Yaw correction method based on wind energy utilization coefficient

Abstract

The invention discloses a yaw correction method based on a wind energy utilization coefficient. The yaw correction method comprises the following steps that zero calibration is conducted on a wind indicator; collecting data; data cleaning; processing data; obtaining unit characteristics; screening data; removing non-operation data; further eliminating the power limiting data; further eliminating data with relatively high dispersion in the power-yaw error characteristics; further eliminating data with relatively high dispersion in the wind speed-power characteristics by adopting a statistical mode; further acquiring data corresponding to the B-C segment; and the inherent yaw error is eliminated. The method comprises the following steps: under the condition of zero calibration of a wind indicator, collecting data, cleaning the data, determining effective data, selecting data of an optimal operation section in a main wind direction of a unit according to data characteristics of the unit for analysis and statistics, and determining an inherent yaw error according to a wind energy utilization coefficient and a characteristic curve of the yaw error. And the fixed yaw error is set as the wind facing angle, so that the influence of the inherent yaw error on the unit power is eliminated, and the generating capacity of the unit is improved.

Description

Technical Field

[0001] The invention relates to the technical field of yaw correction of wind turbines, and in particular to a yaw correction method based on wind energy utilization coefficient. Background Art

[0002] According to wind power related theories, when the yaw error of the upwind wind turbine is zero during operation, that is, when the angle between the incoming wind direction and the central axis of the nacelle is zero, the wind energy absorbed by the wind turbine is the largest. Generally, the wind measuring device of the wind turbine is mechanical and installed at the tail of the nacelle. It is affected by the wake of the impeller, and the measured wind speed and direction signals are affected by the rotating wind wheel, rather than the actual incoming wind speed and direction. At the same time, during the installation and commissioning of the unit, the wind vane will also have physical deviations. The deviation between the measured value of the mechanical wind measuring device installed at the tail of the nacelle and the free stream wind direction value in front of the unit is the inherent yaw error of the unit.

[0003] According to the wind energy calculation formula, the power absorbed by the wind turbine is in a cubic relationship with the wind speed, and the power generation lost due to yaw error is directly related to the yaw error. If the existing inherent yaw error can be corrected, the power generation of the unit can be further improved.

[0004] The Chinese patent "A method and device for determining the static yaw error of a wind turbine" with publication number CN 112031997 A and the Chinese patent "A method for identifying the inherent deviation and step size of the yaw error of a wind turbine based on power curve analysis" with publication number CN 109667727 A both use correction methods based on power with the following problems: 1) The relationship between power and yaw error cannot directly reflect the inherent yaw error, but reflects the yaw control strategy of the unit, that is, whether the yaw system of the unit is zeroed, and this zero position is the 0 position sensed by the main control; 2) The accuracy of the correction is mainly based on the number of divided subsets. As the number of divided subsets increases, the calibration value can only be infinitely close to the actual deviation. Moreover, with the same amount of data, the more subsets are divided, the fewer points each subset will have, and the result will not be representative. Therefore, the existing method cannot directly locate the correction value, but can only be infinitely approximated. Summary of the invention

[0005] In order to solve the problem mentioned in the background technology, the present invention provides a yaw correction method based on wind energy utilization coefficient.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions: A yaw correction method based on wind energy utilization coefficient, comprising the following steps:

[0007] Step 1: Zeroing the wind vane: Zero the wind vane of the operating unit to ensure that the data collected subsequently is the data with the angle between the wind vane zero position and the center axis of the nacelle being zero;

[0008] Step 2: Collect data: The collected data includes wind speed, wind direction, yaw error, power, air density, speed, unit status flag and other unit data. The sampling frequency is 1Hz. The collection time needs to be based on the amount of available data Q obtained from subsequent analysis. 主 To determine, and at the same time obtain the impeller diameter and altitude of the unit;

[0009] Step 3: Data cleaning: First, remove abnormal communication data, data with excessive collection values, and data with abnormal data formats. Then divide the collected data into bins every 3 minutes, and count the number of data in each bin. For data with a collection frequency of 1Hz, there are normally 180 data in each bin. The average value calculated using 180 data is the 3-minute average value. Remove points with less than 180*80%=144 data in each bin to ensure the credibility of the obtained 3-minute average data; obtain 3-minute average time series data such as wind speed, wind direction, yaw error, power, air density, speed, and unit status flag;

[0010] Step 4: Data processing: Apply the 3-minute average to obtain the air density and obtain the corresponding air density time series. Then filter the data based on the calculated air density and remove the data with the calculated air density less than 0 to exclude the false data caused by the abnormal temperature measurement. Then, according to the wind energy calculation formula, apply the air density, wind speed, and impeller diameter to calculate the theoretical power of each 3-minute average wind speed and obtain the theoretical power time series P. j ; The formula is as follows:

[0011] P=0.5*ρ*A*v 3

[0012] Where P is the theoretical power corresponding to the wind speed, ρ is the air density, A is the impeller area, which is calculated based on the diameter, and v is the wind speed.

[0013] Then, according to the 3-minute power time series P obtained in the data cleaning step, i and P j The wind energy utilization coefficient time series C pj , C pj =P i / P j ;

[0014] Step 5: Unit characteristics acquisition: Use time series data to analyze the speed-torque characteristics of the unit, and obtain the speed n corresponding to points B and C in the optimal control interval BC section of the speed-torque characteristic curve B 、n C ; Power PB , P C , yaw error-wind energy utilization coefficient characteristics, the analysis is mainly based on these two characteristics, and the wind speed-power characteristics, power-yaw error characteristics, power-pitch angle characteristics, and wind rose diagram are used to assist the analysis; according to the power P B , P C Get the wind speed V corresponding to the two points from the wind speed-power characteristic B 、V C ; and get the main wind direction from the wind rose diagram;

[0015] Step 6: Data screening: mainly to extract data for power improvement analysis;

[0016] Step 7: Eliminate non-operating data: According to the unit operation model flag, eliminate the data of the unit in the non-operating state;

[0017] Step 8: Further eliminate power-limited data: according to the power-pitch angle characteristic, eliminate the data with power less than the rated power and pitch angle greater than the minimum pitch angle. The minimum pitch angle is obtained according to the power-pitch angle characteristic analysis;

[0018] Step 9: further remove data with high discreteness in the power-yaw error characteristics. The removal principle is to remove data with a yaw error greater than 20 and data with a yaw error less than -20 within the power range of greater than 0 and less than 1.1 times the rated power.

[0019] Step 10: further use statistical methods to eliminate data with high dispersion in wind speed-power characteristics;

[0020] Step 11: Further obtain the data corresponding to the BC segment: that is, obtain the speed greater than n B and less than n C And the power is greater than P B and less than P C Time series data of

[0021] The data is divided into bins according to wind direction, with each 22.5 degrees of wind direction as a data bin. The total wind direction is divided into 16 data bins. The number of each data bin is counted, and the amount of data in the main wind direction data bin in the wind rose diagram is counted to ensure that the data volume Q 主 Greater than 20*(V C +1-V B )*2, the amount of data in other data warehouses Q 非主 If it is greater than 10*(V C +1-V B )*2, it is available, and other insufficient data will be directly eliminated; if the statistics do not obtain satisfactory data warehouses, continue to collect data until the data volume is sufficient;

[0022] After obtaining enough data, the yaw error-wind energy utilization coefficient characteristics of the main wind direction data are analyzed, and its characteristics are divided into three cases:

[0023] The first type: parabolic type, then determine the yaw error θ corresponding to the highest point 高 is the inherent yaw error, and the yaw angle to wind is set to θ 高 At this angle, the power output of the unit is the largest, and there is no need to change it after setting.

[0024] The second type: a straight line with high left and low right, then determine the left boundary value θ of the yaw angle in the feature data 左 , first set θ 左 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined;

[0025] The third type: a straight line with a low left and a high right, then the right boundary value θ of the yaw angle in the feature data is determined 右 , first set θ 右 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined;

[0026] θ 高 ,θ 左 ,θ 右 The corresponding specific value may be any data within the range of plus or minus 10 degrees, including 0 degrees. If the highest point appears at 0 degrees, it means that the wind vane of this unit is not zeroed, and the installation of the wind vane needs to be checked;

[0027] Step 12: Ways to eliminate the inherent yaw error: One is to set the detection angle offset value within the program, and still set the yaw angle to wind to zero degrees. However, this method will cause the correction angle to be lost or invalid as the wind vane hardware is changed and the program is optimized, thus losing the power improvement effect. This method directly sets the yaw angle to wind to the statistical inherent yaw error, rather than at zero degrees. This is more intuitive and allows a clear understanding of the inherent yaw error value of the unit. In addition, even if the wind vane hardware is calibrated to zero or replaced, the set value will not be invalid, thus ensuring the efficiency improvement effect.

[0028] After the setup is completed, the unit is analyzed regularly according to this process. Through the data characteristics of the power-yaw error, if the characteristic results are different from the analysis results, a prompt will be given that the wind vane hardware zero position needs to be recalibrated to ensure the optimal power generation of the unit.

[0029] In summary, the present invention mainly has the following beneficial effects: the present invention provides a yaw correction method based on wind energy utilization coefficient, which collects data, cleans the logarithms, determines valid data, and selects data of the optimal operating section in the main wind direction of the unit for analysis and statistics based on the data characteristics of the unit, determines the inherent yaw error based on the characteristic curve of the wind energy utilization coefficient and the yaw error, and sets the fixed yaw error as the wind angle, so as to eliminate the influence of the inherent yaw error on the power of the unit, thereby improving the power generation of the unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a flow chart of the present invention;

[0031] Figure 2 It is the torque-power characteristic diagram of the present invention;

[0032] Figure 3 is the power-pitch angle characteristic diagram;

[0033] Figure 4 It is the characteristic diagram of yaw error-wind energy utilization coefficient;

[0034] Figure 5 This is the wind speed-wind energy utilization coefficient characteristic diagram. DETAILED DESCRIPTION

[0035] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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 making creative work are within the scope of protection of the present invention.

[0036] A yaw correction method based on wind energy utilization coefficient includes the following steps (please refer to Figure 1 ):

[0037] Step 1: Zeroing the wind vane: Zero the wind vane of the operating unit to ensure that the data collected subsequently is the data with the angle between the wind vane zero position and the center axis of the nacelle is zero, so as to ensure the availability of the data for analysis;

[0038] Step 2: Collect data: The collected data includes wind speed, wind direction, yaw error, power, air density (temperature outside the cabin), speed, unit status flag and other unit data. The sampling frequency is 1Hz, and the collection time needs to be based on the amount of available data Q obtained in the subsequent analysis. 主 To determine, and at the same time obtain the impeller diameter and altitude of the unit;

[0039] Step 3: Data cleaning: First, remove abnormal communication data, data with excessive collection values, and data with abnormal data formats. Then, divide the collected data into bins every 3 minutes, and count the number of bins. For data with a collection frequency of 1Hz, there are normally 180 data in each bin. The average value calculated using 180 data is the 3-minute average value. Remove points where the amount of data in each bin is less than 180*80%=144 data to ensure the credibility of the obtained 3-minute average data. This step obtains 3-minute average time series data such as wind speed, wind direction, yaw error, power, air density (temperature outside the cabin), speed, and unit status flag.

[0040] Step 4: Data processing: Apply the 3-minute average to obtain the air density and obtain the corresponding air density time series. Then filter the data based on the calculated air density and remove the data with the calculated air density less than 0 to exclude false data caused by abnormal measured temperature. The air density calculation formula is as follows:

[0041]

[0042] Where: ρ represents the air density (kg / m 3 ), z is the unit altitude (m), T is the absolute temperature on the Kelvin scale (℃+273).

[0043] The advantage of using this air density calculation method is that it only needs to detect the temperature outside the cabin, and the altitude of the unit can be determined in this way, which is simple and low in cost.

[0044] Then, according to the wind energy calculation formula, the air density, wind speed, and impeller diameter are used to calculate the theoretical power of each 3-minute average wind speed, and the theoretical power time series P is obtained. j ; The formula is as follows:

[0045] P=0.5*ρ*A*v 3

[0046] Where P represents the theoretical power corresponding to the wind speed (kW), and ρ represents the air density (kg / m 3 ), A is the impeller area (m 2 ), calculated based on the diameter, v is the wind speed (m / s).

[0047] Then, according to the 3-minute power time series P obtained in the data cleaning step, i and P j The wind energy utilization coefficient time series C pj , C pj =P i / P j ;

[0048] Step 5: Unit characteristics acquisition: Use time series data to analyze the speed-torque characteristics of the unit (see attached Figure 2 ), obtain the speed n corresponding to points B and C in the optimal control interval BC of the speed-torque characteristic curve B 、n C ; Power P B , P C , yaw error-wind energy utilization coefficient characteristics, the analysis is mainly based on these two characteristics, and the wind speed-power characteristics, power-yaw error characteristics, power-pitch angle characteristics, and wind rose diagram are used to assist the analysis; according to the power P B , P C Get the wind speed V corresponding to the two points from the wind speed-power characteristic B 、V C ; and get the main wind direction from the wind rose diagram;

[0049] Method for determining the main wind direction: The wind direction is divided into data bins and zones according to 22.5 degrees, with 16 data bins for all wind directions. Then, the ratio of the data volume in each data bin to the total data volume is counted, and the wind directions with the top three ratios are selected as the main wind directions.

[0050] The BC section of the unit is selected for analysis because: the optimal control is the maximum power tracking section of the unit. When the unit is running in this section, the speed changes with the wind speed so that the wind energy utilization coefficient of the unit reaches the designed maximum value. When running in this section, the wind energy utilization coefficient is theoretically constant and the value is relatively stable. The relationship curve between wind speed and wind energy utilization coefficient is shown in the attached figure. Figure 5 Therefore, only this segment of data can be used for analysis to reflect the actual wind energy utilization coefficient of the unit. The wind energy utilization coefficient of other wind speed segments changes during operation, and the analysis cannot directly reflect the influencing factors of its changes and is not representative.

[0051] Step 6: Data screening: mainly to extract data for power improvement analysis;

[0052] Step 7: Eliminate non-operating data: According to the unit operation model flag, eliminate the data of the unit in the non-operating state;

[0053] Step 8: Further eliminate power-limited data: Eliminate data with power less than the rated power P according to the power-pitch angle characteristics 额定 , the pitch angle is greater than the minimum pitch angle θ 最小 The minimum pitch angle is obtained based on the power-pitch angle characteristic analysis (see Figure 3 );

[0054] The power-pitch angle characteristic is obtained by partitioning the data according to the power, with each 100kW being a partition, and calculating the mean value of the pitch angle in each partition to obtain the power-pitch angle curve.

[0055] Step 9: further remove data with high discreteness in the power-yaw error characteristics. The removal principle is to remove data with a yaw error greater than 20 and data with a yaw error less than -20 within the power range of greater than 0 and less than 1.1 times the rated power.

[0056] Step 10: Further use statistical methods to eliminate data with high dispersion in wind speed-power characteristics; first divide the wind speed into partitions according to 0.2m / s, and calculate the minimum value and first quantile (Q 1 ), median, third quartile (Q 3 ), maximum value, interquartile range (IQR) = Q 3 -Q 1 , and then by setting the upper limit value U lim =Q 3 +a*(IQR), lower limit L lim =Q 3 -b*(IQR), remove values ​​with large discreteness, where the settings of a and b are obtained through automatic iteration, and the general empirical value is in the range of 1.5-2.0;

[0057] Step 11: Further obtain the data corresponding to the BC segment: that is, obtain the speed greater than n B and less than n C And the power is greater than P B and less than P C Time series data of

[0058] The data is divided into bins according to wind direction, with each 22.5 degrees of wind direction as a data bin. The total wind direction is divided into 16 data bins. The number of each data bin is counted, and the amount of data in the main wind direction data bin in the wind rose diagram is counted to ensure that the data volume Q 主 Greater than 20*(V C +1-V B )*2, the amount of data in other data warehouses Q 非主 If it is greater than 10*(V C +1-V B )*2, it is available, and other insufficient data will be directly eliminated; if the statistics do not obtain satisfactory data warehouses, continue to collect data until the data volume is sufficient;

[0059] After obtaining enough data, the yaw error-wind energy utilization coefficient characteristics of the main wind direction data are analyzed, and its characteristics are divided into three cases (see Appendix Figure 4 ):

[0060] The first type: parabolic type, then determine the yaw error θ corresponding to the highest point 高 is the inherent yaw error, and the yaw angle to wind is set to θ 高At this angle, the power output of the unit is the largest, and there is no need to change it after setting.

[0061] The second type: a straight line with high left and low right, then determine the left boundary value θ of the yaw angle in the feature data 左 , first set θ 左 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined;

[0062] The third type: a straight line with a low left and a high right, then the right boundary value θ of the yaw angle in the feature data is determined 右 , first set θ 右 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined;

[0063] It should be noted that θ 高 ,θ 左 ,θ 右 The corresponding specific value may be any data within the range of plus or minus 10 degrees, including 0 degrees. If the highest point appears at 0 degrees, it means that the wind vane of this unit is not zeroed, and the installation of the wind vane needs to be checked;

[0064] Step 12: Ways to eliminate the inherent yaw error: One is to set the detection angle offset value within the program, and still set the yaw angle to wind to zero degrees. However, this method will cause the correction angle to be lost or invalid as the wind vane hardware is changed and the program is optimized, thus losing the power improvement effect. This method directly sets the yaw angle to wind to the statistical inherent yaw error, rather than at zero degrees. This is more intuitive and allows a clear understanding of the inherent yaw error value of the unit. In addition, even if the wind vane hardware is calibrated to zero or replaced, the set value will not be invalid, thus ensuring the efficiency improvement effect.

[0065] After the setup is completed, the unit is analyzed regularly according to this process. Through the data characteristics of the power-yaw error, if the characteristic results are different from the analysis results, a prompt will be given that the wind vane hardware zero position needs to be recalibrated to ensure the optimal power generation of the unit.

[0066] The method provides a method of obtaining the unit operation data of the optimal control section, and after data quality judgment and processing, obtaining the relationship between the yaw error and the wind energy utilization coefficient of the optimal operation section of the main wind direction to quickly determine the value of the inherent yaw error to correct the inherent yaw error. At the same time, this inherent yaw error is used as the yaw to wind value instead of being set at zero degrees to prevent the problem of no historical record of this correction value when the main control program is wrong. This method can quickly locate the correction angle of the yaw error, thereby improving the power generation of the unit. The method is simple and easy to operate, and there is no need to add high-cost wind measurement equipment such as laser radar.

[0067] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A yaw correction method based on wind energy utilization coefficient, Features: The following steps are involved: Step 1: Zeroing the wind vane: Zero the wind vane of the operating unit to ensure that the data collected subsequently is the data with the angle between the wind vane zero position and the center axis of the nacelle being zero; Step 2: Collect data: The collected data includes wind speed, wind direction, yaw error, power, air density, speed, unit status flag and other unit data. The sampling frequency is 1Hz. The collection time needs to be based on the amount of available data Q obtained from subsequent analysis. 主 To determine, and at the same time obtain the impeller diameter and altitude of the unit; Step 3: Data cleaning: First, remove abnormal communication data, data with excessive collection values, and data with abnormal data formats. Then divide the collected data into bins every 3 minutes, and count the number of data in each bin. For data with a collection frequency of 1Hz, there are normally 180 data in each bin. The average value calculated using 180 data is the 3-minute average value. Remove points with less than 180*80%=144 data in each bin to ensure the credibility of the obtained 3-minute average data; obtain 3-minute average time series data such as wind speed, wind direction, yaw error, power, air density, speed, and unit status flag; Step 4: Data processing: Apply the 3-minute average to obtain the air density and obtain the corresponding air density time series. Then filter the data based on the calculated air density and remove the data with the calculated air density less than 0 to exclude the false data caused by the abnormal temperature measurement. Then, according to the wind energy calculation formula, apply the air density, wind speed, and impeller diameter to calculate the theoretical power of each 3-minute average wind speed and obtain the theoretical power time series P. j ; The formula is as follows: P60.5*ρ*A*v 3 Where P is the theoretical power corresponding to the wind speed, ρ is the air density, A is the impeller area, which is calculated based on the diameter, and v is the wind speed. Then, according to the 3-minute power time series P obtained in the data cleaning step, i and P j The wind energy utilization coefficient time series C pj , C pj =P i / P j ; Step 5: Unit characteristics acquisition: Use time series data to analyze the speed-torque characteristics of the unit, and obtain the speed n corresponding to points B and C in the optimal control interval BC section of the speed-torque characteristic curve B 、n C ; Power P B , P C , yaw error-wind energy utilization coefficient characteristics, the analysis is mainly based on these two characteristics, and the wind speed-power characteristics, power-yaw error characteristics, power-pitch angle characteristics, and wind rose diagram are used to assist the analysis; according to the power P B , P C Get the wind speed V corresponding to the two points from the wind speed-power characteristic B 、V C ; and get the main wind direction from the wind rose diagram; Step 6: Data screening: mainly to extract data for power improvement analysis; Step 7: Eliminate non-operating data: According to the unit operation model flag, eliminate the data of the unit in the non-operating state; Step 8: Further eliminate power-limited data: according to the power-pitch angle characteristic, eliminate the data with power less than the rated power and pitch angle greater than the minimum pitch angle. The minimum pitch angle is obtained according to the power-pitch angle characteristic analysis; Step 9: further eliminate the data with high discreteness in the power-yaw error characteristics; Step 10: further use statistical methods to eliminate data with high dispersion in wind speed-power characteristics; Step 11: Further obtain the data corresponding to the BC segment: that is, obtain the speed greater than n B and less than n C And the power is greater than P B and less than P C Time series data of The data is divided into bins according to wind direction, with each 22.5 degrees of wind direction as a data bin. The total wind direction is divided into 16 data bins. The number of each data bin is counted, and the amount of data in the main wind direction data bin in the wind rose diagram is counted to ensure that the data volume Q 主 Greater than 20*(V C +1-V B )*2, the amount of data in other data warehouses Q 非主 If it is greater than 10*(V C +1-V B )*2, it is available, and other insufficient data will be directly eliminated; if the statistics do not obtain satisfactory data warehouses, continue to collect data until the data volume is sufficient; After obtaining enough data, the yaw error-wind energy utilization coefficient characteristics of the main wind direction data are analyzed, and its characteristics are divided into three cases: The first type: parabolic type, then determine the yaw error θ corresponding to the highest point 高 is the inherent yaw error, and the yaw angle to wind is set to θ 高 , the power generated by the unit is the maximum at this angle, and it does not need to be changed after setting; The second type: a straight line with high left and low right, then determine the left boundary value θ of the yaw angle in the feature data 左 , first set θ 左 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined; The third type: a straight line with a low left and a high right, then the right boundary value θ of the yaw angle in the feature data is determined 右 , first set θ 右 The yaw angle to the wind is then collected continuously, and the above steps are repeated until a parabola appears, and the best yaw angle to the wind setting value is determined; θ 高 ,θ 左 ,θ 右 The corresponding specific value may be any data within the range of plus or minus 10 degrees, including 0 degrees. If the highest point appears at 0 degrees, it means that the wind vane of this unit is not zeroed, and the installation of the wind vane needs to be checked; Step 12: How to eliminate the inherent yaw error: directly set the yaw angle to the wind to the statistical inherent yaw error; after the setting is completed, regularly analyze the unit according to this process, and through the data characteristics of the power-yaw error, if the characteristic results are different from the analysis results, a prompt will be given that the wind vane hardware zero position needs to be recalibrated to ensure the optimal power generation of the unit.

2. A yaw correction method based on wind energy utilization coefficient according to claim 1, Features: In step nine, the principle for eliminating data with high discreteness in the power-yaw error characteristic is: within the power range of greater than 0 and less than 1.1 times the rated power, eliminate data with a yaw error greater than 20, and eliminate data with a yaw error less than -20.

Images