free flow wind speed, θ _e is the yaw error, θ _w and θ _np denote the wind direction and the northing angle of the wind turbine nacelle, respectively.

According to expressions (1) and (2), it can be known that the power P _a captured by the wind turbine is proportional to the third power of the effective value of the wind speed _Ve , which indicates that the larger the yaw error θ _e , the larger the power P _a captured by the wind turbine. the smaller.

The active yaw system is to actively align the axis of the nacelle with the wind direction, that is, according to the average value of the wind direction detected by the calculated wind vane within a period of time, the wind rotor is adjusted to the windward position through the yaw direction adjustment device. When the position of the wind turbine nacelle changes, the absolute value encoder will record the currently adjusted angle, and then start the yaw braking. Through this series of active yaw adjustment actions, it is possible for the wind turbine to capture the maximum wind energy.

Therefore, in order to improve the efficiency of the wind turbine, the yaw system always requires the shortest path to align the wind direction by turning the nacelle perpendicular to the tower, so the relationship between the shortest path of yaw adjustment and the yaw angle is as follows:

(1) When the angle difference between the position of the wind turbine nacelle and the wind direction is less than 180°, the formula for calculating the yaw angle is:

θ _e = θ _w - θ _np (3)

At this time, the yaw system drives the motor to rotate forward, so that the wind turbine nacelle is adjusted clockwise, and its schematic diagram is shown in Figure 2;

(2) When the angle difference between the position of the wind turbine nacelle and the wind direction is greater than 180°, the calculation formula of the yaw angle is:

θ _e = 360°-|θ _w -θ _np | (4)

At this time, the drive motor of the yaw system is reversed, so that the wind turbine nacelle is adjusted counterclockwise. The schematic diagram is shown in Figure 3.

At present, the yaw error under the active yaw control strategy based on wind direction feedback is mainly concentrated in [-15°, 15°]. When the wind direction changes beyond the set range, the yaw system will adjust the cabin position. The following takes a 1.5MWCMYWP fan as an example to introduce the yaw logic control algorithm commonly used in the industry. The schematic diagram of its realization process is shown in Figure 4.

It can be seen from Figure 4 that the traditional active yaw logic control algorithm firstly filters the original wind direction measurement data during the implementation process, and then calculates the average yaw error within the set time according to the filtered wind direction data. value.

Specifically, the control algorithm calculates the average value of the yaw error within the set time according to the following expression:

in, represents the average value of yaw error within 10s, represents the average value of the yaw error within 30s, Indicates the average yaw error within 60s.

Then, the algorithm will judge whether the calculated average value of yaw error exceeds the preset corresponding range. Among them, if the preset range is not exceeded, the yaw system will not act. If it exceeds the preset range, the algorithm will further determine whether the time when the average value of the yaw error exceeds the preset range exceeds the set delay time. Among them, if the time when the average value of the yaw error exceeds the preset range does not exceed the set delay time, the same yaw system will not operate.

If the average value of the yaw error exceeds the preset range for more than the preset delay time, then the algorithm will calculate the running time tyaw of the _yaw system. Specifically, the algorithm can calculate the running time tyaw of the _yaw system according to the following expression:

t _yaw =θ _e /v _yaw (6)

Among them, v _yaw represents the operating speed of the yaw system (ie, the rotational speed of the yaw system).

After obtaining the running time _{ty yaw} of the yaw system, the algorithm can also control the yaw system to act according to the running time _{ty yaw} of the yaw system.

However, the wind is the movement of the air relative to the surface of the earth. Its formation is affected by various factors such as geographical location and meteorological conditions. It has obvious daily and annual cycle effects. In addition, there is a certain relationship between wind speed and wind direction. Figures 5 to 7 show the distribution relationship between wind speed and wind direction in a wind farm in the south.

It can be seen from Figures 5 to 7 that the wind direction changes more frequently in the low wind speed area, and the wind direction tends to be stable as the wind speed increases. In addition, the wind speed and wind direction of each place have obvious regional characteristics, and Table 1 shows the wind speed and wind direction characteristics of this region.

Table 1

It can be seen from Figures 5 to 7 and Table 1 that the wind speed is mainly 9-15m/s during this time period, accounting for 90.48%. The north wind direction is mainly concentrated at 280-330°, mainly northwest, which occupies 83.71% of the total. The average wind speed is 10.18m/s and the standard deviation of the wind speed is 4.02.

Using the previous data to analyze the actual operation results under the traditional yaw control strategy, the results are shown in Figure 8 to Figure 10. According to Figures 8 to 10, it can be seen that under the control of the traditional yaw strategy, the mean and standard deviation of the yaw error of the wind turbine will gradually decrease with the increase of the wind speed. In the wind speed area below 2.5m/s, the yaw system is not activated due to the small wind speed, so the yaw error in this area is large; in the low wind speed area below the rated value, the 2.5-4m/s The average value of the yaw error in the segment is relatively large, and then gradually stabilized; in the high wind speed area above the rated value, the average value of the yaw error is relatively stable.

Through analysis, the inventor found that the traditional yaw control algorithm has the following problems:

(1) The accuracy of the crew yaw to the wind is low. The existing control strategy is based on wind direction feedback control, which completely relies on the accuracy of wind direction measurement. However, the accuracy of the wind direction is not only related to the measurement accuracy of its own wind vane sensor, but also closely related to the installation position of the wind vane. This is because the rotation of the rotor of the wind turbine in the upwind direction will generate wake turbulence, which makes the wind vane in the downwind direction swing continuously, thereby reducing the accuracy of wind direction measurement and the service life of the wind measurement equipment, making the yaw control The system cannot get the ideal wind direction input signal, which leads to low wind accuracy of the unit.

(2) Yaw control lag. The yaw error used by the existing yaw control strategies is a calculated average value over a period of time, and the average value reflects the historical yaw state.

(3) The same control strategy is adopted for the entire wind speed area, which simply relies on the wind direction data without considering the wind speed. According to the above research, there is a certain relationship between wind speed and wind direction. The yaw strategy of most wind turbines in the field does not distinguish the wind speed, so that the tolerance range and delay time of the yaw error angle are fixed values.

(4) The adaptive level of the yaw system is very low. In the actual wind farm, the geographical location of the wind turbine also has a great influence on the yaw system, such as the influence of terrain and different positions. At present, the same control strategy is adopted for the units of different positions and even different wind farms, ignoring the differences in wind conditions of wind farms and the performance differences between units.

(5) It can be seen from Figure 8 to Figure 10 that for the existing yaw control strategy, although the yaw error is relatively stable in the high wind speed area, the average value of the yaw error still exceeds the set 8° .

It can be seen that the effect of the existing yaw control strategy is not satisfactory, so it is necessary to optimize the yaw control system.

The invention first provides a wind speed and wind direction prediction method, which can realize short-term independent prediction of wind speed and wind direction. Since the realization principle and realization process of the method for predicting the wind speed and the wind direction are the same, only the wind speed prediction is taken as an example for description here.

FIG. 11 shows a schematic diagram of the implementation flow of wind speed prediction in this embodiment.

FIG. 11 shows a schematic diagram of the implementation flow of independent prediction of wind speed in this embodiment.

As shown in FIG. 11 , in this embodiment, the method first acquires historical wind speed data in step S1101 . It should be pointed out that the historical wind speed data obtained in step S1101 of the method refer to preferably multiple moments included in a time period of a specific length (the length can be configured as different reasonable values according to actual needs). The average value of the wind speed within the preset time period (including the current moment) (for example, the average value of the wind speed within 10s, 30s, or 60s). The average value of the wind speed within 10s corresponding to the current time represents the average value of the wind speed within 10s before the current time.

Of course, in different embodiments of the present invention, the above-mentioned preset duration can be configured to different reasonable values according to actual needs (for example, a reasonable value within 5s to 240s, etc.) limited.

Since this method is based on the ARMA model for wind speed prediction, and the ARMA model requires the data to be stable, in order to ensure the stability of the data, after obtaining the historical wind speed data, the method will perform decompression on the historical wind speed data in step S1102. Trend processing to obtain detrended wind speed data.

Specifically, in this embodiment, in step S1102, the method preferably detrends the wind speed data at historical moments according to the following expression:

in, represents the detrended wind speed data at time t, represents the wind speed data value at time t, Indicates the historical wind speed trend value (ie, average).

In this embodiment, the average value of historical wind speed data Preferably, it refers to the average value of all wind speed data before the current time or the average value of the wind speed data within a certain period of time before the current time.

After completing one detrending process, the method will also perform stationarity detection on the detrended wind speed data in step S1102. Among them, if the detrended wind speed data is not stable, then the method will differentiate the detrended wind speed data again and perform the stationarity detection again until the obtained detrended wind speed data is stable.

Due to the non-stationarity of the wind speed signal, in order to use the time series method to predict it, it is necessary to change the wind speed signal into a stable random signal. In this embodiment, the method preferably adopts the method of citing an ordered difference operator (ie, ▽=1-B), and performs a first-order ordered difference transformation on the original non-stationary time series {y _t }. That is, there is:

▽y _t =(1-B)y _t =y _t -y _t-1 (8)

Among them, ▽y _t represents the difference between the data at time t (that is, the current time) and time t-1 (that is, the previous time), B represents the proportional coefficient of y _t and y t- ₁ , and y _t and y _t-1 Respectively represent the data at time t (ie the current time) and time t-1 (ie the previous time).

After d-order difference, we can get:

▽ ^d y _t = (1-B) ^d y _t (9)

Among them, ▽ ^d y _t represents the d-order difference operator.

The stationary series obtained after difference can be described by AR, MA and ARMA models, then the original time series can be expressed as:

in, represents the lag operator polynomial, θ(B) represents the prediction error lag operator polynomial, and a _t represents the prediction error.

This is the cumulative autoregressive-moving average model ARIMA(p, d, q).

If it is necessary to keep the data series stationary, then it is also necessary to require that the roots of the equations φ(B)=0 and θ(B)=0 are located outside the unit circle, that is, the modulus values of the roots are all greater than 1. in,

Among them, if the modulo values of the roots of the above equations are all greater than 1, then the wind speed series is stationary. If the stationary reversibility test fails, the difference order can be adjusted appropriately for correction until the adjusted wind speed sequence is stable.

Of course, in other embodiments of the present invention, the method may also adopt other reasonable ways to detect the stationarity of the detrended wind speed data, and the present invention is not limited thereto.

In this embodiment, by detrending the historical wind speed data, the method can also determine the difference order d in the ARMA model.

After the detrending process is completed, in step S1103 , the method determines the drag based on the autocorrelative function (ACF) and partial autocorrelative function (PACF) of the detrended wind speed data obtained in step S1102 Tail truncation mode.

Specifically, in this embodiment, the above-mentioned autocorrelation function and partial autocorrelation function can be respectively expressed as:

Among them, ρ _k represents the autocorrelation coefficient with the lag number k, and Respectively represent the detrended data at time i and the data time after detrend at time i+k, φ _kk represents the partial correlation coefficient with the lag number k, φ _{k-1, j} represents the k-1 order autoregressive process. The jth regression coefficient.

Specifically, in this embodiment, the method determines whether the autocorrelation function of the detrended wind speed data can remain zero after reaching a certain order. Wherein, if possible, the method can determine that the autocorrelation function of the detrended wind speed data has truncation; otherwise, it can determine that the autocorrelation function of the detrended wind speed data has tailing.

Similarly, the method can also determine whether the partial autocorrelation function of the detrended wind speed data can remain zero after reaching a certain order. If possible, the method can determine that the partial autocorrelation function of the detrended wind speed data has truncation; otherwise, it can determine that the partial autocorrelation function of the detrended wind speed data has tailing.

By judging whether the autocorrelation function and partial correlation function of the detrended wind speed data are smear type or truncated type, the method provided in this embodiment can also determine the smear and truncation mode of the detrended wind speed data.

FIG. 12 and FIG. 13 respectively show schematic diagrams of the autocorrelation function and the partial correlation function of the 10s average value of the wind speed sequence in this embodiment. It can be seen from Figure 12 and Figure 13 that both the autocorrelation function and the partial correlation function of the detrended wind speed data are tailing.

Again as shown in FIG. 11 , in this embodiment, after determining the smearing truncation mode of the trended wind speed data, the method will determine the ARMA model based on the determined smearing truncation mode in step S1104 . order to determine the automatic order, moving average order and difference order. The difference order is the number of times of the difference determined in the difference process in step S1102. If the wind speed data is relatively stable, then there is no need to perform differential processing during the detrending process, so the number of differentials (ie, the differential order d) is equal to zero.

After determining the order of automatic regression, the order of moving average and the order of difference in the ARMA model, in this embodiment, the method will use the automatic regression determined in step S1104 based on the ARMA model in step S1105 The order, moving average order and difference order can predict the wind speed data one step ahead according to the detrended wind speed data, so as to calculate the wind speed data at the next moment.

Specifically, in this embodiment, the method preferably determines the wind speed data at the next moment according to the following expression:

Among them, y _t+1 represents the data at time t+1 (ie the next time), y _t represents the data at time t (ie the current time), y _ti represents the data at time ti, δ represents a constant term, represents the i-th autoregressive coefficient, φ _j represents the j-th moving average coefficient, p represents the order of the automatic regression, q represents the order of the moving average, and e _t represents the error term at time t (ie the current time) (ie difference between the predicted value and the observed value at time t).

For wind speed data, there is:

in, Indicates the wind speed data at time t+1 (ie, the next time).

So far, the wind speed data at the next moment has been predicted based on the historical wind speed data.

Based on the same principle and process, the wind speed and direction prediction method provided by the present invention can also predict the wind direction data at the next moment according to the historical wind direction data.

The present invention also adopts a wind speed and direction prediction method, which uses the wind direction circular transformation to predict the next moment wind direction data when the ARMA model is used to determine the next moment wind speed data according to the historical wind speed data.

FIG. 14 shows a schematic flowchart of the implementation of the wind speed and direction prediction method provided in this embodiment.

As shown in FIG. 14 , in this embodiment, the method acquires historical wind speed data and wind direction data in step S1401 . It should be noted that the historical wind speed data obtained by the method in step S1101 preferably refers to the average value of wind speed (for example, 10s, 30s or Average wind speed within 60s). The average value of the wind speed within 10s corresponding to the current time represents the average value of the wind speed within 10s before the current time.

Of course, in different embodiments of the present invention, the above-mentioned preset duration can be configured to different reasonable values according to actual needs (for example, a reasonable value within 5s to 240s, etc.) limited.

In step S1402, the method uses the ARMA model to predict the wind speed data at the next moment according to the historical wind speed data. In this embodiment, the method uses the ARMA model to predict the wind speed data at the next moment according to the historical wind speed data. The content is repeated.

The wind direction is a circular variable, so the method provided in this embodiment uses a prediction method more suitable for circular variables to predict the wind direction data at the next moment. Specifically, in this embodiment, the method performs circular variable transformation on the historical wind direction data in step S1403, so as to obtain the sine and cosine values of the historical wind direction data.

Specifically, the method preferably transforms the historical wind direction data according to the following expression:

in, and represent the sine and cosine values of the wind direction data at time t, respectively, Represents the wind direction data at time t.

Based on Expression (17), the method can obtain the sine and cosine values of the wind direction data at the current moment and at each moment before the current moment.

After determining the sine and cosine values of the wind direction data at the current moment (ie, time t), the method determines the next moment (ie, t+1) according to the sine and cosine values of the wind direction data at the current moment in step S1404 time) of the sine of the wind direction data and cosine

Specifically, in this embodiment, the method preferably uses the ARMA model to determine the sine value of the wind direction data at the next moment according to the sine value and the cosine value of the historical wind direction data respectively. and cosine The specific principles and processes thereof are the same as those described in FIG. 11 , and thus the detailed description of the content will not be repeated here.

As shown in FIG. 14 , in this embodiment, the sine value of the wind direction data at the next moment is obtained and cosine After that, the method will in step S1405 according to the sine value of the wind direction data at a moment and cosine Determine the wind direction data at the next moment

Specifically, in this embodiment, the method preferably determines the wind direction data at the next moment according to the following expression

in, Represents the wind direction data at time t+1 (ie, the next time), and Represent the sine and cosine values of the wind direction data at time t+1, respectively.

It should be pointed out that, in other embodiments of the present invention, the method may also adopt other reasonable ways to calculate the sine value of the wind direction data at the next moment according to the prediction and cosine Determine the wind direction data at the next moment

It should be pointed out that, in other embodiments of the present invention, the prediction of wind speed data can be configured according to actual needs, that is, wind speed data is acquired and predicted when necessary, and not acquired when not required. At the same time, the wind speed data does not predict the wind speed data, and the present invention is not limited thereto. In addition, in other embodiments of the present invention, according to actual needs, the method may also use other reasonable ways to predict the wind speed data, and the present invention is also not limited thereto.

The invention provides a new wind speed and direction prediction method, which regards the wind speed and the wind direction as a vector, and gives the wind vector to predict the wind speed data and the wind direction data at the next moment.

FIG. 15 shows a schematic flowchart of the implementation of the wind speed and direction prediction method provided in this embodiment.

As shown in FIG. 15 , in this embodiment, the wind speed and direction prediction method acquires historical wind speed data and historical wind direction data of the area to be analyzed in step S1501 . It should be noted that the historical wind speed data obtained by the method in step S1101 preferably refers to the average value of wind speed (for example, 10s, 30s or Average wind speed within 60s). The average value of the wind speed within 10s corresponding to the current time represents the average value of the wind speed within 10s before the current time.

Of course, in different embodiments of the present invention, the above-mentioned preset duration can be configured to different reasonable values according to actual needs (for example, a reasonable value within 5s to 240s, etc.) limited.

After obtaining the historical wind speed data and wind direction data, the method will decompose the wind vector according to the historical wind speed data and the historical wind direction data in step S1502, thereby obtaining the historical wind vector abscissa and the wind vector ordinate.

Specifically, in this embodiment, the method preferably decomposes the wind vector according to the following expression:

in, and respectively represent the wind vector abscissa data and the wind vector ordinate data at time t, represents the wind speed data, Represents the wind direction data at time t.

Based on expression (19), the method can obtain the wind vector abscissa data and wind vector ordinate data at the current moment and at each moment before the current moment.

Of course, in other embodiments of the present invention, the method may also adopt other reasonable ways to decompose the wind vector, and the present invention is not limited thereto.

After obtaining the historical wind vector abscissa and the wind vector ordinate, the method will use the ARMA model in step S1503 to determine the wind vector abscissa at the next moment according to the historical wind vector abscissa and the historical wind vector ordinate Zephyr vector ordinate

In this embodiment, the method uses the ARMA model to determine the abscissa of the wind vector at the next moment Zephyr vector ordinate The specific principle and process are similar to those shown in Figure 11 above. On the basis of the method shown in Figure 11, the historical wind speed data is replaced with the historical wind vector abscissa and the historical wind vector ordinate, and the next Wind vector abscissa of moment Zephyr vector ordinate The process will not be repeated here.

As shown in FIG. 15 , in this embodiment, in step S1504 , the method uses the abscissa of the wind vector at the next moment obtained in step S1503 Zephyr vector ordinate Determine the wind speed data and wind direction data at the next moment.

Specifically, in this embodiment, the method preferably determines the wind speed data at the next moment according to the following expression

Determine the wind direction data at the next moment according to the following expression:

in, Indicates the wind direction data at time t+1 (ie, the next time).

It should be pointed out that in other embodiments of the present invention, the method may also adopt other reasonable ways to Zephyr vector ordinate The wind speed data and wind direction data at the next moment are determined, and the present invention is not limited to this.

In order to verify the effectiveness and advantages of the wind speed and direction prediction method provided by the present invention, this embodiment uses the wind speed and wind direction data recorded by the SCADA (Supervisory Control and Data Acquisition System) of a wind farm in the south within 24 hours, a total of 86,400 point. Among them, the original wind direction and the average wind direction under different time periods are shown in Figure 16, the original wind speed and the average wind speed under different time periods are shown in Figure 17, and Figure 18 and Figure 19 respectively show the 10s wind direction prediction obtained by different prediction methods. Results and wind speed prediction results, Figure 20 and Figure 21 respectively show the 30s wind direction prediction results and wind speed prediction results obtained by different prediction methods, Figure 22 and Figure 23 respectively show the 60s wind direction prediction results obtained by different prediction methods and Wind speed forecast results.

According to Figure 16 and Figure 17, since the original wind speed data and wind direction original data fluctuate greatly, calculating the average value of 10s, 30s and 60s is beneficial to filtering and reducing the influence of outliers. Also, when the wind vane is measuring more than 360° in wind direction, the value will start at 0. As a result, the wind direction shown in Figure 16 fluctuates greatly during the 20-22H period, and the practical ARMA model alone predicts the wind direction with low accuracy (especially the circles in Figure 18, Figure 20 and Figure 22). The prediction of the wind direction by the deformation variable method can better reflect the continuity of the wind direction, and the sudden change at the above circle will not be out of the line, so that the wind direction prediction accuracy is higher. As for wind speed prediction, it can be seen from Figure 19, Figure 21 and Figure 23 that the results obtained by the separate prediction method are more stable than the original data.

In order to evaluate the accuracy of the proposed short-term wind speed and wind direction prediction method, in this embodiment, three expressions of the mean absolute error (MAE), the mean absolute percent error (MAPE) and the mean square error (MSE) can be used. Compared with the predicted results, the statistical results are shown in Table 2. Among them, the calculation expressions of mean absolute error (MAE), mean absolute percent error (MAPE) and mean square error (MSE) are:

Among them, N represents the number of data, x _i represents the actual value, represents the predicted value.

Table 2

Combining Figures 16 to 23 and Table 2, it can be seen that for wind direction prediction, the accuracy of the wind direction prediction results obtained by using the ARMA prediction model alone is lower than that of the wind vector method and the circular variable method.

It can be seen from the above descriptions that, compared with the existing wind speed and direction prediction methods, the method provided by the present invention can make the wind direction prediction result more accurate and stable, thus providing a data basis for the yaw control of the wind turbine. .

FIG. 24 shows a schematic flowchart of the implementation of the yaw control method for the wind turbine provided by this embodiment.

As shown in FIG. 24 , in this embodiment, the yaw control method first calculates the average value of the wind speed and the average value of the wind direction within a preset time period according to the obtained wind speed and wind direction in step S2401, thereby obtaining historical wind speed data and wind direction data.

In this embodiment, the above-mentioned preset duration is preferably 10s, 30s and/or 60s. Of course, in different embodiments of the present invention, the above-mentioned preset duration can be configured to different reasonable values according to actual needs (for example, a reasonable value within 5s to 240s, etc.) limited.

After obtaining the historical wind speed data and wind direction data, the method predicts the wind speed data and wind direction data at the next moment according to the historical wind speed data and wind direction data in step S2402. In this embodiment, the method preferably adopts the circular variable method to predict the wind speed data and wind direction data at the next moment, wherein the specific principle and process of wind speed data and wind direction data prediction based on the circular variable method have been described in the above content. Detailed description, so the content of this part will not be repeated here.

Of course, in other embodiments of the present invention, according to the actual situation, the method can also use other reasonable ways to predict the wind speed data and wind direction data at the next moment according to the historical wind speed data and wind direction data, and the present invention is not limited to this. For example, in an embodiment of the present invention, the method can also obtain the wind speed data and wind direction data at the next moment by using the method of respectively predicting the wind speed data and the wind direction data as shown in FIG. 15 to obtain the wind speed data and wind direction data at the next moment.

As shown in FIG. 24 , in this embodiment, after the wind speed data and wind direction data at the next moment are determined, the method will determine the control parameters according to the predicted wind speed data at the next moment in step S2403, and then In step S2404, the yaw control is performed on the wind light-emitting point unit according to the control parameters determined in step S2403 and the wind direction book at the next moment determined in step S2402.

In this embodiment, the method preferably determines the wind speed interval to which the predicted wind speed data at the next moment belongs in step S2403, and determines the control parameters according to the wind speed interval to which it belongs.

Specifically, as shown in FIG. 25 , in this embodiment, the method divides the controllable wind speed area of the wind turbine into 4 sections with cut-in wind speed v _{cut_in} , rated wind speed v _n and cut-out wind speed v _{cut_out} as boundaries.

Preferably, if the wind speed v _w <v _{cut_in} (for example, v _w <2.5m/s) measured by the relevant sensor, this also means that the wind speed v _w (ie, the predicted wind speed data at the next moment) is at the cut-in wind speed v _{cut_in} the following area. Since this area contains less wind energy, the method preferably controls the wind turbine to be in a shutdown state in this wind speed area at this time.

Preferably, if the wind speed v _w >v _{cut_out} (eg v _w >25m/s), it means that the wind speed v _w (ie the predicted wind speed data at the next moment) is in the region above the cut-out wind speed v _{cut_out} . Since the wind speed in this area is relatively high, the excessive wind speed has a great influence on the load of the wind turbine, thereby affecting the safety, reliability and life of the wind turbine. Therefore, in this embodiment, the method preferably controls the wind turbine. Yaw to downwind and stop in this wind speed area.

Preferably, if the wind speed v _w is greater than or equal to the preset cut-in wind speed and less than the first preset wind speed threshold, ie there is v _{cut_in} ≤v _w <v ₁ , then the method will keep the original control parameters unchanged. And if the wind speed v _w is greater than or equal to the first preset wind speed threshold and less than the preset cut-out wind speed, that is, there is v ₁ ≤ v _w <v _{cut_out} , then the method will reduce the original control parameters by a specific value to obtain the required control parameters. It should be noted that, in different embodiments of the present invention, the above-mentioned first preset wind speed threshold may be configured as different reasonable values according to actual wind energy conditions, and the present invention is not limited thereto. At the same time, it should also be pointed out that in other embodiments of the present invention, when v _{cut_in} ≤v _w <v ₁ or v ₁ ≤v _w <v _{cut_out} , the method may also use other reasonable ways to configure the control parameters, The present invention is also not limited to this.

For example, if the first preset wind speed threshold is configured as 4m/s, since the wind energy in the wind speed interval [2.5m/s, 4m/s) accounts for 9.64% of the total wind energy, the mean value and standard deviation of the wind error are relatively large. , and the wind direction in this area is relatively unstable. At the same time, since the energy obtained by the wind turbine from the wind in the wind speed range is relatively small, in this embodiment, the method preferably controls the delay time T _set and/or the yaw start error angle v _set , etc. The parameters keep the original control parameters (ie, the control parameters set according to the existing yaw control method) unchanged.

For the wind speed range [4m/s, 25m/s), in this embodiment, the method preferably reduces the original control parameters by a specific value, so as to obtain new control parameters suitable for the wind speed range. The partial air control method can also control the wind generator set according to the determined new control parameters.

In this embodiment, several wind speed intervals are included between the first preset wind speed threshold and the preset rated wind speed. For these wind speed intervals, the larger the wind speed, the smaller the control parameter corresponding to the wind speed interval.

For example, if 4m/ _s≤vw <9m/s, since the wind energy contained in this wind speed range occupies 13.61% of the total wind energy and is in the low wind speed section below the rated wind speed, the traditional yaw control under the The mean value and standard deviation of the errors are small, and under the action of the traditional yaw controller, it can be improved compared with the [2.5m/s, 4m/s) wind speed range, but the yaw control performance still needs to be improved. Therefore, the method provided in this embodiment also reduces the original control parameter values of control parameters such as the delay time T _set and/or the yaw start error angle v _set by a specific value, so that the yaw control performance can meet the requirements of this requirements for the wind speed range.

If 9m/ _s≤vw <12m/s, since the wind energy contained in this wind speed range occupies 33.07% of the total wind energy and is in the middle and high wind speed section below the rated wind speed, the average value of the wind error under traditional yaw control and the standard deviation will continue to be smaller than the previous wind speed range, but the yaw control performance still needs to be improved. Therefore, in this embodiment, the method will continue to reduce the original control parameter values of the control parameters such as the delay time T _set and/or the yaw start error angle v _set , so that the yaw control performance can meet the wind speed range. Require.

In this embodiment, if the wind speed v _w is greater than or equal to the preset cut-in wind speed and less than the rated wind speed, the method preferably adjusts the tip speed ratio of the wind turbine to achieve the tracking of the optimal power curve and the capture of the maximum wind energy as the target. At this time, the pitch angle is preferably set to 0°. Of course, in other embodiments of the present invention, the pitch angle may also be configured to other reasonable values according to actual needs, and the present invention is not limited thereto.

In this embodiment, if the wind speed data is greater than or equal to the preset rated wind speed and less than the preset cut-out wind speed, then the method will perform yaw control on the wind turbine according to the wind direction data at the next moment to make the wind turbine The yaw error of is within the preset error range. Specifically, in this embodiment, if the wind speed data is greater than or equal to the preset rated wind speed and less than the preset cut-out wind speed, the method will adjust the pitch angle to change the wind energy acquisition coefficient to obtain stable output power to protect the unit equipment.

For example, if 12m/ _s≤vw <25m/s, the wind energy contained in this wind speed range will occupy 40.32% of the total wind energy, and the wind direction in this wind speed range will steadily increase. Since the wind turbines above the rated wind speed need to maintain the rated output power, although the yaw error in this wind speed range does not affect the power generation, the excessive yaw error will affect the overall load of the wind turbine, resulting in the change of the average induced wind speed The amplitude is too large, so in this embodiment, the method will configure the yaw start error angle as [-8°, -8°], so that the yaw of the wind turbine can be controlled by the yaw control of the wind turbine. Navigation error remains at [-8°,-8°].

It can be seen that, for each wind speed interval included in the cut-in wind speed to the cut-off wind speed, the method provided in this embodiment preferably sets the control parameters corresponding to each wind speed interval separately, and the specific setting results can be as shown in Table 3. Show.

table 3

For the yaw control method of the wind turbine provided in this embodiment, the wind speed data (for example, 10s, 30s and/or 60s average wind speed) and wind direction data (for example, 10s, 30s and / or 60s average value of wind direction) can be predicted one step in advance, and then the operation of the yaw system is controlled by judging the predicted wind speed data and wind direction data with the corresponding threshold.

In order to verify the effectiveness of the partition control strategy based on wind speed and wind direction prediction, this embodiment uses the wind speed data and wind direction data as shown in Figure 16 to Figure 23, and uses the traditional control strategy and the method proposed by the present invention in the Matlab/Simulink environment. The regional control strategy based on wind speed and wind direction prediction is controlled, and the experimental results are analyzed. In addition, in order to clearly express the effect of the partition strategy, this embodiment analyzes from five aspects, namely, the average yaw error, the root mean square of the yaw error, the yaw time, the number of yaw, and the power loss coefficient.

Among them, the mean value of yaw error is calculated by the following expression:

The root mean square of the yaw error is calculated by the following expression:

The yaw time is calculated using the following expression:

The number of yaw is calculated by the following expression:

In practical engineering experience, the following expression is commonly used to calculate the power loss coefficient

Among them, θ _ye represents, N represents the number of yaw errors, t _yaw represents the yaw time, represents, C _yaw represents the number of yaw, ξ represents the power loss coefficient, P _red represents the reduced power, P _real represents the ideal output power, represents the equivalent yaw error.

Equivalent yaw error It can be calculated according to the following expression:

in, is the mean value of the error in the j-th yaw error region, which characterizes the probability of this yaw error region.

Figure 26 shows the nacelle position under the conventional control strategy and the zoned control strategy provided by the present invention. The results of Fig. 26 are counted according to the wind speed zones, and the yaw error distribution diagrams shown in Fig. 27 to Fig. 30 are obtained.

Table 4 shows the statistics under different yaw control methods.

Table 4

Combining the statistical results in Figures 26 to 30 and Table 4, it can be seen that in the low wind speed range (eg [2.5m/s, 4m/s)), due to the yaw control strategy adopted by the yaw control method provided in this embodiment Consistent with the traditional strategy, so the yaw error distribution is unchanged.

In the middle and low wind speed range below the rated wind speed (for example [4m/s, 9m/s)), the yaw error obtained by the yaw control method provided in this embodiment is reduced compared with the traditional method, and the wind accuracy Also higher, the yaw error increases from 75.20% to 76.04% in the [-8°, -8°] interval.

In the middle and high wind speed area below the rated wind speed (eg [9m/s, 12m/s)), the yaw error obtained by the yaw control method provided in this embodiment is significantly reduced compared with the traditional method, and the yaw error is within [-8 °,-8°] range increased from 81.75% to 82.62%.

In the high wind speed area above the rated wind speed (eg [12m/s, 25m/s)), the yaw error obtained by the yaw control method provided in this embodiment is significantly reduced compared with the traditional method, and the yaw error is within [-8 °,-8°] range is increased from 83.83% to 84.79%, and the error distribution and yaw error distribution are more concentrated.

Compared with the traditional yaw control method, the yaw number of the yaw control method provided by the present invention is improved compared with the traditional control strategy, but the increased number of times is mainly concentrated in the middle and high wind speed area, so the power loss coefficient is significantly reduced.

It can be seen from this that the partitioned predictive control method provided by the present invention can effectively reduce the yaw error in the middle and high wind speed regions, thereby reducing the power loss coefficient (ie, improving the utilization rate of wind energy).

It is to be understood that the disclosed embodiments of the present invention are not limited to the specific structures or process steps disclosed herein, but should extend to equivalents of these features as understood by those of ordinary skill in the relevant art. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not meant to be limiting.

Reference in the specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases "one embodiment" or "an embodiment" in various places throughout the specification are not necessarily all referring to the same embodiment.

While the above examples serve to illustrate the principles of the invention in one or more applications, it will be apparent to those skilled in the art that the invention can be made in form, usage, and implementation without departing from the principles and spirit of the invention. Various modifications can be made to the details without creative labor. Accordingly, the invention is defined by the appended claims.

Claims (15)

1. A yaw control method for a wind turbine, wherein the yaw control method comprises: Step 1. Calculate the average value of the wind speed and the average value of the wind direction within the preset time length according to the obtained wind speed and wind direction, obtain historical wind speed data and historical wind direction data, and predict the wind speed at the next moment according to the historical wind speed data and historical wind direction data. data and wind direction data; Step 2: Determine control parameters according to the wind speed data at the next moment, and use the control parameters and wind direction data to perform yaw control on the wind turbine; Wherein, in the step 1, the steps of predicting the wind speed data and wind direction data at the next moment include: Decompose the wind vector according to the historical wind speed data and the historical wind direction data to obtain historical wind vector abscissa data and historical wind vector ordinate data; Utilize the ARMA model to determine the wind vector abscissa data and the wind vector ordinate data at the next moment according to the historical wind vector abscissa data and the historical wind vector ordinate data; According to the wind vector abscissa data and wind vector ordinate data at the next moment, respectively determine the wind speed data and wind direction data at the next moment; or, Perform circular variable transformation on the historical wind direction data to obtain the sine and cosine values of the historical wind direction data; Use the ARMA model to determine the sine and cosine values of the wind direction data at the next moment according to the sine and cosine values of the historical wind direction data, and determine the wind direction data at the next moment according to the sine and cosine values of the wind direction data at the next moment. . 2 . The method of claim 1 , wherein, in the step 1, the preset duration is 10s, 30s or 60s. 3 . 3. The method of claim 1, wherein the wind vector is decomposed according to the following expression: in, and Represent the wind vector abscissa data and wind vector ordinate data at time t, respectively, represents the wind speed data, Represents the wind direction data at time t. 4. The method of claim 1, wherein the wind speed data at the next moment is determined according to the following expression: in, represents the wind speed data at time t+1, and Represent the wind vector abscissa data and wind vector ordinate data at time t+1, respectively. 5. The method of claim 1, wherein the wind direction data at the next moment is determined according to the following expression: in, represents the wind direction data at time t+1, and Represent the wind vector abscissa data and wind vector ordinate data at time t+1, respectively. 6. The method of claim 1, wherein a circular variable transformation is performed on the historical wind direction data according to the following expression: in, and respectively represent the sine and cosine values of the wind direction data at time t, Represents the wind direction data at time t. 7. The method of claim 1, wherein the wind direction data at the next moment is determined according to the following expression: in, represents the wind direction data at time t+1, and Represent the sine and cosine values of the wind direction data at time t+1, respectively. 8 . The method according to claim 6 , wherein, in the first step, wind speed data at the next moment is determined according to historical wind speed data by using an ARMA model. 9 . 9. The method of claim 8, wherein the step of determining the wind speed data at the next moment comprises: Step a, performing detrending processing on the historical wind speed data to obtain detrending wind speed data; Step b, according to the autocorrelation function and the partial autocorrelation function of the detrended wind speed data, determine the tail truncation mode; Step c, based on the smear truncation mode, utilize preset criteria to determine the order of the ARMA model, and determine the automatic regression order, the moving average order and the difference order; Step d, based on the ARMA model, using the automatic regression order, the moving average order and the difference order to calculate the wind speed data at the next moment according to the detrended wind speed data. 10. The method according to any one of claims 1 to 9, wherein in the second step, a wind speed interval to which the wind speed data at the next moment belongs is determined, and the control is determined according to the wind speed interval to which it belongs. parameter. 11. The method of claim 10, wherein in the second step, if the wind speed data at the next moment is less than the preset cut-in wind speed, the wind turbine is controlled to be in a shutdown state. 12. The method according to claim 10, wherein in the second step, if the wind direction data at the next moment is greater than or equal to the preset cut-out wind speed, the wind turbine is controlled to yaw to the downwind direction position and in a shutdown state. 13. The method of claim 10, wherein in the second step, If the wind speed data at the next moment is greater than or equal to the preset cut-in wind speed and less than the first preset wind speed threshold, keeping the original control parameters unchanged; And/or, if the wind speed data at the next moment is greater than or equal to the first preset wind speed threshold and less than the preset cut-out wind speed, the original control parameter is reduced by a specific value to obtain the required control parameter. 14. The method of claim 13, wherein in the second step, several wind speed intervals are included between the first preset wind speed threshold and the preset rated wind speed, wherein for these wind speed intervals In other words, the larger the wind speed, the smaller the control parameter corresponding to the wind speed interval. 15. The method according to claim 10, wherein in the second step, if the wind speed data at the next moment is greater than or equal to a preset rated wind speed and less than a preset cut-out wind speed, according to the The wind direction data at the next moment performs yaw control on the wind turbine so that the yaw error of the wind turbine is within a preset error range.