Corrected pitch angle

Abstract

A method of correcting a pitch angle deviation of a blade of a wind turbine is provided. A first blade load measurement measured while the wind turbine is operating in a first operating mode is provided and used to determine a calibration parameter. A second blade load measurement measured while the wind turbine is operating in a second operating mode is provided and the calibration parameter is applied to determine a calibrated blade load measurement. A pitch angle deviation of at least one blade of the turbine is estimated based on an identified difference between the calibrated blade load measurements, and the pitch angle is adjusted to correct the pitch angle deviation.

Description

Technical Field

[0001] The invention relates to a method of modifying the pitch angle of a blade of a wind turbine. Background Art

[0002] Modern wind turbines typically employ a variable pitch design, where each blade of the turbine can be pitched to change its respective angle of attack relative to the incoming wind. The pitch adjustment mechanisms of the blades are typically aligned in their zero pitch position by manual inspection during construction of the turbine. This results in a difference between the reported pitch angle of the blades and the aerodynamically true pitch of the blades. This pitch angle deviation reduces the power output of the wind turbine and increases the loads experienced by the turbine.

[0003] Pitch deviation also causes rotor imbalance, which can be detected. However, rotor imbalance can also be caused by other factors, such as mass differences between blades and ice on the blades. Therefore, conventional control systems cannot identify or correct pitch deviation. Summary of the invention

[0004] A first aspect of the present invention provides a method for correcting a pitch angle deviation of a blade of a wind turbine, the wind turbine comprising a plurality of blades, the method comprising:

[0005] providing a first blade load measurement for each blade of the plurality of blades, the first blade load measurement being measured while the wind turbine is operating in a first operating mode;

[0006] determining calibration parameters to correct for differences between first blade load measurements of the blade;

[0007] providing a second blade load measurement for each blade, the second measurement being measured while the wind turbine is operating in a second operating mode, wherein the wind turbine experiences lower wind speeds during the first operating mode than during the second operating mode;

[0008] applying the calibration parameter to the second blade load measurement to determine a calibrated blade load measurement;

[0009] identifying differences between calibrated blade load measurements of the blades;

[0010] estimating a pitch angle deviation of at least one blade causing the identified discrepancy; and

[0011] The pitch angle of the at least one of the blades is changed to correct the pitch angle deviation.

[0012] In some embodiments of the method, identifying the difference between the calibrated blade load measurements may include: determining a target blade load; and identifying the difference between the calibrated blade load measurement of the at least one blade and the target blade load.

[0013] Estimating the pitch angle deviation may include estimating the pitch angle deviation of the at least one blade causing a difference between a calibrated blade load measurement of the blade and the target blade load.

[0014] In some embodiments, determining the target blade load may include determining an average of calibrated blade load measurements.

[0015] In such an embodiment, the pitch angle deviation may be estimated for each blade of the wind turbine, and the pitch angle of each of the blades may be adjusted to correct the corresponding pitch angle deviation.

[0016] In some embodiments, determining the target blade load may include: identifying one blade as a master blade; and setting a calibrated blade load measurement of the master blade as the target blade load.

[0017] In some embodiments, estimating the pitch angle deviation of the blade may include: modeling differences in expected blade load measurements of the blade for a given pitch angle deviation of the blade; and comparing the modeled differences with the identified differences to estimate the pitch angle deviation of the at least one blade.

[0018] In some embodiments, estimating the pitch angle deviation of the blade may further comprise estimating a magnitude and / or direction of a pitch angle adjustment required to correct the identified difference based on a comparison of the modeled difference with the identified difference.

[0019] In some embodiments, changing the pitch angle of the at least one of the blades to correct the pitch angle deviation may include: incrementally changing the pitch angle of the at least one blade; and re-estimating the pitch angle deviation of the at least one blade.

[0020] In some embodiments, determining a calibration parameter to correct a difference between first blade load measurement results of a blade includes: determining an average value of the first blade load measurement results; and determining a corresponding calibration parameter for each blade to correct a difference between the first blade load measurement result of the blade and the average value.

[0021] In some embodiments, determining calibration parameters to correct differences between first blade load measurement results of blades may include: identifying one of the blades as a master blade; and determining corresponding calibration parameters for each of the other blades to correct differences between the first blade load measurement result of the blade and the first blade load measurement result of the master blade.

[0022] In some embodiments, during the first operating mode, the wind turbine may be operated at idle; and during the second operating mode, the wind turbine may be operated at part load or full load.

[0023] In such an embodiment, determining the calibration parameters may include binning the measurements of the first blade load measurement for each blade according to the rotation angle of the blade associated with the respective measurement. The rotation angle is the position of the blade in the rotor plane relative to a predetermined zero position of the turbine blade.

[0024] In some embodiments, identifying differences between the calibrated blade load measurements for the blades may include binning the calibrated blade load measurements for each blade according to a wind speed associated with each measurement.

[0025] In some such embodiments, identifying a difference between calibrated blade load measurements for the blades may further include determining a relationship between wind speed and a difference between the calibrated blade load measurement for the at least one blade and a target blade load.

[0026] In some such embodiments, estimating the pitch angle deviation of the blade may include: modeling a relationship between wind speed and a difference in a given pitch angle deviation of the at least one blade; and comparing the modeled relationship with the determined relationship to estimate the pitch angle deviation of the at least one blade.

[0027] In some embodiments, during the first operating mode, the wind turbine may be operated at partial load; and during the second operating mode, the wind turbine may be operated at full load.

[0028] In some such embodiments, determining the calibration parameter may include binning the measurements of the first blade load measurement for each blade according to the wind speed associated with the respective measurement.

[0029] In some embodiments, identifying differences between the calibrated blade load measurements for the blades may include binning the calibrated blade load measurements for each blade according to a pitch angle associated with each measurement.

[0030] In such an embodiment, identifying a difference between the calibrated blade load measurements for the blades may further comprise determining a relationship between the pitch angle and the difference between the calibrated blade load measurement for the at least one blade and the target blade load.

[0031] In some such embodiments, the relationship between the pitch angle and the difference may be the rate of change of blade loading as a function of the pitch angle.

[0032] In some embodiments, estimating the pitch angle deviation of the blade may include:

[0033] modeling a relationship between a pitch angle and a difference for a given pitch angle deviation of the at least one blade; and

[0034] The modeled relationship is compared to the determined relationship to estimate a pitch angle deviation of the at least one blade.

[0035] A second aspect of the present invention provides a wind turbine comprising:

[0036] a plurality of blades, each blade being associated with a corresponding blade load sensor; and

[0037] a pitch adjustment mechanism configured to adjust the pitch of each of the blades; and

[0038] a control system configured to receive blade load measurements from the blade load sensors and to control the pitch adjustment mechanism;

[0039] Wherein, the control system is configured to execute the method of any embodiment of the first aspect to correct the pitch angle deviation of at least one of the plurality of blades. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Embodiments of the present invention will now be described with reference to the accompanying drawings, in which:

[0041] Figure 1 is a schematic diagram of a wind turbine;

[0042] Figure 2 is a schematic diagram of a control system for a wind turbine;

[0043] Figure 3 A method of controlling a wind turbine to correct for pitch angle deviation is shown;

[0044] Figure 4 Shows that it is possible to use Figure 3 The increase in energy production achieved by the method shown;

[0045] Figure 5a -d shows a calibration step of one embodiment of the method;

[0046] Figure 6a -d shows the pitch correction step of this embodiment;

[0047] Figure 7a -b shows a calibration step of another embodiment of the method; and

[0048] Figure 8a -c shows the pitch correction step of this other embodiment. DETAILED DESCRIPTION

[0049] Figure 1 An example of a wind turbine 100 is shown in a schematic perspective view. The wind turbine 100 includes a tower 102, a nacelle 103 at the apex of the tower, and a rotor 104 operatively coupled to a generator housed in the nacelle 103. In addition to the generator, the nacelle also houses various components required to convert wind energy into electrical energy and various components required to operate, control and optimize the performance of the wind turbine 100. The rotor 104 of the wind turbine includes a central hub 105 and a plurality of blades 106 protruding outward from the central hub 105. In the illustrated embodiment, the rotor 104 includes three blades 106, but the number can vary. In addition, the wind turbine includes a control system. The control system can be placed in the nacelle or distributed at multiple locations within the turbine and communicated.

[0050] The wind turbine 100 may be included in a group of other wind turbines belonging to a wind power plant (also referred to as a wind farm or wind park) that serves as an electric power generation plant connected to an electric grid via transmission lines. The electric grid typically consists of a network of power generating stations, transmission circuits, and substations connected by a network of transmission lines that transmit electricity to loads in the form of end users and other customers of the electric utility.

[0051] Figure 2 An embodiment of a control system 200 and elements of a wind turbine are schematically shown. The wind turbine includes rotor blades 106, which are mechanically connected to a generator 202 via a gearbox 203. In direct drive systems and other systems, the gearbox 203 may not be present. The electricity generated by the generator 202 is injected into the grid 204 via an electrical converter 205. The generator 202 and the converter 205 may be based on a full-scale converter (FSC) architecture or a doubly fed induction generator (DFIG) architecture, but other types may also be used.

[0052] The control system 200 includes a plurality of elements, including at least one main controller 220 having a processor and a memory, so that the processor can perform computing tasks based on instructions stored in the memory. In general, the wind turbine controller ensures that the wind turbine generates the requested power output level in operation. This is achieved by adjusting the pitch angle of the blades 106 and / or the power extraction of the converter 205. To this end, the control system includes: a pitch system, which includes a pitch controller 207 using a pitch reference 208; and a power system, which includes a power controller 209 using a power reference 206. The wind turbine rotor includes rotor blades that can be pitched by a pitch mechanism. The rotor includes a separate pitch system that enables the rotor blades to be pitched individually, and may include a common pitch system that adjusts all pitch angles of all rotor blades simultaneously. The control system or elements of the control system can be placed in a power plant controller (not shown) so that the turbine can be operated based on instructions provided externally.

[0053] The control system 200 also includes a plurality of sensors 230 (for clarity, Figure 2 Only one sensor is shown in FIG. 1 ). The sensor 230 measures the load on each blade 106. In particular, the sensor 230 may measure the root bending moment of each blade 106.

[0054] When turbine 100 is initially constructed, the pitch of each blade 106 is manually aligned. This may result in a difference between the pitch angle reported by control system 200 for blade 106 and the actual aerodynamic pitch of the blade. Control system 200 uses the pitch to optimize the performance of turbine 100 for the current wind conditions. Any mismatch between the actual pitch of blades 106 and the reported pitch may cause turbine 100 to operate suboptimally, thereby reducing the amount of energy that can be extracted from the wind.

[0055] Figure 3 A method 300 of correcting for deviations in the pitch angle is shown. The method 300 allows for automatic correction of pitch deviations while the turbine is running. It does not require manual re-inspection of the pitch position of the blades and avoids the problems associated with detecting pitch deviations caused by rotor imbalance.

[0056] At step 301, a first blade load measurement for each blade 106 of the turbine 100 is provided. The first blade load measurement is a measurement from when the wind turbine is operating in a first operating mode, such as idling or a partial load region. The first blade load measurement may be a recent measurement, or may be provided from a data store. For example, during an initial startup period of the turbine 100, the first blade load measurement may have been recorded and stored for later use. The first blade load measurement may be a measurement of the root bending moment experienced by each blade 106.

[0057] At step 302, calibration parameters are determined to correct for differences between first blade load measurements of the blades. This step ensures that observed blade load differences are not due to the different blade load sensors 230 themselves. For example, gain multipliers and offset values ​​may be determined for sensors 230 associated with each blade 106. The calibration parameters may be determined using an average value (e.g., mean or median) of the first measurements, or by calibrating the first measurements for a predetermined master blade.

[0058] At step 303, a second blade load measurement is provided for each blade. The second measurement is measured when the wind turbine is operating in a second operating mode. The second operating mode is an operating region of the wind turbine 100 at higher wind speeds than the first operating mode. For example, where the first operating mode is part load operation, the second operating mode may be full load operation. The second blade load measurement may be provided from a data store.

[0059] At step 304, the calibration parameters are applied to the second blade load measurements to determine a calibrated blade load measurement. In particular, the calibration parameters of sensor 320 associated with blade 106 may be applied to the second blade load measurement for that blade. This step substantially eliminates systematic errors due to sensor 320 itself from the second blade load measurement. Any remaining differences between the calibrated blade load measurements of the blades may be related to pitch angle differences.

[0060] At step 305, a difference between the calibrated blade load measurements of the blades 106 is identified. In particular, identifying the difference between the calibrated blade load measurements may include: determining a target blade load; and identifying a difference between the calibrated blade load measurement of the at least one blade and the target blade load. The target blade load may, for example, be an average of the calibrated blade load measurements, or may be a blade load experienced by a predetermined main blade. The target blade load may vary as a function of operating conditions of the wind turbine, for example as a function of wind speed or rotation angle. Comparing the difference to the target blade load provides a common reference point for assessing differences between blades.

[0061] At step 306, a pitch angle deviation of at least one blade 106 contributing to the identified difference is estimated. In embodiments where a difference in blade load is identified relative to a target blade load, estimating the pitch angle deviation may specifically include estimating the pitch angle deviation of the at least one blade contributing to the difference between the calibrated blade load measurement for the blade and the target blade load.

[0062] At step 307 , the pitch angle of at least one of blades 106 is changed to correct the pitch angle deviation.

[0063] It should be noted that steps 301-302 in which calibration parameters are determined may be performed at a separate time from the remaining steps 303-307. For example, the calibration steps 301-301 may be performed during an initial commissioning period of the wind turbine 100. Then, the pitch correction steps 303-307 may be performed after a period of time, such as a period of months or years later.

[0064] Thus, the method 300 corrects the pitch angle deviation based on the differences between the blades, rather than individually correcting each blade based on its absolute position. This may mean that an individual perfectly aligned blade is slightly misaligned by this method. However, it has been recognized that by correcting for the observed differences between the blades, the overall energy output of the turbine can be increased even if individual blades become misaligned.

[0065] The increase in total energy production that may be achieved by method 300 is Figure 4 Shown in. Figure 4 Typical annual energy production (AEP [MWh]) for a given wind turbine type 100 is shown as a function of pitch error (for each blade). Points 401-403 show the initial pitch error of the blades 106 of the turbine 100. As shown by points 401 and 403, two of the blades are misaligned (-1° and 0.3°, respectively). The blade represented by point 402 is initially perfectly aligned.

[0066] In the right Figure 4 In a particular embodiment of method 300 performed for blades 106 represented in FIG. 4 , each blade is adjusted to correct for the difference between its calibrated blade load and an average calibrated blade load (i.e., the target blade load is the average blade load). The average value may be the median or average of the calibrated blade loads. As a result of this correction, the pitch of each blade 106 (including the initially perfectly aligned blade represented by point 402) is adjusted. The new pitch error for each blade is given by Figure 4 It is indicated by point 404 in FIG. 1 . From this, it can be seen that the total energy produced by the three blades together is increased by adjusting to the average value.

[0067] In an alternative embodiment, instead of correcting each blade to an average value, one of the blades 106 may be identified as a master blade. The calibrated blade load of the master blade is then used as the target blade load. The pitch of each of the remaining blades 106 is adjusted to eliminate the difference between the calibrated blade loads of those blades and the calibrated blade load of the master blade. For example, it may be assumed that the first blade 106 is correctly aligned and is therefore identified as the master blade. The other blades 106 of the turbine 100 are then corrected to match the master blade. This may allow for a quick initial calibration of the pitch of the blades 106 of the turbine 100. Extra care may be taken when initially aligning the master blade to ensure that the other blades are corrected to match the well-aligned blade.

[0068] In method 300, step 306 of estimating the pitch angle deviation of the blade may include: modeling the difference in the expected blade load measurement results of the blade for a given pitch angle deviation of the blade; and comparing the modeled difference with the identified difference to estimate the pitch angle deviation of the at least one blade. For example, the expected blade load difference can be modeled for pitch angle deviations in increments of 0.5° or in increments of 1°. For example, the expected blade load difference can be modeled for pitch angle deviations of -2°, -1°, 0°, 1°, and 2°. Then, comparing the modeled difference with the identified difference may include: selecting the modeled difference that is closest to the identified difference (or closest in behavior to the identified difference); and adjusting the pitch angle based on the given pitch angle of the selected modeled difference.

[0069] The model may, for example, relate the operating conditions of the turbine to the predicted blade loads of the blade. The model may include a lookup table. For example, the model may be the thrust coefficient (C t ) table that relates the thrust coefficient to the pitch and tip speed ratio. The tip speed ratio is the ratio of the tangential speed at the tip of the blade to the wind speed.

[0070] In some embodiments, the step 306 of estimating the pitch angle deviation of the blade 106 may also include: estimating the magnitude and / or direction of the pitch angle adjustment required to correct the identified difference based on the comparison of the modeled difference with the identified difference. In other words, it is not necessary to extract the exact value of the pitch deviation. Instead, the direction and / or magnitude of the required correction can be estimated from the modeled difference, and the pitch of one or more blades 106 can then be adjusted based on the estimate. Then, the correction steps 302-307 of the method 300 can be re-executed to test whether the adjustment has reduced the pitch angle deviation.

[0071] In order to minimize the possibility of adjusting the pitch of the blade in a manner that reduces power production, the pitch angle can be adjusted incrementally, for example, with a step size between 0.5° and 1° or between 0.2° and 0.5°. After the incremental adjustment, the correction steps 302-306 of method 300 can be performed again using the new second measurement result to see whether the incremental adjustment has improved or worsened the pitch angle deviation of blade 106. If the pitch angle deviation has been reduced, step 307 can be performed again to adjust the pitch angle of at least one blade by further increments. If the pitch angle deviation increases, the pitch of the blade can be reset to their initial position. These method steps can be repeated until there is no pitch deviation between the blades, or until the estimated blade load deviation is less than 0.5° or less than 0.2°. As the pitch angle deviation is gradually reduced, the size of the increment used can be reduced. Alternatively, the increment can be implemented as a ratio (or gain) of the estimated pitch angle deviation. For example, the ratio can be between 0.3 and 0.8. The adjusted pitch positions of the blades may be maintained for a test period, during which the energy output of turbine 100 may be evaluated. If the output decreases, the pitch positions of the blades may be reset to their initial positions.

[0072] 5-8 illustrate specific implementations of the method 300 .

[0073] 5 and 6 illustrate the calibration step and the correction step, respectively, of a particular embodiment of the method 300. In this embodiment, a first measurement is made when the turbine is idling (e.g., wind speed is less than 4 m / s). A second measurement is made when the turbine is at partial load (e.g., wind speed is between 5 and 9 m / s).

[0074] Figure 5a The blade load experienced by each blade 106 during idling operation as the blade rotates in the rotor plane is shown, plotted against the angular position of the blade in the rotor plane at the time of measurement (i.e., the azimuthal position of the blade). At idling, the load experienced by the blade is primarily due to gravity. This results in periodic variations in the blade load, such as during Figure 5a As can be seen in . Since each blade has approximately the same mass, each blade should experience approximately the same load change. Any differences between the measured blade loads may be due to systematic errors in the sensors 230 measuring the blade loads.

[0075] To determine the difference between the measured blade loads, the measurements are averaged and the difference between the measurement and the average is determined for each blade. Figure 5b Shown with Figure 5aThe difference between the average value (in this case the mean) of the blades x. The calibration parameters are then determined for each blade. In particular, the gain value G is calculated for each blade x. x and offset value O x (where x can be Figure 5a Leaf A, leaf B or leaf C as indicated in the figure to substantially eliminate the difference from the average for that leaf.

[0076] The calibrated blade load measurement for blade x is then BladeLoadx Calibrated The following formula may be used to calculate the raw blade load measurement BladeLoadx measured by the corresponding sensor 230:

[0077] Blade Loadx Calibrated = Bla deLoadx×G x +O x

[0078] In this embodiment, the average is determined as a function of the angular position in the rotor plane.In an alternative example, an average peak-to-peak blade load value may be determined and the calibration parameters generated based on the difference between the individual peak-to-peak loads and the average peak-to-peak.

[0079] To illustrate the effect of the calibration parameters, Figure 5c shows that when the corresponding calibration parameters are applied Figure 5a The measurement results. Figure 5d The differences between these calibrated measurements and their average are shown. Figure 5d and Figure 5b In comparison, it is clear that the calibration parameters significantly reduce the difference between the individual blade loads and the average blade load over all rotor angles.

[0080] Having determined the calibration parameters for each blade in the idling region, the method now continues by correcting the pitch angle deviation using the measurements from the part load region.

[0081] Figure 6a The blade load experienced by each of the blades 106 is shown, plotted against the wind speed experienced by the turbine 100 when the corresponding measurement was taken. The calibration parameters have been applied to Figure 6a The apparent differences between the blade loads experienced by the different blades may be due to pitch angle deviations.

[0082] In this embodiment, the pitch of each blade is modified based on the average blade load, such as Figure 4 For this purpose, the difference between each calibrated blade load and the average blade load is determined as a function of wind speed. The resulting difference is Figure 6b shown.

[0083] The method then proceeds to estimate the pitch angle deviation that will result in a difference between each calibrated blade load and the mean value using a model relating the operating parameters to pitch and blade loading. For example, a thrust coefficient table may be used.

[0084] Figure 6c The estimated blade loads experienced by blade 106 as a function of wind speed are shown for pitch errors of 1°, 0°, and -1°. The blade loads for other given pitch error values ​​can also be calculated, but are not shown for clarity. Figure 6c For example, the blade loads may be calculated for pitch errors in the range of -5° to 5° or -2° to 2° for 0.5° increments or 1° increments.

[0085] Figure 6d Shows Figure 6c The difference between the estimated blade loads and the modelled blade loads for a blade without pitch error. It can be seen that the blade loads produced by different pitch errors behave differently as a function of wind speed. In particular, the gradient of the difference as a function of wind speed is positive for a pitch error of +1° and negative for a pitch error of -1°. The absolute value of the difference between 1° and -1° is about 5%.

[0086] The modeling differences can be compared with Figure 6b The difference between the calibrated measurements shown in is compared to estimate the actual pitch error of the blade. For example, the absolute value of the difference and the gradient of the difference can be compared. For example, Figure 6b Two of the blades have a difference of approximately 5% and a positive gradient. Therefore, the pitch error of these blades can be estimated to be approximately +1°. The remaining blade has a difference of approximately 10% in absolute value and a negative gradient. Assuming a roughly linear relationship between blade load and pitch error, the pitch error of this blade can be estimated to be approximately -2°. The pitch of each blade can then be adjusted to correct its corresponding estimated pitch error.

[0087] In the above example, the pitch angle is specifically estimated using the absolute value of the difference between the calibrated blade load and the average; and the gradient of the difference relative to the wind speed. However, in general, the step 305 of identifying the difference between the calibrated blade load measurements may include: determining any relationship between the wind speed and the difference between the calibrated blade load measurements and the target / average blade load of the blade. The step 306 of estimating the pitch angle deviation of the blade may include: modeling the relationship between the wind speed and the difference between the given pitch angle deviation of the at least one blade; and comparing the modeled relationship with the determined relationship to estimate the pitch angle deviation of the at least one blade. The relationship may be, for example, an approximate relationship, or the slope of a best fit line of binned measurements. The relationship may be determined over a sub-portion of the measured wind speed.

[0088] In order to provide an accurate value of the blade load, the measurement of each blade may be repeated a number of times. For example, the measurements may be taken within a predetermined number of days (e.g. between 3-5 days), or as long as the turbine remains within the operating region required for these measurements. The first / second measurement results may include all recorded measurements, binned with respect to rotor position / wind speed; or may include an average value of the blade load at a specific rotor position / wind speed.

[0089] The embodiments shown in FIGS. 5 and 6 use an idling period to calibrate the blade load measurements. This method is particularly useful during the initial startup of the turbine 100, when restarting the turbine after a shutdown, or when there are periods of very low wind. However, some turbines (particularly offshore turbines) may only be operated very rarely during idling periods. FIGS. 7 and 8 show alternative embodiments of the method 300 that do not require the turbine to be idling.

[0090] 7 and 8 illustrate the calibration step and the correction step, respectively, of an alternative embodiment of method 300. In this embodiment, a first measurement is taken when the turbine is at partial load (e.g., wind speeds between 5 m / s and 9 m / s - when the turbine 100 is operating at a fixed pitch angle). A second measurement is taken when the turbine is at full load (e.g., wind speeds above 12 m / s - when the turbine 100 is operating at a fixed rotor speed). Full load means that the turbine is operating at its rated power. Partial load is when power is being generated, but operating below rated power. It is best not to use measurements taken at wind speeds between 9 m / s and 12 m / s, as this is a transition region between speed control operation and pitch control operation.

[0091] Figure 7aA first measurement of blade load for each blade 106 (labeled in the figure as blade A, blade B and blade C, respectively) measured during part load operation is shown. The blade loads are binned as a function of the wind speed at the time of the measurement. As described above with respect to FIGS. 5 and 6 , the blade loads may be taken over a longer period of time and all measurements binned with respect to wind speed, or an average blade load is found for a particular wind speed value.

[0092] Similar to the process described above with respect to FIG. 5 , the measurements are calibrated by finding calibration parameters that eliminate the difference between the first measurement of each blade and the average of the first measurements. Figure 7b Shows Figure 7a Measurement results with the calibration parameters applied. It can be seen that the differences between the blade loads in the original data have been eliminated by the calibration parameters.

[0093] Having determined the calibration parameters, the method continues by correcting the pitch error based on observed blade load differences in the second operating mode (ie, full load operation).

[0094] Figure 8a The blade load measurements for each blade 106 in the second operating mode are shown, in which the calibration parameters have been applied. In full load operation, pitch control is used to maintain the power output of the turbine 100. Therefore, the blade load for each blade 106 is plotted as a function of the pitch angle of that blade when the measurement is made. In alternative embodiments, the blade loads may instead be binned as a function of wind speed, but using pitch instead of wind speed may provide a more accurate result.

[0095] because Figure 8a The measurements shown in have been calibrated so that differences between the loads on each blade are most likely caused by pitch errors. Figure 8b The difference between the calibrated blade load of each blade and the average blade load as a function of the pitch angle is shown. In this case, the difference is the ratio of the individual blade load to the average blade load of each blade.

[0096] As mentioned previously, the pitch error for each blade can be estimated by comparing the observed difference from the mean with the modeled difference. In particular, the blade load can be estimated using the thrust coefficient C t The thrust coefficient C t The table relates pitch angle and tip speed ratio to blade loads. This model is used to determine the blade loads for a given value of pitch error. t The thrust coefficient determined from the table can itself be used as a proxy for estimating blade loading in this method. Alternatively, the estimated blade loading can be derived from the thrust coefficient. Deriving the blade loading can provide a more accurate result.

[0097] Figure 8c shows the C for pitch errors of +1° and -1° t The modeled C ratio is the ratio of the modeled thrust coefficient for a particular pitch error value divided by the modeled thrust coefficient for a blade with zero pitch error. The modeled C ratio can also be determined for other pitch error values. t ratio, but for clarity not shown in Figure 8b For example, blade load / C t The value can be calculated for pitch errors in the range of -3° to 3° or -2° to 2° with 0.5° increments or 1° increments. In this example, as described above, C t is used as a proxy for the estimated blade load, but in other examples the blade load can be derived from C t The value is derived and the modeled blade load ratio is used instead of C t ratio.

[0098] The pitch deviation of each blade is calculated by dividing the measured difference (i.e. Figure 8b those shown in ) and the modeled blade loads / C t The modeled blade load ratio that best matches the measured difference of the blade is identified, and the pitch error that produces the modeled ratio is considered to be the pitch error of the blade. In particular, the gradient of the modeled ratio as a function of the pitch angle may be compared with the gradient of the measured difference. Alternatively, the gradient of the modeled ratio as a function of the pitch angle may be determined, such as Figure 8c As shown. It can then be assumed that the magnitude of the gradient is linearly proportional to the pitch error. By comparing the magnitude of the gradient of the measured difference with the magnitude of the modeled gradient, the value of the pitch error can be determined. The direction of the pitch error (i.e., + or -) can be determined from the direction of the gradient. For example, if the measured gradient is 0.7 times the modeled gradient, then the pitch error can be assumed to be 0.7 times the modeled pitch error value.

[0099] Once a pitch error is identified, the pitch of each blade 106 is adjusted to correct its respective pitch error. As described above, the pitch of the blades may be adjusted incrementally so that the performance of the turbine in a new pitch arrangement may be evaluated before further adjusting the pitch of the blades.

[0100] Although the invention has been described with reference to one or more preferred embodiments, it will be appreciated that various changes or modifications may be made without departing from the scope of the invention as defined in the appended claims.

Claims

1. A method for correcting a pitch angle deviation of a blade of a wind turbine, the wind turbine comprising a plurality of blades, the method comprising: providing a first blade load measurement for each blade of the plurality of blades, the first blade load measurement being measured while the wind turbine is operating in a first operating mode; determining calibration parameters to correct for differences between first blade load measurements of the blade; providing a second blade load measurement for each blade, the second blade load measurement being measured when the wind turbine is operating in a second operating mode, wherein the wind turbine experiences lower wind speeds during the first operating mode than during the second operating mode; applying the calibration parameter to the second blade load measurement to determine a calibrated blade load measurement; identifying differences between calibrated blade load measurements of the blades; estimating a pitch angle deviation of at least one blade causing the identified discrepancy; and changing the pitch angle of the at least one of the blades to correct the pitch angle deviation, Identifying differences between calibrated blade load measurements involves: determining a target blade load; and identifying a difference between a calibrated blade load measurement of the at least one blade and the target blade load; and Estimating the pitch angle deviation includes estimating a pitch angle deviation of the at least one blade that causes a difference between a calibrated blade load measurement of the blade and the target blade load.

2. The method according to claim 1, wherein: Determining the target blade load includes determining an average of calibrated blade load measurements.

3. The method according to claim 2, wherein: A pitch angle deviation is estimated for each blade of the wind turbine, and wherein the pitch angle of each of the blades is adjusted to correct the respective pitch angle deviation.

4. The method according to claim 1, wherein: Determining the target blade load includes: Identifying a blade as a primary blade; and A calibrated blade load measurement of the primary blade is set as the target blade load.

5. The method according to any one of claims 1 to 4, wherein: Estimating the pitch angle deviation of the blades includes: modeling differences in expected blade load measurements for the blade for a given pitch angle deviation of the blade; and The modeled difference is compared to the identified difference to estimate a pitch angle deviation of the at least one blade.

6. The method according to claim 5, wherein: Estimating the pitch angle deviation of the blades also includes: A magnitude and / or direction of a pitch angle adjustment required to correct the identified difference is estimated based on a comparison of the modeled difference with the identified difference.

7. The method according to any one of claims 1 to 4, wherein: Changing the pitch angle of the at least one of the blades to correct the pitch angle deviation comprises: incrementally changing the pitch angle of the at least one blade; and The pitch angle deviation of the at least one blade is re-estimated.

8. The method according to any one of claims 1 to 4, wherein: During said first operating mode, the wind turbine operates at idle speed; and During the second operation mode the wind turbine is operated at part load or full load.

9. The method according to any one of claims 1 to 4, wherein: Determining the calibration parameter comprises binning the measurement results of the first blade load measurement for each blade according to the rotation angle of the blade associated with the respective measurement result.

10. The method according to any one of claims 1 to 4, wherein: Identifying differences between the calibrated blade load measurements for the blades includes binning the calibrated blade load measurements for each blade according to a wind speed associated with each measurement.

11. The method according to any one of claims 1 to 4, wherein: During said first operating mode, the wind turbine operates at part load; and During the second operating mode, the wind turbine operates at full load.

12. The method according to claim 11, wherein: Determining the calibration parameter comprises binning the measurement results of the first blade load measurement for each blade according to the wind speed associated with the respective measurement result.

13. The method according to claim 11, wherein: Identifying differences between the calibrated blade load measurements for the blades includes binning the calibrated blade load measurements for each blade according to a pitch angle associated with each measurement.

14. A wind turbine comprising: a plurality of blades, each blade being associated with a corresponding blade load sensor; and a pitch adjustment mechanism configured to adjust the pitch of each of the blades; as well as a control system configured to receive blade load measurements from the blade load sensors and to control the pitch adjustment mechanism; Wherein, the control system is configured to perform the method of any preceding claim to correct a pitch angle deviation of at least one of the plurality of blades.

Images