Unbalance estimation of a wind rotor of a wind turbine

Abstract

A method for compensating an imbalance (344, 345) of a wind rotor (110) of a wind turbine (100) is described. The method comprises applying at least one test offset to a parameter (113, 117) characterizing a state of the wind rotor (110), wherein the imbalance (344, 345) depends on the parameter (113, 117); measuring an acceleration of the wind turbine (100) for the at least one test offset, wherein the acceleration depends on the imbalance (344, 345); and determining a compensation offset (341-343, 874-876) based on the measured acceleration, wherein the imbalance (344, 345) is at least partially compensated when the compensation offset (341-343, 874-876) is applied to the parameter (113, 117). Furthermore, a corresponding wind turbine (100) is described, which comprises a control unit (101) configured to perform the above-described method.

Description

Technical Field

[0001] This invention generally relates to the technical field of wind turbines. Specifically, the invention relates to a method for compensating for imbalances in a wind turbine rotor and a wind turbine including a control unit configured to perform this method. Background Technology

[0002] Many turbines operate with mass and pitch imbalances. This imbalance increases fatigue loads on the turbine and its base. A balanced rotor reduces these fatigue loads, thereby reducing the cost of the turbine and its base.

[0003] Traditionally, offline tools have been used to estimate mass imbalances. However, such tools are incapable of pitch balancing and require several hours of turbine operating data. Pitch calibration is typically performed based on markings in the blades and hub or using photographic methods.

[0004] Therefore, a rapid and effective method for compensating for various imbalances in the wind turbine rotor may be needed, as well as a corresponding wind turbine with a control unit configured to perform this method. Summary of the Invention

[0005] This need can be met by the subject matter of the basic scheme according to the invention. Advantageous embodiments of the invention are described by preferred embodiments of the invention.

[0006] According to a first aspect of the invention, a method for compensating for imbalance in a wind turbine rotor is described. The method includes applying at least one test offset to a parameter characterizing the state of the wind turbine rotor, wherein the imbalance depends on the parameter. The method further includes measuring the acceleration of the wind turbine for the at least one test offset, wherein the acceleration depends on the imbalance. The method further includes determining a compensation offset based on the measured acceleration, wherein when the compensation offset is applied to the parameter, the imbalance is at least partially compensated.

[0007] According to another aspect of the invention, a wind turbine for generating electricity includes a tower, a wind turbine rotor, and a motor configured as a generator, the wind turbine rotor being disposed at the top portion of the tower and including at least one blade, the motor being mechanically coupled to the wind turbine rotor. The wind turbine also includes a control unit configured to perform the methods described herein.

[0008] In the context of this application, a "wind turbine" is any device configured to convert wind energy into electrical energy, particularly for distribution to the power grid and / or for local energy supply. A wind turbine may include a tower, wind turbine rotor, nacelle, shaft (particularly a low-speed shaft and / or a high-speed shaft), gearbox, generator, braking assembly, pitch drive, and / or yaw drive. The wind turbine rotor may be a rotor assembly and may include one or more blades and a central hub to which the blades are attached. The wind turbine rotor may include one, two, three, four, five, or more than five blades. The wind turbine rotor may also include a pitch bearing connecting the rotor hub to the blades. Using the pitch bearing, the blade orientation can be adjustable, particularly the pitch angle that determines the angle of attack of the blades.

[0009] An “imbalance” in a wind turbine rotor can be any kind of asymmetry in the rotor, particularly manifesting itself during the operation of the wind turbine, especially during the rotation of the rotor. Imbalance can be any asymmetry that deviates from the symmetry of the rotor as defined by its axis of rotation and / or the arrangement of the rotor blades. Imbalance can be static or dynamic.

[0010] Imbalance can be caused by an asymmetrical distribution of rotor mass, particularly by a distribution where the rotor's center of gravity is not located on the rotor's axis of rotation. Mass imbalance can also result from uneven mass distribution within the blades, especially from different mass distributions on different blades of the wind turbine rotor. Mass imbalance can be defined by a virtual mass, particularly a virtual point mass, located at a certain distance from the wind turbine rotor's axis of rotation. This virtual mass can rotate at the same speed as the wind turbine rotor.

[0011] Mass imbalance in a wind turbine rotor can be caused by factors such as blade icing, blade tip wear, or manufacturing tolerances. This imbalance can lead to centrifugal forces on the rotor corresponding to the rotating virtual mass. This can cause oscillations in the wind turbine, particularly the rotor, nacelle, and / or tower. These oscillations can have the same frequency as the rotating virtual mass and / or the rotor. The oscillations can be left-right, but can also be, for example, front-back.

[0012] Imbalance can be caused by asymmetry in the aerodynamic properties of the wind turbine rotor, particularly by differences in the aerodynamic properties between different blades of the rotor. For example, imbalance can be pitch imbalance, where different blades have different pitches, especially different pitch angles.

[0013] For example, the pitch angles of different blades can differ by an order of 1°. Imbalance can also be due to shape differences between different blades, such as blade deformation, which may be caused, for example, by mechanical shocks and / or wear on the blades. In addition, aerodynamic imbalance can be caused by changes in the surface properties of individual blades, for example, due to icing. Furthermore, imbalance can be caused at least in part by asymmetry in components of the wind turbine other than the wind turbine rotor (e.g., the shaft).

[0014] Aerodynamic imbalances can lead to asymmetric forces on a wind turbine rotor, for example, due to differences in the angle of attack of different blades and / or differences in blade profiles. This force imbalance can cause oscillations in the wind turbine, particularly the rotor, nacelle, and / or tower. The frequency of the oscillations can correspond to the rotational frequency of the rotor. The oscillations can be particularly back-and-forth, but can also be, for example, left-and-right.

[0015] Imbalance can be torque imbalance, for example, measured in kg*m. Torque imbalance can capture the magnitude of mass imbalance and / or equivalent pitch imbalance. It can capture the magnitude of total imbalance. The corresponding phase imbalance can be correlated with torque imbalance.

[0016] The parameter characterizing the state of the wind turbine rotor can be any property or characteristic of the wind turbine rotor. It can be a changeable property of the wind turbine rotor, particularly a controllable one. This parameter can be changed and / or controlled by the operator and / or the wind turbine's control unit. During wind turbine operation, especially during energy generation, the parameter can be variable and / or controllable. Alternatively, it can be variable and / or controllable when the wind turbine is not in operation, such as during maintenance or repair. This parameter can be controllable to determine the wind turbine's energy yield.

[0017] The parameter could be, for example, the mass distribution of the wind turbine rotor, particularly the mass distribution of the blades. This parameter could indicate the presence or absence of one or more counterweights on each blade of the wind turbine, specifically counterweights located at a predetermined distance from the axis of rotation of the wind turbine rotor. Such counterweights could, for example, have a mass of 18 kg. This parameter could also indicate the presence of one or more counterweights of variable mass at variable distances from the axis of rotation of the wind turbine.

[0018] Alternatively or alternatively, the parameters may relate to the aerodynamic properties of the wind turbine rotor. For example, the parameters may represent the pitch angle of each blade of the wind turbine. Alternatively or additionally, the parameters may represent the geometric profile of each blade of the wind turbine. Further alternatively or additionally, the parameters may represent the surface properties of each of the blades of the wind turbine rotor, such as the surface roughness distribution on each of the blades of the wind turbine rotor.

[0019] Parameters can be scalars, vectors, or higher-dimensional quantities such as matrices or tensors. Parameters may include several sub-parameters; for example, each sub-parameter might characterize a different blade on a wind turbine rotor. Each sub-parameter itself can be a scalar, vector, or higher-dimensional quantity such as a matrix or tensor. Parameters can be time-dependent, for example, time-dependent in a periodic manner, where the period can be determined by the rotation (number of revolutions) of the wind turbine.

[0020] The “offset” of a parameter can be a difference added to that parameter. This difference can be positive, negative, and / or zero. This difference can be defined relative to a base value for the parameter. The base value can be any value of the parameter used for comparison. The base value can correspond to the current value of the parameter or a previous value of the parameter. The base value can correspond to a parameter value calculated or modeled, for example, by the wind turbine’s control unit based on conditions such as wind speed, wind direction, and / or desired energy output. The base value can be determined by the wind turbine’s operating mode, in which imbalances in the wind turbine rotor are not compensated. The base value can be determined relative to the wind turbine’s baseline operation, in which standard operating procedures unsuitable for a particular wind turbine are employed, such as those unsuitable for manufacturing tolerances or differences arising from wear on the wind turbine.

[0021] Here, the terms value sum and difference can be understood in a general sense as referring to a scalar, vector, or even higher-dimensional quantity, depending on the nature of the parameter. Each vector component or matrix term can be negative, positive, and / or equal to zero. Furthermore, the terms value sum and difference can refer to the sum and difference of the values ​​of different sub-parameters of a parameter, such as the sub-parameters characterizing the different blades of a wind turbine rotor.

[0022] Offset can be a discrepancy intentionally introduced, for example, by maintenance personnel and / or by a control unit configured to control the state of the wind turbine rotor. Offset can be constant, or it can vary over time, particularly periodically with a period determined by the rotational cycle of the wind turbine rotor.

[0023] The “test offset” for a parameter can be any offset applied to the parameter. Test offsets can be applied for testing and / or data collection purposes. Test offsets can correspond to intentionally introduced imbalances in the wind turbine rotor. Test offsets can be used to examine or map the wind turbine’s response to parameter variations, particularly to examine variations in imbalances that depend on said parameters.

[0024] Test offsets may include individual pitch angle offsets for each blade of the wind turbine rotor. Alternatively or additionally, test offsets may include individual balancing masses attached to each blade of the wind turbine rotor. Test offsets may be zero and / or may be different from zero. For example, test offsets, particularly mass test offsets, may include no balancing mass. The at least one test offset may be one, two, three, four, five, six, more than six, nine, more than nine, twelve, or more than twelve test offsets, which may be applied one after another. At least one test offset may be at least two test offsets, which may be applied one after another.

[0025] The “compensation offset” for a parameter can be any offset applied to the parameter. The compensation offset can be applied for compensation purposes, such as to at least partially compensate for an imbalance in a wind turbine. An imbalance can be at least partially compensated if any wear on the wind turbine is reduced due to the imbalance and / or if the loss of energy output from the wind turbine is reduced due to the imbalance. An imbalance can also be at least partially compensated if the measure of the imbalance is reduced. For example, if the corresponding acceleration of the wind turbine decreases or if the amplitude of the corresponding oscillation of the wind turbine decreases, the imbalance can be at least partially compensated.

[0026] Specifically, mass imbalance can be at least partially compensated if the virtual mass characterizing the mass imbalance decreases and / or if the virtual mass moves closer to the axis of rotation. As another example, aerodynamic imbalance can be at least partially compensated if the aerodynamic properties of the different blades are adjusted relative to each other. Aerodynamic imbalance can be compensated for by pitch adjustment. If the blade angles of the different blades are better aligned relative to each other, pitch imbalance, which can be considered an aerodynamic imbalance, can be at least partially compensated.

[0027] The “acceleration” of a wind turbine can be the acceleration of any component of the wind turbine, such as the tower, nacelle, or even the wind turbine rotor. Acceleration can be measured by one or more accelerometers mounted on these components of the wind turbine. Acceleration can be measured relative to the environment, particularly the base of the wind turbine. Acceleration can be a three-dimensional acceleration measured at one or more locations on the wind turbine, or it can include one or more components of the three-dimensional acceleration at these locations, such as components parallel to a plane or a given direction. Specifically, acceleration can include or may be the forward-backward acceleration of the wind turbine. Alternatively or additionally, acceleration can include or may be lateral acceleration. Here, lateral acceleration can be defined as perpendicular to the vertical axis of the tower but parallel to the rotor plane of the wind turbine rotor. Forward-backward acceleration can be defined as perpendicular to the vertical axis of the tower and perpendicular to the lateral acceleration of the tower. Alternatively, the forward-backward direction can be defined relative to the rotor plane, particularly perpendicular to the rotor plane.

[0028] "Measuring acceleration" may include sensing acceleration using a sensor device such as an accelerometer. It may also include processing the sensed acceleration. For example, measuring acceleration may include determining oscillations, particularly the frequency of the oscillations. As another example, measuring acceleration may include performing a Fourier analysis of the sensed acceleration. Measuring acceleration at at least one test offset may include measuring acceleration at zero test offset, i.e., against the baseline value of the parameter.

[0029] Measuring acceleration may include binning (for statistical analysis) the acceleration at at least one number of revolutions, particularly several, of the wind turbine rotor. It may include binning the acceleration according to the rotor azimuth angle. The acceleration may be binned at revolutions of one, two, three, four, five, more than five, ten, or more than ten. Measuring acceleration may include measuring the acceleration several times to determine the average, standard deviation, and / or confidence interval. It may include measuring the acceleration until the standard deviation and / or confidence interval are below a predetermined value and / or a predetermined maximum number of measurements are reached.

[0030] Determining the compensation offset based on measured acceleration may include calculating and / or modeling the compensation offset based on measured acceleration. The compensation offset can be determined solely based on the measured acceleration; that is, it is not necessary to measure any other quantities. When determining the compensation offset, other conditions or circumstances characterizing the wind turbine and / or the wind turbine environment, such as wind speed, wind direction, and / or the absolute values ​​of different blade pitch angles, may be considered.

[0031] The methods described above for compensating for imbalances in wind turbine rotors and corresponding wind turbines may be advantageous because imbalances can lead to increased wear or fatigue loads on the wind turbine, particularly on various components such as turbine blades, blade pitch control devices, or bearings in the drivetrain. Therefore, at least partially compensating for imbalances can result in longer wind turbine lifespans, longer maintenance intervals, and / or reduced tower and base costs, which are due to lower design constraints related to imbalance.

[0032] Furthermore, this method and the corresponding wind turbine can be advantageous because the compensation offset can be determined solely based on acceleration measurements, particularly nacelle acceleration measurements and / or rotor azimuth measurements using standard turbine sensors. Simultaneously, this method enables both pitch balancing and the calculation of the balancing mass required to balance the wind turbine. For example, optical measurements, such as optical measurements of blade angles, or photographic analysis may not be necessary.

[0033] Accordingly, this method is likely to be rapid, requiring only about an hour or less for the balancing process, while conventional processes require at least several hours. Furthermore, the balancing procedure, particularly the pitch balancing procedure, may be performed online during wind turbine operation, especially during energy generation.

[0034] Applying at least one test offset to a parameter characterizing the wind turbine rotor's state can facilitate a detailed examination of the relationship between the parameter and the imbalance, enabling the determination of a compensation offset with high precision, where the imbalance depends on that parameter. The test offset can be interpreted from the perspective of an explicit supplementary imbalance introduced in addition to the existing imbalance of the wind turbine rotor, particularly an existing imbalance that needs to be at least partially compensated. These supplementary imbalances allow for the collection of relevant information about the wind turbine's behavior that would not be available if only acceleration were measured without a test offset. Due to this additional information, the determined compensation offset will generally be more accurate and / or more effective in compensating for the imbalance compared to a compensation offset determined solely based on measured acceleration without a test offset.

[0035] According to an embodiment of the invention, determining the compensation offset includes determining the frequency components of the spectrum of the measured acceleration for each of at least one test offset, particularly the amplitude and / or phase of the frequency components.

[0036] Determining the frequency components may include performing a Fourier analysis of the acceleration being measured. This may include applying a Fourier transform to the measured acceleration, such as a continuous and / or discrete Fourier transform, like a fast Fourier transform. It may include applications of the Goertzel algorithm to the measurement. One frequency component may be determined for each of at least one test offset. The frequency component may be the same for each of the at least one test offset, i.e., corresponding to the same frequency, particularly the 1P frequency, or the frequency component may be different for different test offsets of the at least one test offset. More than one frequency component may be determined for at least one of the at least one test offset.

[0037] This implementation may be advantageous because the determined frequency components can be correlated with the oscillations of the wind turbine, which can be measured based on acceleration. The oscillations of the wind turbine can be oscillations at least in part determined by, for example, an imbalance in the wind turbine. The imbalance itself can manifest as one or a few oscillations of the wind turbine. Therefore, the imbalance can be adequately examined based on only one or a few frequency components corresponding to these oscillations. The Goertzel algorithm may be particularly effective for determining a single or a few frequency components of the measured acceleration.

[0038] According to an exemplary embodiment, determining the frequency component includes repeatedly determining the frequency component until certain characteristics of the frequency component, particularly the amplitude and phase or the real and imaginary parts, are known within a predetermined confidence interval. The frequency component can be repeatedly determined after a predetermined number of oscillations of the wind turbine and / or after a predetermined number of revolutions (e.g., ten revolutions) of the wind turbine rotor.

[0039] According to another embodiment of the invention, the frequency of each frequency component is based on, and in particular equal to, the rotor frequency of the wind turbine rotor, especially the current rotor frequency. Therefore, the frequency component can be a so-called 1P excitation or 1P frequency component. The rotor frequency can be any other frequency, such as a multiple or fraction of the rotor frequency.

[0040] This embodiment can be advantageous if, for example, the imbalance is due to a difference in properties between one blade and the others. Such an imbalance can cause oscillations in the wind turbine at the current rotor frequency. These oscillations can be optimally analyzed by determining the frequency component corresponding to the rotor frequency.

[0041] According to another embodiment of the present invention, determining the compensation offset includes mapping each frequency component to a coordinate system, determining a center defined by each frequency component, and calculating the compensation offset based on the center.

[0042] The coordinate system can be a two-dimensional coordinate system. In particular, the coordinate system can be a complex plane, and the frequency components can be complex numbers mapped onto the complex plane. The coordinate system can also be a one-dimensional coordinate system, a three-dimensional coordinate system, or even a coordinate system with more than three dimensions.

[0043] A single frequency component can be determined for each of at least one test offset. The center can be defined as the average value of each frequency component. This average value can be a weighted average, where different frequency components have different weights. They can also be assigned the same weight. The center can be defined as the centroid of a geometry defined by each frequency component. The geometry can be, for example, a polygon, triangle, hexagon, circle, or ellipse. The angles of the geometry can correspond to frequency components. One or more, and in particular all, frequency components can be located in the circumferential portion or periphery of the geometry.

[0044] The compensation offset calculation based on the center may include calculating the compensation offset based on at least one of the following: the distance of the center from the axis of the coordinate system, the distance of the center from the origin of the coordinate system, the coordinates of the center, the distance of the center from one or more frequency components, the average distance of the center from one or more frequency components, the orientation of one or more frequency components, the perimeter of the geometry, and the orientation of the geometry.

[0045] Such an implementation can be advantageous because it provides a simple and efficient process for determining the compensation offset. This process allows for the determination of the compensation offset with high accuracy, thereby largely compensating for the imbalance. This process can be particularly advantageous when combined with an appropriate selection of the test offset. For example, the test offset can be selected in a way that the frequency components exhibit a certain symmetry in the coordinate system and / or that they approximate a certain shape in the coordinate system.

[0046] According to another embodiment of the invention, determining the compensation offset includes fitting a circle based on each of the frequency components and calculating the compensation offset based on the circle.

[0047] A single frequency component can be determined for each of at least one test offset. Fitting a circle based on each frequency component can include mapping the frequency components to a coordinate system, particularly a two-dimensional coordinate system. It can include fitting the circle using least squares methods, particularly nonlinear least squares methods. It can include determining the fitted frequency components, each corresponding to a specific frequency component. The fitted frequency components can be located on the circumference of the circle, particularly equidistant from each other. For example, there can be six fitted frequency components corresponding to six frequency components, and the six fitted frequency components can be spaced 60° apart from each other on the circumference of the circle.

[0048] Calculating the compensation offset based on a circle can include calculating the compensation offset based on at least one of the following: the diameter of the circle, the coordinates of the center of the circle, the orientation of one or more frequency components, and the orientation of one or more fitted frequency components. There can be at least three frequency components, and in particular six frequency components, based on which a circle is fitted.

[0049] Such an implementation can be advantageous because it provides a simple and efficient process for determining the compensation offset. This process allows the compensation offset to be determined with high accuracy, thereby largely compensating for the imbalance. This process can be particularly advantageous when combined with an appropriate selection of the test offset. For example, the test offset can be selected such that, for instance, the frequency components approximate a circle when mapped to a coordinate system.

[0050] According to another embodiment of the invention, determining the compensation offset includes setting the compensation offset to one of at least one test offset, particularly a test offset corresponding to the frequency component with the minimum amplitude. The at least one test offset may include a test offset with a value of zero and / or at least one non-zero test offset.

[0051] Such an implementation can be advantageous because it requires minimal modeling or computation. Therefore, it can be particularly well-suited for online monitoring systems and / or for compensating for imbalances during wind turbine operation, especially during energy generation. This implementation can be particularly advantageous when combined with an iterative process, in which a compensation offset is determined based on a set of test offsets, and then further compensation offsets are determined based on another set of test offsets. Further iterations are possible, and given appropriate selections of the test offsets and additional test offsets, some convergence of the total compensation offset can be expected.

[0052] According to another embodiment of the invention, after determining the compensation offset, the method further includes applying the compensation offset to a parameter; applying at least one additional test offset to the parameter, each of the at least one additional test offset being less than a corresponding test offset; measuring the acceleration of the wind turbine against the at least one additional test offset; and determining another compensation offset based on the measured acceleration. Applying another compensation offset to the parameter can compensate for imbalances to a greater extent than applying only the compensation offset to the parameter.

[0053] If the absolute value of the other test offset is smaller, for example, if the other test offset and the test offset are scalars or vectors, then the other test offset may be smaller than the corresponding test offset. If at least one entry in the vector or matrix describing the other test offset is smaller than the corresponding entry of the test offset, then the other test offset may be smaller than the corresponding test offset. If another test offset of at least one sub-parameter of a parameter is smaller than the corresponding test offset of the sub-parameter, particularly for all sub-parameters, then the other test offset of the parameter may be smaller than the corresponding test offset of the parameter. Corresponding test offsets may be determined, for example, by the order in which they are applied and / or by their relative similarity. In an alternative embodiment, it is sufficient that at least one of at least one additional test offset is smaller than the corresponding test offset.

[0054] After determining the other compensation offset, the method may further include applying the other compensation offset to the parameter. It may also include determining and applying a third compensation offset based on at least one third test offset, and optionally, determining and applying a fourth compensation offset based on at least one fourth test offset. The method may include any number of iterations for determining subsequent compensation offsets, e.g., five, more than five, more than ten, or more than one hundred iterations. The number of iterations can be determined by the convergence of the total compensation offset and / or by the convergence of the compensation offset sequence to zero. The total compensation offset can be defined as the sum of the compensation offset sequences.

[0055] Such an implementation may be advantageous because it requires minimal modeling and computation. Therefore, it can be particularly well-suited for online monitoring systems and / or for compensating for imbalances during wind turbine operation, especially during power generation.

[0056] According to another embodiment of the invention, applying at least one test offset includes applying multiple test offsets sequentially, particularly applying multiple test offsets in a time-series order. Multiple test offsets can be applied one after another.

[0057] The multiple test offsets can consist of three test offsets, more than three test offsets, six test offsets, more than six test offsets, nine test offsets, or more than nine test offsets. The number of test offsets is divisible by the number of blades on the wind turbine rotor. Additionally, there may be test offsets with a value of zero.

[0058] This implementation may be advantageous because the test offset sequence allows for a thorough examination of the wind turbine's behavior during imbalance changes. This can allow for the determination of compensation offsets with high accuracy and efficient compensation for imbalances.

[0059] According to another embodiment of the invention, each of the plurality of test offsets includes a blade test offset for each of the plurality of blades of the wind turbine rotor, wherein the values ​​of the blade test offsets are arranged for different test offsets, and / or wherein the signs of the values ​​of the blade test offsets are reversed for different test offsets, in particular all signs.

[0060] Multiple test offsets may include test offsets for which the values ​​of the blade test offsets are arranged such that a corresponding blade test offset exists for each of the multiple blades. For example, in the case of three blades, there may be corresponding test offsets (a,b,b), (b,a,b), and (b,b,a), where the first term represents the blade offset of the first blade, the second term represents the blade offset of the second blade, and the third term represents the blade offset of the third blade, respectively. Similarly, with regard to sign reversal, there may be corresponding test offsets (a,b,c) and (-a,-b,-c), where the values ​​a, b, and c can be positive, negative, or zero.

[0061] Such an embodiment can be advantageous because it allows for a systematic examination of how the behavior of a wind turbine depends on the imbalance. This allows for the determination of compensation offsets with high accuracy and efficient compensation of the imbalance. When the corresponding frequency components are mapped to a coordinate system, the symmetry of the test offsets, involving arrangement and / or sign reversal, can be reflected. Therefore, the above embodiment may be particularly advantageous when determining the compensation offset based on the center of the frequency components and / or based on the fitted circle.

[0062] According to an exemplary embodiment, the number of test offsets is a multiple of the number of blades on the wind turbine rotor, specifically a multiple of twice the number of blades on the wind turbine rotor. This embodiment can reflect the symmetrical selection of test offsets as described above, wherein a factor of two is generated when paired test offsets with reversed signs are required.

[0063] According to an exemplary embodiment, for at least one of at least one test offset, and in particular all test offsets, the sum of the blade test offset values ​​is zero. This may be advantageous for introducing test offsets in a particularly balanced or symmetrical manner.

[0064] According to an exemplary embodiment, the compensation offset is a blade compensation offset, and the method further includes applying the blade compensation offset and determining the mass compensation offset when applying the blade compensation offset.

[0065] According to another embodiment of the invention, the parameters include a plurality of blade parameters, each blade parameter characterizing a different blade among a plurality of blades of the wind turbine rotor, wherein each blade parameter includes the blade pitch and / or blade balance mass of the corresponding blade. The blade balance mass can be determined by the number and / or mass of the balance weights attached to the corresponding blade. The masses of the different balance weights can be different, or the masses of all balance weights can be the same. The balance weights can be attached to the corresponding blades at a variable distance from the axis of rotation of the wind turbine rotor, or they can be attached only at a fixed distance from the axis of rotation. The mass of a single balance weight can, for example, be between 5 kg and 50 kg, particularly between 10 kg and 20 kg.

[0066] This embodiment may be advantageous because both blade pitch and blade balance mass are suitable parameters for compensating for imbalances in the wind turbine rotor. For example, blade balance mass may be suitable for compensating for mass imbalances in the wind turbine rotor, while blade pitch may be suitable for compensating for aerodynamic and / or mass imbalances in the wind turbine rotor.

[0067] According to another embodiment of the invention, at least one test offset includes a blade pitch test offset with an absolute value between 0.05 degrees and 2 degrees, particularly between 0.1 degrees and 1 degree, and particularly between 0.3 degrees and 0.6 degrees.

[0068] This embodiment may be advantageous because the blade pitch test offset of that order of magnitude may be large enough to cause significant excitation, but small enough not to lead to serious load problems.

[0069] According to another embodiment of the invention, the acceleration includes the forward and backward acceleration and / or the left and right acceleration of the wind turbine tower and / or the nacelle of the wind turbine.

[0070] The lateral acceleration can be defined as perpendicular to the vertical axis of the tower but parallel to the rotor plane of the wind turbine rotor. The forward / backward direction can be defined as the lateral acceleration perpendicular to both the vertical axis of the tower and the tower itself. Alternatively, the forward / backward direction can be defined relative to the rotor plane, specifically perpendicular to it. When measuring acceleration, deviations from these directions may be possible, for example, a deviation of 15 degrees or less, particularly 5 degrees or less, and especially 1 degree or less.

[0071] This implementation may be advantageous because these accelerations and corresponding oscillations may be well correlated with the imbalance to be reduced.

[0072] According to another embodiment of the invention, the imbalance includes aerodynamic imbalance, which is compensated based on forward and backward acceleration, and mass imbalance, which is compensated based on left and right acceleration. Aerodynamic imbalance may be pitch imbalance.

[0073] Aerodynamic imbalances, particularly pitch imbalances, can be compensated first, followed by mass imbalances, or vice versa. For large mass imbalances, multiple iterations may be required, such as iterations of pitch balancing and mass balancing, or iterations of mass balancing only. Pitch balancing can at least partially compensate for mass imbalances.

[0074] This embodiment may be advantageous because forward and backward acceleration, especially forward and backward oscillation, may be related to aerodynamic imbalance, while left and right acceleration, especially left and right oscillation, may be related to mass imbalance.

[0075] According to another embodiment of the invention, the compensation offset is determined based on the blade pitch angle of the wind turbine rotor, particularly the absolute blade pitch angle of the wind turbine rotor, and / or based on the blade load of the wind turbine rotor. The blade pitch angle may be one of the pitch angles of the blades of the wind turbine rotor, or it may be the average of all the pitch angles of the blades.

[0076] This embodiment may be advantageous for effectively compensating for imbalances, especially when the imbalance depends on the current blade pitch angle.

[0077] According to an exemplary embodiment, the method is performed during the operation of a wind turbine, particularly during energy generation.

[0078] Some embodiments have been described with reference to the claims of the device type, while others have been described with reference to the claims of the method type. However, those skilled in the art will understand from the above and following description that, unless otherwise indicated, any combination of features relating to the method type claims and features relating to the device type claims is disclosed herein.

[0079] The above and other aspects of the invention will be apparent from the examples of the embodiments described below, and will be explained with reference to the examples of the embodiments. Attached Figure Description

[0080] Figure 1 A wind turbine having a control unit according to an exemplary embodiment of the present invention is shown.

[0081] Figure 2 A circle fitted to a frequency component is shown according to an exemplary embodiment of the present invention.

[0082] Figure 3 The pitch offset according to an exemplary embodiment of the present invention is shown, and Figure 4 The corresponding imbalance is shown.

[0083] Figure 5 The pitch offset according to an exemplary embodiment of the present invention is shown, and Figure 6 The corresponding imbalance is shown.

[0084] Figure 7 The pitch offset according to an exemplary embodiment of the present invention is shown, and Figure 8 The corresponding imbalance is shown.

[0085] Figure 9 The correlation between imbalance and pitch angle is shown in an exemplary embodiment of the present invention.

[0086] Figure 10 and Figure 11 A comparison of frequency components with different damping curves according to an exemplary embodiment of the present invention is shown.

[0087] Figure 12 The rotor speed, blade pitch, and wind speed according to an exemplary embodiment of the present invention are shown, and Figure 13 The corresponding equilibrium mass of the blade is shown. Detailed Implementation

[0088] The illustrations in the accompanying drawings are schematic. Similar or identical elements are given the same reference numerals in different drawings. For clarity and understandability, reference numerals are sometimes omitted for features that have already been provided with reference numerals in previous drawings.

[0089] Before describing exemplary embodiments in further detail with reference to the accompanying drawings, some basic considerations will be summarized, upon which exemplary embodiments of the present invention have been developed.

[0090] The developed method can be used to aerodynamically balance the pitch of any wind turbine and then calculate the mass to be added to the rotor to achieve mass balance. A monitor (control unit) using the mass balance calculation method operates continuously to monitor the rotor's balance. This monitor can be used to detect imbalances caused by any reason, such as icing, an unbalanced rotor, blade tip losses, etc.

[0091] This method applies a series of six pitch test offsets while the turbine rotor rotates at a selected speed. A Goertzel filter is used to determine the amplitude and phase (or real and imaginary parts) of the 1P excitation (frequency component) based on nacelle acceleration, which is binned as a function of rotor azimuth over several rotations. For pitch balancing, tower fore-and-aft acceleration is used. Once the sequence of six pitch test offsets has been determined by checking confidence intervals, the resulting data points form a circle in the complex plane. The circle is fitted to the data using a nonlinear least squares method. Then, using different nonlinear least squares methods, six data points with 60° intervals on the circle are fitted to the data.

[0092] Based on the pitch test offset used in the pitch sequence, the diameter of the circle, and the orientation of the data points, the pitch compensation offset required to aerodynamically balance the rotor can be calculated. When applying the pitch compensation offset, the nacelle side-to-side acceleration is used to calculate the mass compensation offset required to balance the rotor. There is a strong coupling between aerodynamic imbalance and mass imbalance; therefore, for large mass imbalances, multiple balancing iterations may be required.

[0093] This method can be used during normal turbine operation. Pitch compensation offset can be calculated as a function of the turbine pitch angle and can be scheduled as a function of the pitch angle to maintain tight balance tolerances for all turbine operating conditions. Rotor imbalance has been observed to depend on the turbine pitch angle.

[0094] As an alternative to directly calculating the pitch compensation offset required for balancing the rotor, an iterative method can be used based on the same 1P data collection method. For this method, a baseline condition without a pitch test offset and six pitch test offset cases are evaluated, and the case with the lowest 1P excitation is selected. The lowest 1P excitation can be either the case with no test offset or the case with one of the six test offsets in the sequence. The condition with the lowest 1P excitation now becomes the baseline, and the amplitude of the pitch test offset used in the sequence is reduced, resulting in further pitch test offsets. A sequence of six additional pitch test offsets is applied, and again the case with the lowest excitation is selected. This process continues until the method converges. A lower bound on the pitch test offset amplitude will be used to provide some differences in the excitation.

[0095] The advantage of the method described above is that it uses data from a pitch offset test sequence, which can be used to calculate the pitch compensation offset required for balancing. Equally important is the use of multiple calculations to estimate convergence and to calculate the average of several rotor revolutions. Once the pitch compensation offset is calculated, this value is maintained to calculate the mass required to balance the rotor. Forward and backward accelerations are used for pitch balancing, and left and right accelerations are used for mass balancing. The method described here requires only standard turbine nacelle acceleration measurements and is capable of both pitch balancing and calculating the mass required to balance a wind turbine rotor.

[0096] Figure 1 A wind turbine 100 is shown, comprising a tower 102 mounted on a base (not shown). The tower 102 is aligned with a vertical axis 109. A nacelle 104 is arranged on top of the tower 102. Between the tower 102 and the nacelle 104, a yaw angle adjustment device 103 is provided, which can be controlled to "yaw" the nacelle 104 for alignment with the current wind direction.

[0097] The wind turbine 100 also includes a wind rotor 110 with three blades 111. Figure 1 Only two of these blades 111 are visible. The wind turbine rotor 110 can rotate about the axis of rotation 118. The blades 111, mounted at the hub 114, extend radially from the axis of rotation 118.

[0098] Between each of the hub 114 and blades 111, a corresponding blade adjustment device 112 is provided to adjust the blade pitch angle 117 of each blade 111 by rotating the corresponding blade 111 about an axis substantially parallel to the longitudinal extension of the blade 111. By controlling the blade adjustment device 112, the blade pitch angle 117 of the corresponding blade 111 can be adjusted in such a way that, at least under normal wind conditions, the maximum wind energy can be recovered from the currently available wind energy. However, the blade pitch angle can also be intentionally adjusted to a position that captures only reduced wind energy.

[0099] Within the nacelle 104, a generator 100 comprising a stator assembly and a rotor assembly is provided. A wind turbine rotor 110 is rotatably connected to the rotor assembly via a transmission system implemented as a rotatable shaft 115. A bearing assembly 116, schematically depicted, is provided to hold both the wind turbine rotor 110 and the rotor assembly in place. Figure 1 As can be seen, shaft 115 extends along the axis of rotation 118. To provide an AC power signal synchronized with the power signal of the utility power grid, the electrical output of stator assembly 110 is electrically connected to power converter 105.

[0100] The wind turbine 100 also includes a control unit 101 for operating the wind turbine 100 in an efficient manner. The control unit 101 may be located inside or outside the wind turbine 100. In addition to controlling, for example, a yaw angle adjustment device 112, the control unit 101 is also used to adjust the blade pitch angle 117 of the blades 111 of the wind rotor 110 in an optimized manner.

[0101] The control unit is configured to perform a method for compensating for an imbalance in the wind turbine rotor 110 of the wind turbine 100. The method includes applying at least one test offset to a parameter characterizing the state of the wind turbine rotor 110, wherein the imbalance depends on the parameter. The method also includes measuring the acceleration of the wind turbine 100 for at least one test offset, such as acceleration along the longitudinal direction 107 and / or along the lateral direction 108, wherein the acceleration depends on the imbalance. The method concludes by determining a compensation offset based on the measured accelerations 107, 108, wherein when the compensation offset is applied to the parameter, the imbalance is at least partially compensated. The test offset and / or compensation offset can be applied by changing the pitch angle 117 of the blades 111 and / or by adding or removing balance weights 113 of the blades 111.

[0102] Figure 2 Frequency components 221-226 are shown mapped to circle 237, which is used for the pitch balancing process. During this pitch balancing process, the following turbine operating conditions are applied: The turbine is offline at low wind speeds in the order of 4 m / s to 7 m / s, and the rotor speed of the wind turbine rotor is regulated by the blade pitch, for example, at the rated rotor speed.

[0103] Then, known pitch test offsets (e.g., 0.3, -0.15, -0.15) are applied. The forward and backward acceleration data are binned as a function of azimuth. Based on this data, the acceleration amplitude and phase of the corresponding frequency components 221-226 are calculated every N (~10) revolutions using a Goertzel filter until the desired confidence interval or the maximum allowed iteration value is reached.

[0104] Based on this process, data for the positive and negative test offsets of each blade are collected, for example, data is collected for the following six test offsets:

[0105]

[0106] Each test offset corresponds to frequency components 221-226. Frequency components 221-226 are mapped to a coordinate system using Goertzel amplitude and Goertzel phase, i.e., the corresponding amplitude and phase of each of frequency components 221-226.

[0107] The parameters of circle 237, fitted with acceleration data, are estimated using a nonlinear least squares method. Circle 237 has a center 238. Based on different nonlinear least squares methods, six fitted frequency components 231-236 are determined using a 60-degree offset at an unknown number of revolutions for each fitted frequency component 231-236. This reduces directional sensitivity. The six fitted frequency components 231-236 lie on the circumference of circle 237.

[0108] The pitch compensation offset is calculated to center the circle based on the fit of the data points and the known pitch offset used to generate the data, such as a Coleman transform based on a nonlinear least-squares fit from positive and negative pitch test offsets. For example, the pitch compensation offset can be calculated based on the following equation:

[0109] A tilt =-y center Pitch check / r circle

[0110] A yaw =-x center Pitch check / rcircle

[0111]

[0112] Here, r circle It is the radius of circle 237, for example, 0.0153, x center and y center These are the coordinates of the center 238 of circle 237, for example, -0.01382 and -0.0061, pitch. check It is the amplitude of the pitch offset used to generate data, and It is determined by the orientation 239 of at least one of the fitted frequency components 231-236. Finally, CO pitchA CO pitchB CO pitchC It is the calculated pitch compensation offset.

[0113] The described method can be used during commissioning and can be periodically activated as a check. A periodic or continuous variation applies positive and negative pitch test offsets to each blade, selects the minimum 1P acceleration as the new pitch offset reference, then reduces the pitch test offset amplitude, and repeats the process.

[0114] Figure 3 The blade pitch compensation offset 341-343 is shown, and Figure 4 The corresponding imbalances 344 and 345 are shown, which have been at least partially compensated. Figure 3 and Figure 4 The calculations for compensating for mass and pitch imbalance are shown, which were specifically introduced for testing purposes. Specifically, the balance blocks are moved to create a mass imbalance; two balance blocks are removed from blade B and placed into blade A along with three other balance blocks. Furthermore, a pitch offset has been applied, i.e., a +0.5° pitch imbalance offset has been introduced on blade C.

[0115] To determine the blade pitch compensation offset 341-343, five pitch and mass balance tests were performed corresponding to the number of iterations. Figure 3 The blade pitch compensation offsets 341-343, in degrees, depending on the number of iterations, are shown. As can be seen, the pitch compensation offsets 341-343 calculated by the methods described herein are consistent across five tests. The average compensation offsets 341-343 calculated for pitch and mass are: 0.598° for blade A, -0.005° for blade B, and -0.594° for blade C.

[0116] These compensation offsets appear credible and consistent, as when considering... Figure 3 and Figure 4Subtracting the compensation offset for mass imbalance only (0.474° for blade A, 0.167° for blade B, and -0.333° for blade C) from the compensation offsets for pitch and mass imbalance determined in this embodiment yields the following differences: 0.124° for blade A, 0.200° for blade B, and -0.325° for blade C. These values ​​are equivalent to the zero-mean correction for a +0.5° pitch imbalance offset on blade C, i.e., 0.167° for blade A, 0.167° for blade B, and -0.333° for blade C.

[0117] Figure 4 The torque imbalance 344, measured in kg*m, and the corresponding phase imbalance 345, measured in degrees, are shown according to the number of iterations. The confidence interval 346 for the torque imbalance 344 is also depicted. The corresponding imbalances 344 and 345, as depicted, have been at least partially compensated by applying blade pitch compensation offsets 341-343. Consistent pitch balance conditions are obtained. The average amplitude of the torque imbalance 344, maintaining the pitch offset at the value used to balance the pitch, is 1984.0 kg*m.

[0118] For general balancing, pitch balance can first be calculated by applying a controllable test offset. Pitch compensation offset can be applied to balance the pitch, and then the mass imbalance can be calculated. For large mass imbalances, at least two iterations of rotor imbalance estimation may be required, especially since large mass imbalances affect pitch balance calculations. Only mass is corrected in the first iteration. The first mass correction is calculated using data from both the mass and pitch imbalance data. The second mass correction is calculated using data from the first mass correction. In at least some cases, two iterations seem sufficient for both pitch and mass corrections. Figure 3 and Figure 4 The rotor imbalance estimate during the first iteration is shown. Figure 5 and Figure 6 The rotor imbalance estimate after the first iteration is shown. Figure 7 and Figure 8 The rotor imbalance estimate after the second iteration is shown.

[0119] Figure 5 The blade pitch compensation offset 341-343 is shown, and Figure 6 The corresponding imbalances 344 and 345 are shown, which have been at least partially compensated. Figure 5 and Figure 6 It shows that in such Figure 3 and Figure 4The rotor imbalance estimate shown is the rotor imbalance estimate after the first iteration with only mass compensation.

[0120] To determine the blade pitch compensation offset 341-343, five pitch and mass balance tests were performed corresponding to the number of iterations. Figure 5 The blade pitch compensation offsets 341-343, in degrees, depending on the number of iterations, are depicted. As can be seen, the pitch compensation offsets 341-343 calculated by the methods described herein are consistent across five tests. The calculated average compensation offsets 341-343 for the pitch are: 0.405° for blade A, 0.057° for blade B, and -0.463° for blade C.

[0121] Figure 6 The corresponding torque imbalance 344, measured in kg*m, and the corresponding phase imbalance 345, measured in degrees, depending on the number of iterations, are depicted. The confidence interval 346 for the torque imbalance 344 is also depicted. The corresponding imbalances 344 and 345, as depicted, have been at least partially compensated by applying blade pitch compensation offsets 341-343. Consistent pitch balance conditions are obtained. The average amplitude of the torque imbalance 344, maintaining the pitch offset at the value used to balance the pitch, is 586.4 kg*m.

[0122] Figure 7 The blade pitch compensation offset 341-343 is shown, and Figure 8 The corresponding imbalances 344 and 345 are shown, which have been at least partially compensated. Figure 7 and Figure 8 It shows that in such Figure 5 and Figure 6 The rotor imbalance estimate shown is the rotor imbalance estimate after the second iteration with second mass compensation and pitch compensation.

[0123] To determine the blade pitch compensation offset 341-343, ten pitch and mass balance tests were performed corresponding to the number of iterations. Figure 7 The blade pitch compensation offsets 341-343, in degrees, depending on the number of iterations, are depicted. As can be seen, the pitch compensation offsets 341-343 calculated by the method described in this paper are consistent across ten tests. The average compensation offsets 341-343 for pitch are: -0.0056° for blade A, -0.0323° for blade B, and 0.0388° for blade C. Therefore, the pitch is well balanced.

[0124] Figure 8The corresponding torque imbalance 344, measured in kg*m, and the corresponding phase imbalance 345, measured in degrees, are depicted depending on the number of iterations. The confidence interval 346 for the torque imbalance 344 is also depicted. The corresponding imbalances 344 and 345, as depicted, have been at least partially compensated by applying blade pitch compensation offsets 341-343. Consistent pitch balance conditions are obtained. The average amplitude of the torque imbalance 344, maintaining the pitch offset at the value used to balance the pitch, is 295.3 kg*m. The mass imbalance is within a mass balance block tolerance.

[0125] Figure 9 Confidence intervals for torque imbalance 344 (in kg*m), phase imbalance 345 (in degrees), and torque imbalance 346, dependent on the blade pitch angle of the wind turbine rotor, are depicted. As can be seen, imbalances 344 and 345 increase with the blade pitch angle. Therefore, it may be necessary to calculate pitch compensation offset based on the blade pitch angle to balance the rotor. Accordingly, the pitch compensation offset can be scheduled according to the pitch to achieve a tight balance target.

[0126] Figure 10 and Figure 11 The simulated frequency components compared to damping curves 751-754 are shown in order to determine the damping constant ζ. Figure 10 and Figure 11 In this embodiment, a mass asymmetry of 125 kg has been introduced at 22.18 m on one of the blades. Each damping curve corresponds to a different damping constant ζ; for example, damping curve 751 corresponds to a damping constant of 0.02, damping curve 752 corresponds to a damping constant of 0.05, damping curve 753 corresponds to a damping constant of 0.10, and damping curve 754 corresponds to a damping constant of 0.15. The amplitude and phase of the frequency components are determined based on the Goertzel algorithm.

[0127] exist Figure 10 The figure shows the amplitudes of the frequency components 761-765 and ω / ω. n The relationship is given by ω, where ω is the rotational frequency of the wind turbine rotor, and ω n It is the natural eigenfrequency of the tower. Figure 10 and Figure 11 In the embodiments, ω n The frequency component amplitude is 0.265 Hz. The amplitude of the frequency component 761-765 is compared with different damping curves 751-754. The amplitude consistently follows the damping curve 751 with a damping constant of 0.02.

[0128] exist Figure 11 The diagram shows the phase of the frequency components 766-769 and ω / ω. nThe phases 766-769 correspond to amplitudes 762-765, respectively. The phases 761-765 of the frequency components are compared with different damping curves 751-754. As in the case of amplitude, the phases also follow the damping curve 751 very well with a damping constant of 0.02.

[0129] Based on the determined damping constant, the balancing mass to be added to the wind turbine blades can be calculated to compensate for mass imbalance. For example, if the mass imbalance is due, it can be compensated for by a virtual mass M at a certain distance r from the rotation axis of the wind turbine rotor. Rot This can be represented as, and if the virtual mass is located in the direction of one of the blades, the balance mass to be added to the other blades can be calculated.

[0130] For example, if leaf A is heavy, that is, if in This is the angle between the vertical axis of the wind turbine tower and blade A when the virtual mass is positioned at 90° to the vertical axis. Then, the mass to be added to blades B and C can be calculated as follows:

[0131]

[0132] M Rotr That is, the product of the virtual mass and the distance of the virtual mass to the axis of rotation can be calculated based on the damping constant ζ and the frequency of the wind turbine rotor ω.

[0133] Figure 12 The relationship between rotor speed 871 in rpm, blade pitch 872 in degrees, and wind speed 873 in m / s and time in seconds is shown. Figure 13 The corresponding balance masses 874-876 (in kg) are shown, which also depend on time in seconds. Balance mass 874 is added to blade A, balance mass 875 is added to blade B, and balance mass 876 is added to blade C. The balance mass 876 of blade C is zero because... Figure 12 and Figure 13 In one embodiment, the imbalance of the heavy blade C is compensated.

[0134] During mass balancing, the following operating conditions are applied: The turbine is offline under low wind speeds of 4 to 7 m / s. The rotor speed of the wind turbine rotor is adjusted via the blade pitch, for example, at the rated rotor speed.

[0135] The tower's left and right acceleration data are binned as a function of azimuth. The acceleration amplitude and phase of the frequency components are calculated every N(10) revolutions using a Goertzel filter until the desired confidence interval or the maximum allowed iteration value is reached. The balancing mass to be added to each blade for balancing is then calculated. For this, the tower's left and right frequencies and the tower head mass must be known, and low damping (ζ = 0.01–0.02) must be assumed. Figure 13 As shown, after several calculations, the determined balance masses 875 and 876 for blades A and B converge. Instead of calculating the individual balance mass for each blade, the magnitude and direction of the imbalance can be calculated.

[0136] For overall balance, i.e., pitch and mass balance, one can begin with pitch balance because pitch offset can be controlled to estimate pitch imbalance. Pitch offset can be applied to aerodynamic balancing. The mass required for balancing can be calculated according to the mass balancing process. The pitch offset required for aerodynamic balancing and the mass to be added to each blade for mass balancing can be reported or stored in non-volatile memory.

[0137] It should be noted that the term "comprising" does not exclude other elements or steps, and the use of the article "a" (aoran) does not exclude multiple (unless a quantity is specified before the noun, multiple is not excluded). Furthermore, elements described for different embodiments may be combined. It should also be noted that reference numerals in the claims should not be construed as limiting the scope of the claims.

Claims

1. A method for compensating for imbalance (344, 345) in the wind turbine rotor (110) of a wind turbine (100), the method comprising: At least two test offsets are applied one after another to parameters (113, 117) characterizing the state of the wind turbine rotor (110), wherein the imbalance (344, 345) depends on the parameters (113, 117). The acceleration of the wind turbine (100) is measured for the at least two test offsets, wherein the acceleration depends on the imbalance (344, 345); and The compensation offset (341-343, 874-876) is determined based on the measured acceleration, wherein when the compensation offset (341-343, 874-876) is applied to the parameters (113, 117), the imbalance (344, 345) is at least partially compensated, wherein determining the compensation offset (341-343, 874-876) includes: For each of the at least two test offsets, determine the frequency components (221-226) of the spectrum of the measured acceleration. Map each of the frequency components (221-226) to a coordinate system. Determine the center (238) defined by each of the frequency components (221-226), wherein the center is defined as the centroid of the geometry defined by each of the frequency components, and The compensation offset (341-343, 874-876) is calculated based on the center (238). The parameters (113, 117) include multiple blade parameters, each blade parameter representing a different blade (111) among the multiple blades (111) of the wind turbine rotor (110). Each blade parameter includes the blade pitch (117).

2. The method of claim 1, wherein determining the compensation offset (341-343, 874-876) comprises determining the amplitude (761-765) and / or phase (766-769) of the frequency component (221-226) of the spectrum of the measured acceleration for each of the at least one test offset.

3. The method according to claim 1 or 2, wherein, The frequency of each of the frequency components (221-226) is based on the rotor frequency of the wind turbine (110), and in particular equal to the rotor frequency of the wind turbine (110).

4. The method according to claim 1 or 2, wherein, Determining the compensation offset (341-343, 874-876) includes fitting a circle (237) based on each of the frequency components (221-226), and calculating the compensation offset (341-343, 874-876) based on the circle (237), wherein the circle (237) is fitted based on at least three frequency components (221-226).

5. The method according to claim 1 or 2, wherein, Determining the compensation offset (341-343, 874-876) includes setting the compensation offset (341-343, 874-876) to one of the at least one test offset.

6. The method according to claim 1 or 2, wherein, After determining the compensation offsets (341-343, 874-876), the method further includes: Apply the compensation offsets (341-343, 874-876) to the parameters (113, 117); At least one additional test offset is applied to the parameters (113, 117), each of which is less than the corresponding test offset; The acceleration of the wind turbine (100) is measured against at least one additional test offset; and Additional compensation offsets (341-343, 874-876) are determined based on the measured acceleration.

7. The method of claim 1 or 2, wherein applying the at least one test offset comprises sequentially applying a plurality of test offsets.

8. The method according to claim 7, wherein, Each of the plurality of test offsets includes a blade test offset for each of the plurality of blades (111) of the wind turbine rotor (110), wherein the values ​​of the blade test offsets are arranged for different test offsets, and / or wherein the signs of the values ​​of the blade test offsets are reversed for different test offsets.

9. The method according to claim 1 or 2, wherein, Each blade parameter includes the blade balance mass of the corresponding blade (111).

10. The method according to claim 1 or 2, wherein, At least one test offset includes blade pitch test offsets with absolute values ​​between 0.05 degrees and 2 degrees, particularly between 0.1 degrees and 1 degree, and particularly between 0.3 degrees and 0.6 degrees.

11. The method according to claim 1 or 2, wherein, The acceleration includes the forward and backward acceleration (107) and / or the left and right acceleration (108) of the tower (102) and / or the nacelle (104) of the wind turbine (100).

12. The method according to claim 11, wherein, The imbalance (344, 345) includes an aerodynamic imbalance compensated for based on the front-to-back acceleration (107), and the imbalance (344, 345) includes a mass imbalance compensated for based on the left-to-right acceleration (108).

13. The method according to claim 1 or 2, wherein, The compensation offsets (341-343, 874-876) are determined based on the blade pitch angle (872) of the wind turbine rotor (110) and / or the blade load of the wind turbine rotor.

14. The method according to claim 1 or 2, wherein, Determining the compensation offset (341-343, 874-876) includes setting the compensation offset (341-343, 874-876) to a test offset corresponding to the frequency component with the minimum amplitude.

15. A wind turbine (100) for generating electricity, said wind turbine (100) comprising: Tower (102); A wind turbine rotor (110) is arranged at the top portion of the tower (102) and includes at least one blade (111); An electric motor (106) configured as a generator (106) is mechanically connected to the wind turbine rotor (110); as well as A control unit configured to perform the method according to any one of claims 1 to 14.

Images