A method and device for yaw optimization analysis of wind turbine generator set

Abstract

The present invention discloses a method and device for optimizing the yaw analysis of a wind turbine generator set, belonging to the technical field of wind power generation. The method performs adaptive abnormality identification and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting on the original data of a wind farm; data post-processing includes calculating the representative wind direction change degree threshold and wind speed interval duration threshold in each wind speed interval based on probability distribution; calculating the optimal global wind direction change degree threshold and global wind speed interval duration threshold by weighted average method; optimizing the yaw strategy of the wind turbine set in the wind farm by the optimal global wind direction change degree threshold and global wind speed interval duration threshold; feeding back the optimized yaw strategy of the wind turbine in the wind farm to the wind turbine set in the wind farm, and controlling the wind turbine set in the wind farm to terminate the yaw angle execution action outside the optimal yaw interval. The present invention reduces the mechanical wear of the yaw system, reduces energy loss, can cope with variable wind conditions, and has high power generation efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of wind power generation, and in particular relates to a method and a device for yaw optimization analysis of a wind generator set. Background Art

[0002] The yaw system of a wind turbine is a key part of the operation of the wind turbine. Its main function is to ensure that the blades of the wind turbine are always facing the incoming wind direction to maximize the efficiency of wind energy capture. The yaw system adjusts the airport direction of the wind turbine according to the wind direction data measured by the wind direction sensor so that the wind rotor always keeps aligned with the wind direction. Through precise yaw control, wind turbines can effectively improve power generation efficiency while reducing the impact of lateral forces on the wind rotor. At the same time, wind is a vector, including both wind speed and wind direction. Both of these parameter variables are uncertain, with strong randomness and intermittency. Generally, the greater the wind speed, the smaller the wind direction fluctuation; the smaller the wind speed, the more violent the wind direction fluctuation. Traditional yaw control strategies are difficult to cope with this complexity. Different wind speed ranges may require different yaw thresholds to more reasonably reflect the trend of wind direction changes.

[0003] In the prior art, the yaw control system of the wind turbine usually has the problem of too frequent or too delayed yaw action due to the use of a fixed wind direction deviation threshold. This unreasonable adjustment frequency will not only increase the mechanical wear of the yaw system, but also cause energy loss. It may also cause cable entanglement problems and increase maintenance costs. The traditional yaw control method fails to fully consider the differences in wind direction changes in different wind speed ranges. Due to the complex relationship between wind speed and wind direction, the wind direction fluctuation range in each wind speed range is not the same, and the fixed threshold cannot adapt to all situations. For example: in the low wind speed range, the wind direction fluctuates greatly, but too high a yaw frequency will reduce the efficiency of the wind turbine; in the high wind speed range, the wind direction is relatively stable, but too high a threshold setting may cause a delayed response.

[0004] In addition, traditional technologies rely on fixed yaw thresholds, and the yaw system often lacks a dynamic adjustment mechanism, that is, when wind speed and wind direction change frequently, the yaw threshold should be randomly and dynamically adjusted, which cannot cope with changing wind conditions. Especially in environments with high wind turbulence intensity, the yaw system of the wind turbine is prone to unnecessary frequent adjustments, and even affects the long-term reliability of the generator set. The setting of the yaw threshold directly affects the energy capture efficiency of the wind turbine. If the yaw system reacts too frequently, the wind turbine may lose some wind energy due to frequent adjustments and reduce power generation efficiency; conversely, if the reaction is too slow, the wind rotor cannot be aligned with the incoming wind direction in time, which will also cause energy loss. How to optimize the yaw adjustment of the wind turbine without increasing mechanical wear and improve the overall power generation efficiency is still a difficult problem in the existing technology. Summary of the invention

[0005] To this end, the present invention provides a method and device for optimizing the yaw analysis of a wind turbine generator set, so as to solve the problems in the prior art that, due to the use of a fixed wind direction deviation threshold, the yaw action is prone to being too frequent or too delayed, increasing the mechanical wear of the yaw system, causing energy loss, possibly causing cable entanglement, being unable to cope with changeable wind conditions, and having low power generation efficiency.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a method for optimizing the yaw of a wind turbine generator set, comprising:

[0007] Acquiring original data of the wind farm, wherein the original data of the wind farm includes original wind speed data and original wind direction data;

[0008] Performing data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting;

[0009] The raw data of the wind farm after the data pre-processing is subjected to data post-processing, wherein the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; after obtaining the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval, calculating the optimal global wind direction change degree threshold and the global wind speed interval duration threshold by weighted average method;

[0010] The yaw strategy of wind turbines in wind farms is optimized through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold.

[0011] The optimized yaw strategy of the wind turbines in the wind farm is fed back to the wind turbine group in the wind farm, so as to control the wind turbine group in the wind farm to terminate the execution of the yaw angle outside the optimal yaw range, so that the wind turbine group can operate in the best windward state.

[0012] As a preferred solution of the yaw optimization analysis method for wind turbine generator sets, during the process of performing the adaptive anomaly identification and interpolation completion on the original wind speed data:

[0013] According to the local mean of wind speed and wind speed standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind speed data If the wind speed data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind speed data before and after and The mean of is interpolated and completed;

[0014] In the process of performing the adaptive anomaly identification and interpolation completion on the original wind direction data:

[0015] According to the average wind direction and wind direction standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind direction data If the wind direction data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind direction data and The mean value of is interpolated.

[0016] As a preferred solution of the yaw optimization analysis method for wind turbine generator sets, the formula for performing the adaptive time synchronization and filtering on the original wind speed data is:

[0017] ;

[0018] In the formula, is the wind speed data after filtering; It is the wind speed data that has been filtered at the last moment; is the wind speed sampling time interval; is the time adjustment item; Control the weight distribution between new wind speed data and historical data, Used to correct the wind speed sampling time difference, Through acceleration Dynamically control the sensitivity of the wind speed filter;

[0019] The formula for performing the adaptive time synchronization and filtering on the original wind direction data is:

[0020] ;

[0021] In the formula, is the wind direction data after filtering; It is the wind direction data that has been filtered at the last moment; Used to correct the difference in wind direction sampling time; is the wind direction sampling time interval.

[0022] As a preferred solution of the yaw optimization analysis method for wind turbine generator sets, during the nonlinear normalization and distribution fitting process of the original wind speed data:

[0023] Logarithmic normalization is used to convert the wind speed into a standardized form. The normalization formula is:

[0024] ;

[0025] In the formula, is the normalized wind speed data; is the minimum wind speed data; is the maximum wind speed data;

[0026] Assuming that the wind speed conforms to the log-normal distribution, the probability density of the wind speed is expressed by the distribution function :

[0027] ;

[0028] In the formula, is the wind speed; is the standard deviation of wind speed; is the log mean of the lognormal distribution;

[0029] In the process of performing the nonlinear normalization and distribution fitting on the original wind direction data:

[0030] Normalize the wind direction data using the cosine function:

[0031] ;

[0032] In the formula, is the normalized wind direction data; is the original wind direction data, indicating the current wind direction; is the reference wind direction, which is the relative angle when the reference value is used for normalization;

[0033] Use von Mises to model the distribution of wind direction data:

[0034] ;

[0035] In the formula, To control the concentration of the distribution, represents the average wind direction. is the modified Bessel function.

[0036] As a preferred solution for the yaw optimization analysis method of wind turbines, the formula for calculating the threshold value of representative wind direction change in each wind speed interval based on probability distribution is:

[0037] ;

[0038] In the formula, For wind speed data in the interval Inside, wind direction change angle The cumulative probability of The code indicating the wind speed range; The interval wind speed data Probability density function of internal wind direction change; 0.6 is the set cumulative probability threshold;

[0039] The formula for calculating the wind speed interval duration threshold based on probability distribution is:

[0040] ;

[0041] In the formula, It is the wind speed data in the interval Duration The cumulative probability of is the probability density function of the duration of wind direction change within the wind speed; 0.6 is the set cumulative probability threshold.

[0042] As the preferred solution for the wind turbine yaw optimization analysis method, the formula for calculating the optimal global wind direction change degree threshold by weighted average method is:

[0043] ;

[0044] In the formula, is the global wind direction change degree threshold; It is the correction factor adjusted according to the characteristics of wind resources within the wind speed range; It is the interval wind speed data The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval;

[0045] The formula for calculating the global duration threshold by weighted average method is:

[0046] ;

[0047] In the formula, is the global wind direction change degree threshold.

[0048] As a preferred solution for the yaw optimization analysis method of wind turbines, the formula for optimizing the yaw strategy of wind turbines in wind farms through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is:

[0049] ;

[0050] In the formula, is the optimized yaw strategy, , , is the change in power loss, which is a constant parameter adjusted according to the target wind farm and wind turbine characteristics; It's the wind speed.

[0051] The present invention also provides a wind turbine generator set yaw optimization analysis device, comprising:

[0052] A wind farm original data acquisition module is used to acquire the original data of the wind farm, wherein the original data of the wind farm includes original wind speed data and original wind direction data;

[0053] A data pre-processing module, used for performing data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting;

[0054] A data post-processing module is used to perform data post-processing on the raw data of the wind farm after the data pre-processing, wherein the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; after obtaining the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval, calculating the optimal global wind direction change degree threshold and the global wind speed interval duration threshold by weighted average method;

[0055] A yaw strategy optimization module is used to optimize the yaw strategy of the wind turbine group in the wind farm through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold;

[0056] The yaw strategy feedback control module is used to feed back the optimized yaw strategy of the wind farm wind turbine to the wind farm wind turbine group, control the wind farm wind turbine group to terminate the yaw angle execution action outside the optimal yaw range, and make the wind turbine group operate in the best windward state.

[0057] As a preferred solution of the yaw optimization analysis device for wind turbine generator set, in the data pre-processing module:

[0058] According to the local mean of wind speed and wind speed standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind speed data If the wind speed data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind speed data before and after and The mean value of is interpolated.

[0059] As a preferred solution of the yaw optimization analysis device for wind turbine generator set, in the data pre-processing module:

[0060] According to the average wind direction and wind direction standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind direction data If the wind direction data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind direction data and The mean value of is interpolated.

[0061] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the data pre-processing module, the formula for performing the adaptive time synchronization and filtering on the original wind speed data is:

[0062] ;

[0063] In the formula, is the wind speed data after filtering; It is the wind speed data that has been filtered at the last moment; is the wind speed sampling time interval; is the time adjustment item; Control the weight distribution between new wind speed data and historical data, Used to correct the wind speed sampling time difference, Through acceleration Dynamically control the sensitivity of the wind speed filter.

[0064] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the data pre-processing module, the formula for performing the adaptive time synchronization and filtering on the original wind direction data is:

[0065] ;

[0066] In the formula, is the wind direction data after filtering; It is the wind direction data that has been filtered at the last moment; To control the weight distribution between new data and historical data of wind direction; Used to correct the difference in wind direction sampling time; is the wind direction sampling time interval;

[0067] In the data pre-processing module, logarithmic normalization is used to convert the wind speed into a standardized form. The normalization formula is:

[0068] ;

[0069] In the formula, is the normalized wind speed data; is the minimum wind speed data; is the maximum wind speed data;

[0070] In the data pre-processing module, it is assumed that the wind speed conforms to the log-normal distribution and the probability density of the wind speed is expressed by the distribution function :

[0071] ;

[0072] In the formula, is the wind speed; is the standard deviation of wind speed; is the log mean of the lognormal distribution.

[0073] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the data pre-processing module, the cosine function is used to normalize the wind direction data:

[0074] ;

[0075] In the formula, is the normalized wind direction data; is the original wind direction data, indicating the current wind direction; is the reference wind direction, which is the relative angle when the reference value is used for normalization;

[0076] In the data pre-processing module, von Mises is used to perform distribution modeling on wind direction data:

[0077] ;

[0078] In the formula, To control the concentration of the distribution, represents the average wind direction. is the modified Bessel function.

[0079] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the data post-processing module, the formula for calculating the representative wind direction change degree threshold in each wind speed interval based on probability distribution is:

[0080] ;

[0081] In the formula, For wind speed data in the interval Inside, wind direction change angle The cumulative probability of The code indicating the wind speed range; For interval wind speed data Probability density function of internal wind direction change; 0.6 is the set cumulative probability threshold;

[0082] In the data post-processing module, the formula for calculating the wind speed interval duration threshold based on probability distribution is:

[0083] ;

[0084] In the formula, It is the wind speed data in the interval Duration The cumulative probability of is the probability density function of the duration of wind direction change within the wind speed; 0.6 is the set cumulative probability threshold.

[0085] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the data post-processing module, the formula for calculating the optimal global wind direction change degree threshold by weighted average method is:

[0086] ;

[0087] In the formula, is the global wind direction change degree threshold; It is the correction factor adjusted according to the characteristics of wind resources within the wind speed range; It is the interval wind speed data The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval;

[0088] In the data post-processing module, the formula for calculating the global duration threshold by weighted average method is:

[0089] ;

[0090] In the formula, is the global wind direction change degree threshold.

[0091] As a preferred solution of the yaw optimization analysis device for wind turbine generator sets, in the yaw strategy optimization module, the formula for optimizing the yaw strategy of the wind turbine set in the wind farm by using the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is:

[0092] ;

[0093] In the formula, is the optimized yaw strategy, , , is the change in power loss, which is a constant parameter adjusted according to the target wind farm and wind turbine characteristics; It's the wind speed.

[0094] The present invention has the following advantages:

[0095] The present invention performs a series of complete data pre-processing and data post-processing on the collected wind resource data, including anomaly identification and interpolation, time synchronization filtering, normalized distribution fitting, and data distribution form to improve data quality and evaluation accuracy, and can provide the best data basis for subsequent threshold calculation;

[0096] The present invention processes the wind speed interval in sections and combines the probability distribution analysis of the wind direction change section, so that the yaw threshold can be more adapted to the wind direction change characteristics under different wind speed conditions, and optimizes the yaw threshold calculation and control strategy formulation in different intervals;

[0097] The present invention comprehensively considers the distribution of wind speed and wind direction, takes the wind resource data of the target wind farm site as the calculation basis, and accurately calculates the optimal yaw threshold of the wind turbine in the form of cumulative probability density; it can dynamically adjust the yaw strategy according to different regions and seasonal conditions of the wind farm, and maximize the improvement of quality and efficiency;

[0098] The present invention avoids the occurrence of too frequent or too delayed yaw actions, reduces mechanical wear of the yaw system, reduces energy loss, effectively prevents cable entanglement, can cope with changeable wind conditions, and improves power generation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0100] Figure 1 A schematic diagram of the flow chart of the yaw optimization analysis method for a wind turbine generator set provided in an embodiment of the present invention;

[0101] Figure 2 A technical implementation route of the wind turbine yaw optimization analysis method provided in an embodiment of the present invention;

[0102] Figure 3 A comparison of the yaw times before and after the threshold value optimization of the yaw optimization analysis method for a wind turbine generator set provided in an embodiment of the present invention;

[0103] Figure 4 A comparison of the power output of a wind turbine generator set before and after the threshold value is optimized in the yaw optimization analysis method provided in an embodiment of the present invention;

[0104] Figure 5 Schematic diagram of the structure of the wind turbine yaw optimization analysis device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0105] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0106] Example 1

[0107] See also Figure 1 and Figure 2 The embodiment of the present invention provides a method for optimizing the yaw of a wind turbine generator set, comprising the following steps:

[0108] S1. Acquire original data of a wind farm, wherein the original data of a wind farm includes original wind speed data and original wind direction data;

[0109] S2. performing data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting;

[0110] S3, performing data post-processing on the original data of the wind farm after the data pre-processing, wherein the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; after obtaining the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval, calculating the optimal global wind direction change degree threshold and the global wind speed interval duration threshold by weighted average method;

[0111] S4, optimizing the yaw strategy of the wind turbines in the wind farm by using the optimal global wind direction change degree threshold and the global wind speed interval duration threshold;

[0112] S5. Feedback the optimized yaw strategy of the wind turbines in the wind farm to the wind turbine group in the wind farm, control the wind turbine group in the wind farm to terminate the yaw angle execution action outside the optimal yaw range, and make the wind turbine group operate in the best windward state.

[0113] In this embodiment, in step S1, infrared equipment (such as laser radar) is used in the wind farm and a wind tower is installed to collect data such as wind speed and wind direction to obtain an initial sample. For the initial sample, corresponding pre-processing is performed in subsequent steps to make the data complete and reliable.

[0114] In this embodiment, in step S2, outliers in the wind speed and wind direction data are identified and corrected, and the data points of the missing data are supplemented. The distribution characteristics of wind speed and wind direction (such as normal distribution or periodic distribution) are used to determine the threshold of outliers. Outliers that exceed the dynamically determined range will be eliminated and corrected by the interpolation method of the adjacent points before and after.

[0115] Specifically, in the process of performing the adaptive anomaly recognition and interpolation completion on the original wind speed data, the specific formula is as follows:

[0116] ;

[0117] Specifically, according to the local mean of wind speed and wind speed standard deviation , set the adaptive threshold range [ ], where It is a regulating factor used to control the judgment criteria of abnormal values. The regulating factors of wind farms in different regions Dynamic adjustment; if wind speed data If the wind speed data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind speed data before and after and The mean of is interpolated and completed;

[0118] In step S2, since the wind direction data has periodic characteristics, the abnormal value determination process is similar to the wind speed processing, and needs to be optimized for the periodicity of the wind direction. Specifically, in the process of performing the adaptive abnormality identification and interpolation completion on the original wind direction data, the specific formula is as follows:

[0119] ;

[0120] Specifically, according to the average wind direction and wind direction standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind direction data If the wind direction data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind direction data and The mean of is interpolated and completed. Therefore, through adaptive anomaly recognition and interpolation completion, abnormal wind speed and wind direction data can be effectively eliminated, and smooth interpolation and completion can be performed on them to avoid subsequent analysis errors caused by data anomalies.

[0121] In this embodiment, in step S2, time synchronization adjustment and noise filtering are combined to handle the unstable sampling frequency of wind speed and wind direction data and filter out high-frequency noise. The adaptive filter dynamically adjusts the filtering strength according to the rate of change of the data to ensure that the key features of the signal are retained while removing noise.

[0122] Specifically, assuming that the wind speed sampling time There are changes, by adding time adjustment items To correct the wind speed data. At the same time, the filter uses weights `` and the second derivative of wind speed change Dynamically adjust the output to balance data smoothness and response speed. Specifically, the formula for performing the adaptive time synchronization and filtering on the original wind speed data is:

[0123] ;

[0124] In the formula, is the wind speed data after filtering; It is the wind speed data that has been filtered at the last moment; is the wind speed sampling time interval; is the time adjustment item; Control the weight distribution between new wind speed data and historical data, Used to correct the wind speed sampling time difference, Through acceleration Dynamically control the sensitivity of the wind speed filter.

[0125] The time synchronization and filtering of wind direction data are similar to those of wind speed, and wind direction data will also be adjusted in time synchronization in the same way and dynamically smoothed by the filter. Specifically, the formula for the adaptive time synchronization and filtering of the original wind direction data is:

[0126] ;

[0127] In the formula, is the wind direction data after filtering; It is the wind direction data that has been filtered at the last moment; To control the weight distribution between new data and historical data of wind direction; Used to correct the difference in wind direction sampling time; is the wind direction sampling time interval.

[0128] Therefore, through adaptive time synchronization and filtering, the impact of unstable sampling time on the data can be effectively solved, and the wind speed and wind direction data can be efficiently filtered to eliminate noise.

[0129] In this embodiment, in step S2, data normalization is combined with probability distribution fitting to ensure that the scale of wind speed and wind direction data is unified, and their distribution characteristics are modeled, thereby laying a foundation for subsequent statistical analysis and evaluation.

[0130] Among them, the normalization of wind speed and the logarithmic normal distribution fitting use logarithmic normalization to convert the wind speed into a standardized form for subsequent analysis. Specifically, in the process of performing the nonlinear normalization and distribution fitting on the original wind speed data:

[0131] Logarithmic normalization is used to convert the wind speed into a standardized form. The normalization formula is:

[0132] ;

[0133] In the formula, is the normalized wind speed data; is the minimum wind speed data; is the maximum wind speed data;

[0134] Assuming that the wind speed conforms to the log-normal distribution, the probability density of the wind speed is expressed by the distribution function :

[0135] ;

[0136] In the formula, is the wind speed; is the standard deviation of wind speed; is the log mean of the lognormal distribution; this distribution provides better descriptive statistics for wind speed data that have asymmetric distribution characteristics.

[0137] The normalization process of the wind direction data uses a cosine function to ensure that its periodic characteristics are retained. Specifically, in the process of performing the nonlinear normalization and distribution fitting on the original wind direction data:

[0138] Normalize the wind direction data using the cosine function:

[0139] ;

[0140] In the formula, is the normalized wind direction data; is the original wind direction data, indicating the current wind direction; is the reference wind direction, which is the relative angle when the reference value is used for normalization;

[0141] Use von Mises to model the distribution of wind direction data:

[0142] ;

[0143] In the formula, To control the concentration of the distribution, represents the average wind direction. is a modified Bessel function. This distribution can accurately reflect the periodic changes of wind direction data.

[0144] In step S2, the original data of the wind farm is processed through adaptive anomaly identification and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting, which can effectively eliminate outliers in the data, smooth data fluctuations, ensure time synchronization, and finally lay the foundation for subsequent analysis through normalization and distribution fitting.

[0145] In this embodiment, in step S3, for each wind speed interval First, we need to determine the representative wind direction change degree threshold This threshold is determined by the cumulative probability distribution function and is defined as the degree of wind direction change that satisfies the set cumulative probability value (usually 60%), that is, the threshold is the angle that ensures that 60% of the wind directions in the interval will not exceed. The formula for calculating the representative wind direction change degree threshold in each wind speed interval based on the probability distribution is:

[0146] ;

[0147] In the formula, For wind speed data in the interval Inside, wind direction change angle The cumulative probability of The code indicating the wind speed range; For interval wind speed data The probability density function of wind direction change; 0.6 is the set cumulative probability threshold; by solving this equation, the representative wind direction change degree threshold in different wind speed ranges can be obtained .

[0148] In this embodiment, in step S3, the wind speed data in the same interval Based on the calculated wind direction change threshold , count the duration of wind direction changes, and determine the duration threshold in each wind speed interval through the probability distribution function The cumulative probability method is also used, which is defined as the duration that meets the set cumulative probability (such as 60%). The formula for calculating the wind speed interval duration threshold based on the probability distribution is:

[0149] ;

[0150] In the formula, It is the wind speed data in the interval Duration The cumulative probability of is the probability density function of the duration of wind direction change within the wind speed range; 0.6 is the set cumulative probability threshold. After the above calculation, the duration threshold of each wind speed range can be obtained: .

[0151] In this embodiment, in step S3, after obtaining the representative wind direction change degree threshold and duration threshold in each wind speed interval, the global optimal wind direction change degree threshold and duration threshold are calculated by weighted average method. This weighted average takes into account the wind resource characteristics of different wind speed intervals and the corresponding wind direction change impact.

[0152] Specifically, use the wind speed interval weight and correction factor The weighted average method is used to calculate the global wind direction change threshold. The formula for calculating the optimal global wind direction change degree threshold by the weighted average method is:

[0153] ;

[0154] In the formula, is the global wind direction change degree threshold; It is a correction factor adjusted according to the characteristics of wind resources within the wind speed range to improve the accuracy of the formula; It is the interval wind speed data The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval;

[0155] The formula for calculating the global duration threshold by weighted average method is:

[0156] ;

[0157] In the formula, is the global wind direction change degree threshold.

[0158] In this embodiment, in step S4, the yaw strategy of the wind turbine group can be optimized by the optimal global wind direction change degree threshold and the global wind speed interval duration threshold to reduce power loss. Specifically, the formula for optimizing the yaw strategy of the wind turbine group in the wind farm by the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is:

[0159] ;

[0160] In the formula, is the optimized yaw strategy, , , is the change in power loss, which is a constant parameter adjusted according to the target wind farm and wind turbine characteristics; It's the wind speed.

[0161] In this embodiment, in step S5, the optimized yaw strategy of the wind farm wind turbine is fed back to the wind farm wind turbine group, and the wind farm wind turbine group is controlled to terminate the yaw angle execution action outside the optimal yaw interval, so that the wind turbine group operates in the optimal windward state. Then, the new yaw threshold is fed back to the wind turbine group, and the yaw angle execution action outside the optimal yaw interval is terminated, so that the wind turbine operates in the optimal windward state, and the yaw frequency is reduced, thereby increasing the power generation of the unit.

[0162] To verify the optimization effect of the embodiment of the present invention, a simulation analysis is performed using the historical SCADA data of a 2.0MW wind turbine. The wind direction and wind speed in the historical data are used as input, the initial position of the cabin is referred to, and the corresponding yaw angle and duration threshold are selected for different wind speed intervals in combination with the above threshold calculation method. The optimized cabin position is simulated and 24 hours of second-level data is selected for calculation. The results are as follows: Figure 2 As shown; and using the wind power calculation formula combined with the unit parameters, the power output before and after the threshold optimization is simulated and calculated. The results are as follows Figure 3 According to the change of the cabin position, the number of yaws of the unit within 24 hours is counted, and the power output growth within 24 hours is counted. The results are shown in Table 1:

[0163] Table 1 Comparison of yaw times and power output before and after threshold optimization

[0164]

[0165] The simulation test proves that the technical solution of the embodiment of the present invention can significantly reduce the number of yaw times of the unit and significantly improve the power output by about 2.12%, which verifies the effectiveness of the method of the present invention.

[0166] In summary, the embodiment of the present invention obtains original data of a wind farm, and the original data of the wind farm includes original wind speed data and original wind direction data; performs data pre-processing on the original data of the wind farm, and the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting; performs data post-processing on the original data of the wind farm after the data pre-processing, and the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; after obtaining the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval, calculate the optimal global wind direction change degree threshold and the global wind speed interval duration threshold by weighted average method; optimize the yaw strategy of the wind turbine group in the wind farm by the optimal global wind direction change degree threshold and the global wind speed interval duration threshold; feed back the optimized yaw strategy of the wind turbine in the wind farm to the wind turbine group in the wind farm, control the wind turbine group in the wind farm to terminate the yaw angle execution action outside the optimal yaw interval, and make the wind turbine group operate in the optimal windward state. The present invention performs a series of complete data pre-processing and data post-processing on the collected wind resource data, including anomaly identification and interpolation, time synchronization filtering, normalized distribution fitting, and data distribution form to improve data quality and evaluation accuracy, and can provide the best data basis for subsequent threshold value calculation; the present invention performs segmented processing on the wind speed interval, and combines the probability distribution analysis of the wind direction change section, so that the yaw threshold can be more adapted to the wind direction change characteristics under different wind speed conditions, and optimizes the yaw threshold calculation and control strategy formulation in different intervals; the present invention comprehensively considers the wind speed and wind direction distribution forms, takes the target wind farm site wind resource data as the calculation basis, and accurately calculates the optimal wind turbine yaw threshold in the form of cumulative probability density; the yaw strategy can be dynamically adjusted according to different regions and seasonal conditions of the wind farm, so as to maximize the improvement of quality and efficiency; the present invention avoids the occurrence of yaw actions that are too frequent or too delayed, reduces the mechanical wear of the yaw system, reduces energy loss, effectively prevents cable entanglement, can cope with variable wind conditions, and improves power generation efficiency.

[0167] Example 2

[0168] See also Figure 5 Embodiment 2 of the present invention further provides a wind turbine generator set yaw optimization analysis device, comprising:

[0169] A wind farm original data acquisition module 100 is used to acquire wind farm original data, wherein the wind farm original data includes original wind speed data and original wind direction data;

[0170] A data pre-processing module 200 is used to perform data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting;

[0171] The data post-processing module 300 is used to perform data post-processing on the raw data of the wind farm after the data pre-processing, wherein the data post-processing includes calculating the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval based on the probability distribution; after obtaining the representative wind direction change degree threshold and the wind speed interval duration threshold in each wind speed interval, calculating the optimal global wind direction change degree threshold and the global wind speed interval duration threshold by weighted average method;

[0172] A yaw strategy optimization module 400 is used to optimize the yaw strategy of the wind turbine group in the wind farm through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold;

[0173] The yaw strategy feedback control module 500 is used to feed back the optimized yaw strategy of the wind farm wind turbine to the wind farm wind turbine group, control the wind farm wind turbine group to terminate the yaw angle execution action outside the optimal yaw range, and make the wind turbine group operate in the best windward state.

[0174] In this embodiment, in the data pre-processing module 200:

[0175] According to the local mean of wind speed and wind speed standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind speed data If the wind speed data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind speed data before and after and The mean value of is interpolated.

[0176] In this embodiment, in the data pre-processing module 200:

[0177] According to the average wind direction and wind direction standard deviation , set the adaptive threshold range [ ], where is the adjustment factor; if the wind direction data If the wind direction data exceeds the adaptive threshold range, is an outlier, and uses the adjacent wind direction data and The mean value of is interpolated.

[0178] In this embodiment, in the data pre-processing module 200, the formula for performing the adaptive time synchronization and filtering on the original wind speed data is:

[0179] ;

[0180] In the formula, is the wind speed data after filtering; It is the wind speed data that has been filtered at the last moment; is the wind speed sampling time interval; is the time adjustment item; Control the weight distribution between new wind speed data and historical data, Used to correct the wind speed sampling time difference, Through acceleration Dynamically control the sensitivity of the wind speed filter.

[0181] In this embodiment, in the data pre-processing module 200, the formula for performing the adaptive time synchronization and filtering on the original wind direction data is:

[0182] ;

[0183] In the formula, is the wind direction data after filtering; It is the wind direction data that has been filtered at the last moment; To control the weight distribution between new data and historical data of wind direction; Used to correct the difference in wind direction sampling time; is the wind direction sampling time interval;

[0184] In the data pre-processing module 200, logarithmic normalization is used to convert the wind speed into a standardized form, and the normalization formula is:

[0185] ;

[0186] In the formula, is the normalized wind speed data; is the minimum wind speed data; is the maximum wind speed data;

[0187] In the data pre-processing module 200, it is assumed that the wind speed conforms to the log-normal distribution and the probability density of the wind speed is represented by the distribution function: :

[0188] ;

[0189] In the formula, is the wind speed; is the standard deviation of wind speed; is the log mean of the lognormal distribution.

[0190] In this embodiment, in the data pre-processing module 200, the wind direction data is normalized using a cosine function:

[0191] ;

[0192] In the formula, is the normalized wind direction data; is the original wind direction data, indicating the current wind direction; is the reference wind direction, which is the relative angle when the reference value is used for normalization;

[0193] In the data pre-processing module 200, von Mises is used to perform distribution modeling on the wind direction data:

[0194] ;

[0195] In the formula, To control the concentration of the distribution, represents the average wind direction. is the modified Bessel function.

[0196] In this embodiment, in the data post-processing module 300, the formula for calculating the representative wind direction change degree threshold in each wind speed interval based on the probability distribution is:

[0197] ;

[0198] In the formula, For wind speed data in the interval Inside, wind direction change angle The cumulative probability of The code indicating the wind speed range; For interval wind speed data Probability density function of internal wind direction change; 0.6 is the set cumulative probability threshold;

[0199] In the data post-processing module 300, the formula for calculating the wind speed interval duration threshold based on probability distribution is:

[0200] ;

[0201] In the formula, It is the wind speed data in the interval Duration The cumulative probability of is the probability density function of the duration of wind direction change within the wind speed; 0.6 is the set cumulative probability threshold.

[0202] In this embodiment, in the data post-processing module 300, the formula for calculating the optimal global wind direction change degree threshold by weighted average method is:

[0203] ;

[0204] In the formula, is the global wind direction change degree threshold; It is the correction factor adjusted according to the characteristics of wind resources within the wind speed range; It is the interval wind speed data The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval;

[0205] In the data post-processing module 300, the formula for calculating the global duration threshold by weighted average method is:

[0206] ;

[0207] In the formula, is the global wind direction change degree threshold.

[0208] In this embodiment, in the yaw strategy optimization module 400, the formula for optimizing the yaw strategy of the wind turbine group in the wind farm by using the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is:

[0209] ;

[0210] In the formula, is the optimized yaw strategy, , , is the change in power loss, which is a constant parameter adjusted according to the target wind farm and wind turbine characteristics; It's the wind speed.

[0211] It should be noted that the information interaction, execution process and other contents between the modules of the above-mentioned device are based on the same concept as the method embodiment in Example 1 of the present application, and the technical effects they bring are the same as those of the method embodiment of the present application. For specific contents, please refer to the description in the method embodiment shown in the previous part of the present application, and will not be repeated here.

[0212] Example 3

[0213] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code of a wind turbine yaw optimization analysis method is stored, and the program code includes instructions for executing the wind turbine yaw optimization analysis method of embodiment 1 or any possible implementation thereof.

[0214] Computer-readable storage media can be any available media that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives, SSDs), etc.

[0215] Example 4

[0216] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0217] The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the wind turbine yaw optimization analysis method of embodiment 1 or any possible implementation thereof.

[0218] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor implemented by reading software codes stored in a memory. The memory can be integrated in the processor or can be located outside the processor and exist independently.

[0219] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center.

[0220] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0221] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A method for yaw optimization analysis of a wind turbine generator set, characterized in that: include: Acquiring original data of the wind farm, wherein the original data of the wind farm includes original wind speed data and original wind direction data; Performing data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting; Performing data post-processing on the raw data of the wind farm after the data pre-processing, wherein the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; After obtaining the representative wind direction change degree threshold and wind speed interval duration threshold in each wind speed interval, the optimal global wind direction change degree threshold and global wind speed interval duration threshold are calculated by weighted average method; The yaw strategy of wind turbines in wind farms is optimized through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold. Feedback the optimized yaw strategy of the wind turbines in the wind farm to the wind turbine group in the wind farm, control the wind turbine group in the wind farm to terminate the yaw angle execution action outside the optimal yaw range, and make the wind turbine group operate in the best windward state; In the process of performing the adaptive anomaly identification and interpolation completion on the original wind speed data: According to the local mean of wind speed μ v and wind speed standard deviation σ v , set the adaptive threshold range [μ v -α·σ v ,μ v +α·σ v ], where α is the adjustment factor; if the wind speed data v i If the wind speed data v exceeds the adaptive threshold range, i is an outlier, and uses the adjacent wind speed data v i-1 and v i+1 The mean of is interpolated and completed; In the process of performing the adaptive anomaly identification and interpolation completion on the original wind direction data: According to the average wind direction μ θ and wind direction standard deviation σ θ , set the adaptive threshold range [μ θ -α`·σ θ ,μ θ +α`·σ θ ], where α` is the adjustment factor; if the wind direction data μ i If the wind direction data μ exceeds the adaptive threshold range, i is an outlier, and uses the adjacent wind direction data μ i-1 and μ i+1 The mean of is interpolated and completed; The formula for performing the adaptive time synchronization and filtering on the original wind speed data is: In the formula, v filtered,i is the wind speed data after filtering; v filtered,i-1 is the wind speed data that has been filtered at the previous moment; Δt i is the wind speed sampling time interval; β·Δt i is the time adjustment term; α controls the weight distribution between the new wind speed data and the historical data, β is used to correct the wind speed sampling time difference, and γ is adjusted by acceleration. Dynamically control the sensitivity of the wind speed filter; The formula for performing the adaptive time synchronization and filtering on the original wind direction data is: i filtered,i =α```(θ i +β`·Δt i `)+(1-α```)θ filtered,i-1 In the formula, θ filtered,i is the wind direction data after filtering; θ filtered,i-1 is the wind direction data that has been filtered at the last moment; α``` controls the weight distribution between the new wind direction data and the historical data; β` is used to correct the wind direction sampling time difference; Δt i ` is the wind direction sampling time interval.

2. A wind turbine yaw optimization analysis method according to claim 1, characterized in that: In the process of performing the nonlinear normalization and distribution fitting on the original wind speed data: Logarithmic normalization is used to convert the wind speed into a standardized form. The normalization formula is: In the formula, v norm,i is the normalized wind speed data; v min is the minimum wind speed data; v max is the maximum wind speed data; Assuming that the wind speed conforms to the log-normal distribution, the probability density of wind speed is expressed by the distribution function f(v): Where v is the wind speed; σ is the standard deviation of wind speed; μ is the logarithmic mean of the lognormal distribution; In the process of performing the nonlinear normalization and distribution fitting on the original wind direction data: Normalize the wind direction data using the cosine function: In the formula, θ norm,i is the normalized wind direction data; θ i is the original wind direction data, indicating the current wind direction; θ0 is the reference wind direction, which is the relative angle when the reference value is used for normalization; Use von Mises to model the distribution of wind direction data: Where κ is the concentration of the control distribution, μ θ represents the average wind direction, and I0 is the modified Bessel function.

3. A wind turbine yaw optimization analysis method according to claim 2, characterized in that: The formula for calculating the representative wind direction change degree threshold in each wind speed interval based on probability distribution is: In the formula, P(θ threshold,i |v interval,i ) is the wind speed data v in the interval interval,i The cumulative probability of wind direction changing angle θ within θ; v interval,i The code for wind speed range; f(θ|v interval,i ) is the interval wind speed data v interval,i Probability density function of internal wind direction change; 0.6 is the set cumulative probability threshold; The formula for calculating the wind speed interval duration threshold based on probability distribution is: In the formula, P(T threshold,i |v interval,i ) is the wind speed data v in the interval interval,i The cumulative probability of duration T; f(T|v interval,i ) is the probability density function of the duration of wind direction change within the wind speed; 0.6 is the set cumulative probability threshold.

4. A method for yaw optimization analysis of a wind turbine generator set according to claim 3, characterized in that: The formula for calculating the optimal global wind direction change degree threshold by weighted average method is: In the formula, θ glbal is the global wind direction change threshold; α i is the correction factor adjusted according to the characteristics of wind resources within the wind speed range; ω i is the interval wind speed data v i The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval; The formula for calculating the global duration threshold by weighted average method is: Where, T global is the global wind direction change degree threshold.

5. A method for yaw optimization analysis of a wind turbine generator set according to claim 4, characterized in that: The formula for optimizing the yaw strategy of wind turbines in wind farms by using the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is: Where ΔP is the optimized yaw strategy, C1, C2, and C3 are the changes in power loss, which are constant parameters adjusted according to the target wind farm and wind turbine characteristics; v is the wind speed.

6. A wind turbine yaw optimization analysis device, characterized in that: include: A wind farm original data acquisition module is used to acquire the original data of the wind farm, wherein the original data of the wind farm includes original wind speed data and original wind direction data; A data pre-processing module, used for performing data pre-processing on the raw data of the wind farm, wherein the data pre-processing includes adaptive anomaly recognition and interpolation completion, adaptive time synchronization and filtering, and nonlinear normalization and distribution fitting; A data post-processing module, used for performing data post-processing on the raw data of the wind farm after the data pre-processing, wherein the data post-processing includes calculating a representative wind direction change degree threshold and a wind speed interval duration threshold in each wind speed interval based on probability distribution; After obtaining the representative wind direction change degree threshold and wind speed interval duration threshold in each wind speed interval, the optimal global wind direction change degree threshold and global wind speed interval duration threshold are calculated by weighted average method; A yaw strategy optimization module is used to optimize the yaw strategy of the wind turbine group in the wind farm through the optimal global wind direction change degree threshold and the global wind speed interval duration threshold; The yaw strategy feedback control module is used to feed back the optimized yaw strategy of the wind farm wind turbine to the wind farm wind turbine group, control the wind farm wind turbine group to terminate the yaw angle execution action outside the optimal yaw range, and make the wind turbine group operate in the best windward state; In the data pre-processing module: According to the local mean of wind speed μ v and wind speed standard deviation σ v , set the adaptive threshold range [μ v -α·σ v ,μ v +α·σ v ], where α is the adjustment factor; if the wind speed data v i If the wind speed data v exceeds the adaptive threshold range, i is an outlier, and uses the adjacent wind speed data v i-1 and v i+1 The mean of is interpolated and completed; In the data pre-processing module: According to the average wind direction μ θ and wind direction standard deviation σ θ , set the adaptive threshold range [μ θ -α`·σ θ ,μ θ +α`·σ θ ], where α` is the adjustment factor; if the wind direction data μ i If the wind direction data μ exceeds the adaptive threshold range, i is an outlier, and uses the adjacent wind direction data μ i-1 and μ i+1 The mean of is interpolated and completed; In the data pre-processing module, the formula for performing the adaptive time synchronization and filtering on the original wind speed data is: In the formula, v filtered,i is the wind speed data after filtering; v filtered,i-1 is the wind speed data that has been filtered at the previous moment; Δt i is the wind speed sampling time interval; β·Δt i is the time adjustment term; α controls the weight distribution between the new wind speed data and the historical data, β is used to correct the wind speed sampling time difference, and γ is adjusted by acceleration. Dynamically control the sensitivity of the wind speed filter; In the data pre-processing module, the formula for performing the adaptive time synchronization and filtering on the original wind direction data is: i filtered,i =α```(θ i +β`·Δt i `)+(1-α```)θ filtered,i-1 In the formula, θ filtered,i is the wind direction data after filtering; θ filtered,i-1 is the wind direction data that has been filtered at the last moment; α``` controls the weight distribution between the new wind direction data and the historical data; β` is used to correct the wind direction sampling time difference; Δt i ` is the wind direction sampling time interval.

7. A wind turbine yaw optimization analysis device according to claim 6, characterized in that: In the data pre-processing module, logarithmic normalization is used to convert the wind speed into a standardized form. The normalization formula is: In the formula, v norm,i is the normalized wind speed data; v min is the minimum wind speed data; v max is the maximum wind speed data; In the data pre-processing module, it is assumed that the wind speed conforms to the log-normal distribution, and the probability density of the wind speed f(v) is represented by the distribution function: Where v is the wind speed; σ is the standard deviation of wind speed; μ is the logarithmic mean of the lognormal distribution; In the data pre-processing module, the wind direction data is normalized using the cosine function: In the formula, θ norm,i is the normalized wind direction data; θ i is the original wind direction data, indicating the current wind direction; θ0 is the reference wind direction, which is the relative angle when the reference value is used for normalization; In the data pre-processing module, von Mises is used to perform distribution modeling on wind direction data: Where κ is the concentration of the control distribution, μ θ represents the average wind direction, and I0 is the modified Bessel function.

8. The yaw optimization analysis device for a wind turbine generator set according to claim 7, characterized in that: In the data post-processing module, the formula for calculating the representative wind direction change degree threshold in each wind speed interval based on probability distribution is: In the formula, P(θ threshold,i |v inyerval,i ) is the wind speed data v in the interval interval,i The cumulative probability of wind direction changing angle θ within θ; v interval,i The code for wind speed range; f(θ|v interval,i ) is the interval wind speed data v interval,i Probability density function of internal wind direction change; 0.6 is the set cumulative probability threshold; In the data post-processing module, the formula for calculating the wind speed interval duration threshold based on probability distribution is: In the formula, P(T threshold,i |v interval,i ) is the wind speed data v in the interval interval,i The cumulative probability of duration T; f(T|v interval,i ) is the probability density function of the duration of wind direction change within the wind speed; 0.6 is the set cumulative probability threshold; In the data post-processing module, the formula for calculating the optimal global wind direction change degree threshold by weighted average method is: In the formula, θ glbal is the global wind direction change threshold; α i is the correction factor adjusted according to the characteristics of wind resources within the wind speed range; ω i is the interval wind speed data v i The weight is calculated based on the frequency of wind speed or wind energy contribution within the interval; In the data post-processing module, the formula for calculating the global duration threshold by weighted average method is: Where, T global is the global wind direction change degree threshold; In the yaw strategy optimization module, the formula for optimizing the yaw strategy of the wind turbine group in the wind farm by using the optimal global wind direction change degree threshold and the global wind speed interval duration threshold is: Where ΔP is the optimized yaw strategy, C1, C2, and C3 are the changes in power loss, which are constant parameters adjusted according to the target wind farm and wind turbine characteristics; v is the wind speed.

Images