A method for correcting the balance of a wind turbine rotor

Abstract

This invention discloses a method for balancing and correcting a wind turbine rotor, comprising: acquiring vibration signals, rotor azimuth angle signals, and operating parameters of the wind turbine; performing azimuth synchronization processing on the vibration signals based on the rotor azimuth angle signals to extract the 1P vibration vector; performing operating condition normalization processing on the 1P vibration vector based on the operating parameters to obtain a normalized unbalance vector; applying a preset pitch disturbance to each blade and obtaining the change in the normalized unbalance vector before and after the disturbance; establishing an aerodynamic unbalance sensitivity matrix based on the change; decomposing the current normalized unbalance vector into an aerodynamic unbalance component and a mass unbalance component based on the aerodynamic unbalance component; determining the blade pitch offset correction amount based on the aerodynamic unbalance component, and determining the counterweight and counterweight orientation based on the residual mass unbalance component. This invention can distinguish between aerodynamic unbalance and mass unbalance, reduce blind counterweighting, and improve the accuracy and efficiency of rotor balancing and correction.

Description

Technical Field

[0001] This invention belongs to the field of wind turbine generators, and specifically relates to a method for balancing and correcting the rotor of a wind turbine generator. Background Technology

[0002] During operation, the rotor system of a wind turbine typically consists of a hub and multiple blades mounted on the hub. Under wind load, the blades drive the rotor to rotate, converting mechanical energy into electrical energy through a transmission chain or direct-drive structure. Due to the large size and mass of wind turbine blades, and their exposure to complex environments such as alternating wind loads, temperature variations, rain and snow erosion, blade contamination, and structural fatigue, the rotor system is prone to varying degrees of imbalance. Rotor imbalance increases vibration at a certain rotational frequency, causing additional loads on the main bearings, gearbox, generator, tower, and hub connection structure. In severe cases, this can lead to vibration alarms, reduced load operation, shutdown for maintenance, or even premature component damage, affecting the wind turbine's operational safety and power generation efficiency.

[0003] Existing methods for balancing wind turbine rotors typically rely on vibration signals. By measuring the vibration amplitude and phase at the nacelle, main bearing, or tower, the location of rotor imbalance is determined, and counterweights are added to the hub or blade root to reduce periodic vibrations during rotor rotation. These methods are effective for mass imbalances such as blade mass variations, blade center-of-gravity deviations, or hub counterweight deviations. However, wind turbine rotor imbalances are not solely caused by mass factors; they can also originate from aerodynamic factors such as blade pitch zero-position deviation, blade installation angle errors, blade surface contamination, blade leading-edge damage, and differences in aerodynamic shape after blade maintenance. These aerodynamic imbalances also manifest as increased rotor vibration at a specific rotational frequency, and their vibration characteristics are easily confused with those of mass imbalances in field data.

[0004] In actual operation and maintenance, if all detected single-rotor frequency vibrations are directly attributed to mass imbalance and counterweight correction is performed accordingly, problems such as inaccurate correction direction, repeated adjustment of counterweight, and increased downtime may occur. Especially when the main source of imbalance is aerodynamic imbalance, simply adding counterweights often fails to achieve the desired effect and may even introduce new mass deviations, making the rotor vibration state more complex. On the other hand, some methods reduce rotor vibration by adjusting the pitch angle of individual blades, but if mass imbalance and aerodynamic imbalance cannot be distinguished, mass deviation may be misjudged as pitch deviation, resulting in unreasonable pitch correction and affecting blade load distribution and unit power generation performance. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method for balancing wind turbine rotors. This method can accurately identify rotor imbalance under field operating conditions using vibration signals, rotor azimuth signals, and operating parameters. Furthermore, it distinguishes between mass imbalance components and aerodynamic imbalance components, thereby employing pitch offset correction and counterweight correction based on different imbalance sources. This improves the accuracy, reliability, and field implementation efficiency of rotor balancing.

[0006] The technical solution provided by this invention is as follows: A method for balancing and correcting a wind turbine rotor includes the following steps: S1. Collect vibration signals, rotor azimuth angle signals, and operating parameters of the wind turbine during operation; S2. Perform azimuth synchronization processing on the vibration signal based on the rotor azimuth angle signal, and extract the 1P vibration vector corresponding to the rotor's first rotational frequency. S3. Normalize the 1P vibration vector according to the operating condition parameters to obtain the normalized unbalance vector. S4. Apply a preset pitch disturbance to each blade of the wind turbine generator and obtain the normalized unbalance vector change of each blade before and after applying the preset pitch disturbance. S5. Establish the aerodynamic imbalance sensitivity matrix based on the normalized imbalance vector change corresponding to each blade. S6. Based on the aerodynamic imbalance sensitivity matrix, decompose the current normalized imbalance vector into aerodynamic imbalance components and mass imbalance components. S7. Determine the blade pitch offset correction amount based on the aerodynamic imbalance component, and determine the counterweight and counterweight orientation based on the residual mass imbalance component after blade pitch offset correction, so as to perform graded balance correction on the wind turbine rotor.

[0007] In some implementations, in step S1, the operating parameters include at least two of the following: rotor speed, wind speed, yaw error, generator power, and blade pitch angle.

[0008] In some implementations, a valid data filtering step is included after step S1. The valid data filtering step includes selecting a data window that meets the stable operating conditions from the collected data. The stable operating conditions include the rotor speed fluctuation value being less than a first preset threshold, the wind speed fluctuation value being less than a second preset threshold, the yaw error being less than a third preset threshold, and the wind turbine not being in a start-stop, power-limited, or fault-reduced state.

[0009] In some implementations, step S2, the azimuth synchronization processing includes: converting the vibration signal from time domain data to azimuth domain data based on the rotor azimuth angle, and extracting the first-order frequency component of the azimuth domain data to obtain a 1P vibration vector with amplitude and phase.

[0010] In some implementations, step S3, the operating condition normalization process includes: establishing operating condition correction coefficients based on at least two of rotor speed, wind speed, yaw error, generator power, and blade pitch angle, and using the operating condition correction coefficients to normalize the amplitude of the 1P vibration vector, and / or performing equivalent conversion on the 1P vibration vector under different operating conditions.

[0011] In some implementations, in step S4, the preset pitch disturbance is a short-term reversible pitch offset applied to a single blade, and the preset pitch disturbance is applied to only one blade within the same disturbance test cycle.

[0012] In some implementations, in step S5, the aerodynamic imbalance sensitivity matrix is ​​composed of the mapping relationship between the pitch disturbance of each blade and the corresponding normalized imbalance vector change, and is used to characterize the degree of influence of the pitch change of each blade on the rotor 1P imbalance state.

[0013] In some implementations, step S6, decomposing the current normalized unbalance vector into an aerodynamic unbalance component and a mass unbalance component, includes: based on the aerodynamic unbalance sensitivity matrix, solving for the blade pitch offset combination that minimizes the residual vector of the current normalized unbalance vector, and taking the vector corresponding to the blade pitch offset combination as the aerodynamic unbalance component, and taking the remaining vector after deducting the aerodynamic unbalance component from the current normalized unbalance vector as the mass unbalance component.

[0014] In some implementations, in step S7, when the ratio of the amplitude of the aerodynamic imbalance component to the amplitude of the current normalized imbalance vector is greater than a first proportional threshold, the blade pitch offset correction amount is determined preferentially based on the aerodynamic imbalance component; when the ratio of the amplitude of the mass imbalance component to the amplitude of the current normalized imbalance vector is greater than a second proportional threshold, the counterweight and counterweight orientation are determined based on the mass imbalance component.

[0015] In some implementations, after the balance correction is completed, the vibration signal and operating condition parameters of the wind turbine are collected again during operation, the corrected normalized unbalance vector is recalculated, and when the corrected normalized unbalance vector is greater than the preset balance threshold, blade pitch offset correction and / or counterweight correction are performed based on the corrected normalized unbalance vector.

[0016] In summary, the beneficial effects of this invention are: (1) This invention collects vibration signals, rotor azimuth angle signals, and operating parameters during the operation of a wind turbine generator, and performs azimuth synchronization processing on the vibration signals according to the rotor azimuth angle. This allows the vibration signals to be converted to rotor azimuth coordinates for analysis, thereby accurately extracting the 1P vibration vector corresponding to the rotor's first rotational frequency. Compared to methods that rely solely on time-domain vibration amplitude for judgment, this invention improves the accuracy of unbalanced vibration phase identification and reduces the impact of rotational speed fluctuations on the rotor unbalance judgment results.

[0017] (2) The present invention performs condition normalization processing on the 1P vibration vector according to the operating condition parameters, which can reduce the interference of changes in operating conditions such as wind speed, rotational speed, yaw error, generator power and blade pitch angle on the vibration response, and make the 1P vibration vectors obtained under different operating conditions more comparable. As a result, it can avoid misjudging wind fluctuations, yaw errors or unit load changes as rotor imbalance, and improve the stability and reliability of the balance correction judgment results. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0019] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. The following embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0020] The embodiments of the present invention are illustrated using a three-bladed horizontal axis wind turbine as an example, but the method is not limited to three-bladed wind turbines, and can also be applied to rotor balance correction scenarios of two-bladed, multi-bladed, or other wind turbines with independent pitch control functions.

[0021] A method for balancing and correcting the rotor of a wind turbine includes the following steps: data acquisition, effective data screening, azimuth synchronization processing, 1P vibration vector extraction, operating condition normalization, single-blade pitch disturbance testing, aerodynamic unbalance sensitivity matrix establishment, unbalance component decomposition, graded correction, and closed-loop verification.

[0022] In this embodiment, the wind turbine includes a hub, a first blade, a second blade, and a third blade mounted on the hub, a pitch actuator for driving the blades to adjust pitch, an azimuth detection unit for detecting the rotor azimuth angle, a vibration detection unit for detecting the unit's vibration state, and a main control system for acquiring the unit's operating status. The vibration detection unit can be an acceleration sensor, a velocity sensor, or a displacement sensor, preferably installed at a location near the main bearing housing, nacelle floor, tower top, or hub that reflects rotor imbalance excitation. The vibration detection unit can detect vibrations in the nacelle's forward / backward direction, left / right direction, or vertical direction, and can also detect vibration signals in multiple directions simultaneously.

[0023] Specifically, in step S1, vibration signals, rotor azimuth angle signals, and operating parameters are collected during the operation of the wind turbine. These operating parameters include one or more of the following: rotor speed, wind speed, yaw error, generator power, blade pitch angle, ambient temperature, and unit operating mode. Preferably, at least rotor speed, wind speed, yaw error, generator power, and blade pitch angle are collected to allow for subsequent correction of the vibration response under different operating conditions.

[0024] The sampling frequency of the vibration signal can be 50Hz-2000Hz, preferably 100Hz-1000Hz. The rotor azimuth angle signal can be obtained through an encoder, main shaft speed sensor, generator encoder, hub azimuth sensor, or existing azimuth angle data from the wind turbine main control system. For ease of subsequent calculation, the rotor azimuth angle corresponding to the first blade being in a vertically upward position can be defined as 0°, and one rotation of the rotor corresponds to 0°-360°.

[0025] After step S1, valid data filtering can be performed. The purpose of valid data filtering is to eliminate unstable operating data such as start-up and shutdown, emergency stop, fault load reduction, strong turbulence, yaw adjustment, and power limitation, so as to avoid misjudging external operating condition disturbances as rotor imbalance.

[0026] Specifically, the collected data is divided into multiple data windows, each with a length of 30s-600s, preferably 60s-300s. For each data window, it is determined whether stable operating conditions are met. Stable operating conditions may include: rotor speed fluctuation within the data window is less than a first preset threshold, wind speed fluctuation is less than a second preset threshold, yaw error is less than a third preset threshold, and the wind turbine is not in a start-stop, power-limited, or fault-reduced load state.

[0027] For example, the first preset threshold can be set to 2%-5% of the average rotational speed; the second preset threshold can be set to 10%-20% of the average wind speed; and the third preset threshold can be set to 5°-15°. When a data window meets all of the above conditions, it is considered a valid data window; otherwise, it is discarded.

[0028] In step S2, the vibration signal is subjected to azimuth synchronization processing based on the rotor azimuth angle signal to extract the 1P vibration vector corresponding to one rotor rotation frequency. The 1P vibration vector refers to the vibration component generated once per rotor revolution, which can usually reflect the periodic excitation caused by rotor mass imbalance, blade aerodynamic imbalance, or blade installation angle deviation.

[0029] Specifically, the vibration signals within the effective data window are converted from time-domain data to rotor azimuth-domain data. During the conversion, the vibration signals are resampled at equal angles based on the rotor azimuth angle. For example, the rotor revolution is divided into 360, 720, or 1024 equal-angle sampling points, and the corresponding vibration value is obtained by interpolation at each predetermined azimuth angle. Through this processing, even if there are slight fluctuations in the rotor speed, the vibration data can be unified under the rotor azimuth coordinate system for analysis.

[0030] For the vibration signal after azimuth synchronization, the 1P component can be extracted using Fourier analysis, least squares sine fitting, or synchronous demodulation. Taking least squares fitting as an example, the azimuth domain vibration signal can be represented as: x(θ)=a·cosθ+b·sinθ+c Where x(θ) is the vibration signal corresponding to the rotor azimuth angle θ, a and b are the first-order rotational frequency component coefficients, and c is the DC term. Based on a and b, the 1P vibration vector V can be obtained, with its amplitude A and phase φ as follows: A = √(a² + b²) φ = atan2(b, a) Therefore, the 1P vibration vector can be expressed as V=A∠φ, or as a complex number V=A·ejφ. If vibration signals from multiple directions are collected, the 1P vibration vector in each direction can be calculated separately, or it can be synthesized into a comprehensive 1P vibration vector according to preset weights. For example, the left-right vibration of the cabin can be used as the main unbalance evaluation quantity, and the forward-backward vibration of the cabin can be used as an auxiliary evaluation quantity.

[0031] In step S3, the 1P vibration vector is normalized according to the operating condition parameters to obtain the normalized unbalance vector. Since the vibration response of a wind turbine is affected by factors such as wind speed, rotational speed, power, pitch angle, and yaw error, the same rotor unbalanced state may exhibit different 1P vibration amplitudes under different operating conditions. Therefore, it is necessary to perform operating condition normalization.

[0032] Specifically, a reference operating condition can be selected, which includes reference wind speed v0, reference engine speed n0, reference power P0, reference pitch angle β0, and reference yaw error γ0. The operating condition correction factor K is established based on historical operating data, unit simulation models, or field calibration data. The operating condition correction factor K can be expressed as: K = f(v, n, P, β, γ) Where v is wind speed, n is rotor speed, P is generator power, β is blade pitch angle, and γ is yaw error. The function f can be a lookup table function, a polynomial function, a piecewise linear function, or a regression function trained from historical data.

[0033] Then, the 1P vibration vector is normalized using the working condition correction factor K to obtain the normalized unbalance vector U: U=V / K Where V is the measured 1P vibration vector and U is the normalized unbalance vector. For cases where the phase is less affected by operating conditions, only the amplitude of the 1P vibration vector can be normalized; for cases where the phase of the unit's structural response changes significantly with the rotational speed, the phase can also be equivalently converted based on the rotational speed range or modal response characteristics.

[0034] In a preferred embodiment, the operating condition correction coefficient K is determined at least based on the rotor speed and wind speed. For example, the wind speed is divided into several wind speed ranges, and the rotor speed is divided into several speed ranges, with corresponding correction coefficients pre-established for each wind speed range and speed range. When the current wind speed and speed are collected, the current correction coefficient is obtained through table lookup and interpolation. This approach reduces the requirements for on-site implementation of complex algorithms and facilitates deployment in the wind turbine main control system or back-end diagnostic platform.

[0035] In step S4, a preset pitch disturbance is applied to each blade of the wind turbine, and the normalized unbalance vector change of each blade before and after the preset pitch disturbance is obtained. The preset pitch disturbance is a short-term, reversible, small-amplitude pitch offset applied to a single blade. Within the same disturbance test cycle, only one blade is applied to the preset pitch disturbance, while the remaining blades maintain their original pitch control state or remain consistent with the main control system command.

[0036] For example, a pitch offset of +0.2° is applied to the first blade for 60-180 seconds. Vibration signals and operating parameters are collected during this disturbance period, and the corresponding normalized unbalance vector U1+ is calculated. Then, the pitch disturbance of the first blade is canceled, restoring it to its original pitch control state. The same pitch disturbance is then applied to the second blade, resulting in U2+; the same pitch disturbance is then applied to the third blade, resulting in U3+.

[0037] To improve the accuracy of sensitivity calculation, positive and negative perturbations can be applied to each blade separately. For example, applying a +Δβ perturbation to the first blade yields U1+, and applying a -Δβ perturbation yields U1-. The aerodynamic sensitivity h1 of the first blade can then be calculated as follows: h1=(U1+-U1-) / (2Δβ) Where Δβ is the pitch disturbance. Similarly, the aerodynamic sensitivity h2 of the second blade and the aerodynamic sensitivity h3 of the third blade can be calculated.

[0038] The preset pitch disturbance can be 0.05°-1.0°, preferably 0.1°-0.5°. This disturbance should be sufficient to cause a detectable change in the 1P vibration vector, but should not cause significant fluctuations in unit power or trigger safety protection. During the disturbance, if there is a sudden change in wind speed, excessive yaw error, or the unit enters a power-limited, load-reduced, or faulty operating state, the disturbance test should be terminated, and a stable operating window should be selected for testing again.

[0039] In step S5, an aerodynamic imbalance sensitivity matrix is ​​established based on the normalized imbalance vector change corresponding to each blade. The aerodynamic imbalance sensitivity matrix is ​​used to characterize the degree of influence of the blade pitch change on the rotor 1P imbalance state.

[0040] For a three-bladed wind turbine, the aerodynamic imbalance sensitivity matrix H can be expressed as: H = [h1 h2 h3] Where h1, h2, and h3 represent the aerodynamic sensitivities of the first, second, and third blades, respectively. Each aerodynamic sensitivity can be a complex vector or a two-dimensional column vector composed of real and imaginary parts. For example, if h1 = ΔU1 / Δβ1, then it represents the normalized unbalance vector change caused by a unit change in the pitch angle of the first blade.

[0041] In one implementation, to avoid interference with rotor imbalance detection caused by the simultaneous increase or decrease of the same pitch angle by all three blades, constraints can be set for the pitch offset combination: δβ1+δβ2+δβ3=0 Wherein, δβ1, δβ2, and δβ3 are the pitch offsets of the first, second, and third blades, respectively. This constraint means that only the relative pitch deviation between the blades is considered, and the overall pitch change is not taken as an aerodynamic imbalance correction.

[0042] In step S6, the current normalized imbalance vector is decomposed into an aerodynamic imbalance component and a mass imbalance component based on the aerodynamic imbalance sensitivity matrix. Specifically, the current normalized imbalance vector can be denoted as Uc, and the blade pitch offset combination δβ is solved based on the aerodynamic imbalance sensitivity matrix H, so that the aerodynamic equivalent vector generated by the pitch offset combination is as close as possible to the aerodynamic component in the current normalized imbalance vector.

[0043] In the specific solution, the least squares method can be used: δβ*=argmin‖Uc-Hδβ‖ Where δβ is the blade pitch offset combination obtained by the solution, and Hδβ is the aerodynamic equivalent vector corresponding to this pitch offset combination. Based on the solution results, Hδβ is taken as the aerodynamic unbalance component Ua, and Uc-Ua is taken as the mass unbalance component Um, that is: Ua=Hδβ* Um = Uc - Ua Wherein, Ua represents the aerodynamic imbalance component caused by blade pitch deviation, blade aerodynamic shape difference, blade contamination or blade installation angle error, etc.; Um represents the mass imbalance component caused by blade mass deviation, blade center of gravity deviation, hub counterweight deviation, etc.

[0044] To avoid excessive pitch correction, upper and lower limits can be set for δβ*. For example, the pitch offset correction for each blade can be limited to between -1.0° and +1.0°, preferably between -0.5° and +0.5°. When the solution exceeds this range, it is limited to the allowable range, and the remaining portion that cannot be offset by the pitch offset is attributed to the mass imbalance component.

[0045] In step S7, the blade pitch offset correction amount is determined based on the aerodynamic imbalance component, and the counterweight and counterweight orientation are determined based on the residual mass imbalance component after blade pitch offset correction, so as to perform graded balance correction on the wind turbine rotor.

[0046] Specifically, first determine the ratio of the amplitude of the aerodynamic imbalance component Ua to the amplitude of the current normalized imbalance vector Uc. When this ratio is greater than a first proportional threshold, it indicates that the aerodynamic imbalance accounts for a high proportion of the current rotor imbalance, and blade pitch offset correction is performed first. The first proportional threshold can be 30%-70%, preferably 40%-60%. For example, when |Ua| / |Uc| is greater than 50%, δβ is preferentially corrected. Or with δβ The compensation amount in the opposite direction is written into the pitch control system as the blade pitch offset correction amount.

[0047] The blade pitch offset correction amount can be determined in the following way: δβcorr=-δβ* Here, δβcorr is the blade pitch offset correction amount used to compensate for aerodynamic imbalance components. In practical applications, δβcorr can be smoothed out and applied gradually over several operating cycles to avoid sudden changes in the unit load caused by abrupt changes in the pitch angle.

[0048] After completing the pitch offset correction, valid data is collected again, and the normalized unbalance vector is recalculated to obtain the residual unbalance vector Ur. If Ur is still greater than the preset balance threshold, and the residual unbalance is mainly manifested as a mass unbalance component, the counterweight and counterweight orientation are determined based on the residual mass unbalance component.

[0049] The counterweight weight and orientation can be calculated based on the pre-established counterweight influence coefficient. The counterweight influence coefficient C represents the change in the normalized unbalance vector when a unit mass of counterweight is installed at a preset radius and orientation. Let the residual mass unbalance component be Um, and the counterweight influence coefficient be C, then the theoretical counterweight weight m can be determined as follows: m=|Um| / |C| The orientation ψ of the counterweight can be determined based on the phase of the residual mass imbalance component and the phase of the counterweight influence coefficient: ψ = arg(-Um / C) Here, arg represents the vector phase. The counterweight orientation can correspond to the preset counterweight mounting hole position on the hub, or it can correspond to the preset counterweight mounting position at the blade root, the inner wall of the hub, or inside the fairing.

[0050] When the theoretical counterweight orientation ψ does not perfectly match the preset mounting hole positions, the closest mounting hole position can be selected for counterweighting, or the theoretical counterweight can be distributed according to two adjacent preset mounting hole positions. For example, if the theoretical counterweight orientation is located between the first and second mounting hole positions, the counterweight can be distributed to the two mounting hole positions according to the ratio of the angle between the two mounting hole positions and the theoretical orientation, thereby synthesizing an equivalent counterweight vector.

[0051] In one specific embodiment, the wind turbine is a three-bladed wind turbine with a rated power of 2MW-6MW. Vibration sensors are installed on the outside of the main bearing housing, with a sampling frequency of 500Hz; the rotor azimuth angle is obtained from the encoder signal of the main control system. A data window with wind speeds of 6m / s-10m / s, yaw error less than 8°, and rotational speed fluctuation less than 3% of the average rotational speed is selected as the effective data window. The vibration signal within each effective data window is resampled into 720 equal-angle points according to the rotor azimuth angle, and a 1P vibration vector is extracted through sine fitting.

[0052] In this specific embodiment, the current normalized unbalance vector Uc is first calculated without any pitch disturbance. Then, a +0.3° pitch disturbance is applied to the first blade and held for 120 seconds, and the corresponding normalized unbalance vector U1+ is calculated. After removing the disturbance, a -0.3° pitch disturbance is applied to the first blade and held for 120 seconds, and U1- is calculated. The second and third blades are tested in the same way, and U2+, U2-, U3+, and U3- are obtained respectively. Based on the above data, h1, h2, and h3 are calculated respectively, and an aerodynamic unbalance sensitivity matrix H is established.

[0053] Then, the least squares method is used to solve for the blade pitch offset combination δβ*. If the calculated amplitude of the aerodynamic imbalance component Ua accounts for 65% of the amplitude of the current normalized imbalance vector Uc, then the aerodynamic imbalance is determined to be the primary cause, and pitch offset correction is performed first. For example, if the calculated compensation is +0.18° for the first blade, -0.07° for the second blade, and -0.11° for the third blade, then this set of pitch offset corrections is written into the pitch control system, and the maximum rate of change is set in the control program to apply the correction gradually.

[0054] After pitch offset correction is completed, stable operating data is reacquired and the residual normalized unbalance vector Ur is calculated. If Ur is below the preset balance threshold, the balance correction is considered complete. If Ur is still above the preset balance threshold, the counterweight and counterweight orientation are calculated based on the residual mass unbalance component Um. For example, if the calculation shows that a 1.8kg counterweight needs to be added at a 210° orientation at a preset radius on the hub, and the adjacent preset mounting holes on the hub are at 200° and 230° respectively, the 1.8kg counterweight can be distributed to the two mounting holes at 200° and 230° using vector decomposition to achieve an equivalent correction effect.

[0055] In another implementation, when the calculated magnitude of the mass imbalance component Um is greater than the magnitude of the current normalized imbalance vector Uc than a second proportional threshold, counterweight correction can be performed directly. The second proportional threshold can be 30%-70%, preferably 50%. For example, when |Um| / |Uc| is greater than 60%, and the aerodynamic imbalance component is small, it can be determined that the current imbalance is mainly caused by blade mass differences or hub counterweight deviations. In this case, the counterweight weight and counterweight orientation are calculated first, without the need for significant adjustments to the blade pitch offset.

[0056] In another implementation, when both aerodynamic and mass imbalance components are significant, aerodynamic correction is performed first, followed by mass correction. This is because aerodynamic imbalance can be compensated online through the pitch control system, typically without requiring shutdown and counterweight adjustment. Direct counterweight correction might mistake aerodynamic deviations for mass deviations, resulting in significant 1P vibration even after counterweight adjustment. By correcting the aerodynamic imbalance first and then correcting the residual mass imbalance with counterweight, the number of repeated shutdowns and counterweight adjustments can be reduced.

[0057] After completing the balance correction, a closed-loop verification step is performed. Specifically, vibration signals, rotor azimuth angle signals, and operating parameters during wind turbine operation are collected again, and the corrected normalized unbalance vector is recalculated according to the aforementioned steps. If the corrected normalized unbalance vector is less than the preset balance threshold, or the corrected 1P vibration amplitude is reduced to below a preset ratio compared to before correction, the rotor balance correction is considered complete. The preset balance threshold can be determined based on the turbine model, tower height, rated power, and main bearing vibration limits.

[0058] If the corrected normalized imbalance vector is still greater than the preset balance threshold, the corrected normalized imbalance vector is used as the new current normalized imbalance vector, and blade pitch offset correction and / or counterweight correction are continued. To avoid overcorrection, a maximum number of iterations can be set, for example, no more than 3 times. If the balance requirement cannot be met after exceeding the maximum number of iterations, a check prompt indicating abnormal blade structure, severe blade contamination, sensor malfunction, or hub connection malfunction can be output.

[0059] Through the above implementation method, this approach does not directly apply counterweight based on the 1P vibration phase. Instead, it establishes an aerodynamic imbalance sensitivity matrix using small-amplitude blade pitch disturbances, then decomposes the total rotor imbalance into aerodynamic and mass imbalance components, and processes them separately using blade pitch offset correction and counterweight correction. This effectively distinguishes different sources of imbalance, such as blade pitch zero-position deviation, blade aerodynamic shape differences, blade mass deviation, and hub counterweight deviation, reducing the risk of blindly applying counterweight and improving the accuracy and efficiency of on-site balancing correction of wind turbine rotors.

[0060] The above are merely preferred embodiments of the present invention. Those skilled in the art, inspired by the technical concept of the present invention, may make various adjustments and substitutions to the yarn material, weaving structure, cross angle, anchoring method, heat treatment conditions, and coating structure. All such adjustments and substitutions should be considered to fall within the protection scope of the present invention.

Claims

1. A method for balancing and correcting the rotor of a wind turbine generator, characterized in that, Includes the following steps: S1. Collect vibration signals, rotor azimuth angle signals, and operating parameters of the wind turbine during operation; S2. Based on the rotor azimuth angle signal, perform azimuth synchronization processing on the vibration signal and extract the 1P vibration vector corresponding to the rotor's first rotational frequency; S3. Perform condition normalization processing on the 1P vibration vector according to the operating condition parameters to obtain the normalized unbalance vector. S4. Apply a preset pitch disturbance to each blade of the wind turbine generator and obtain the normalized unbalance vector change of each blade before and after applying the preset pitch disturbance. S5. Establish the aerodynamic imbalance sensitivity matrix based on the normalized imbalance vector change corresponding to each blade. S6. Based on the aerodynamic imbalance sensitivity matrix, decompose the current normalized imbalance vector into aerodynamic imbalance components and mass imbalance components. S7. Determine the blade pitch offset correction amount based on the aerodynamic imbalance component, and determine the counterweight and counterweight orientation based on the residual mass imbalance component after blade pitch offset correction, so as to perform graded balance correction on the wind turbine rotor.

2. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S1, the operating condition parameters include at least two of the following: rotor speed, wind speed, yaw error, generator power, and blade pitch angle.

3. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, Step S1 is followed by a valid data filtering step, which includes selecting a data window that meets the stable operating conditions from the collected data. The stable operating conditions include rotor speed fluctuation value being less than a first preset threshold, wind speed fluctuation value being less than a second preset threshold, yaw error being less than a third preset threshold, and the wind turbine not being in a start-stop, power-limited, or fault-reduced state.

4. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S2, the azimuth synchronization processing includes: converting the vibration signal from time domain data to azimuth domain data based on the rotor azimuth angle, and extracting the first-order frequency component from the azimuth domain data to obtain the 1P vibration vector with amplitude and phase.

5. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S3, the operating condition normalization process includes: establishing operating condition correction coefficients based on at least two of rotor speed, wind speed, yaw error, generator power, and blade pitch angle, and using the operating condition correction coefficients to perform amplitude normalization processing on the 1P vibration vector, and / or performing equivalent conversion on the 1P vibration vector under different operating conditions.

6. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S4, the preset pitch disturbance is a short-term reversible pitch offset applied to a single blade, and the preset pitch disturbance is applied to only one blade within the same disturbance test cycle.

7. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S5, the aerodynamic imbalance sensitivity matrix is ​​composed of the mapping relationship between the pitch disturbance of each blade and the corresponding normalized imbalance vector change, and is used to characterize the degree of influence of the pitch change of each blade on the rotor 1P imbalance state.

8. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S6, decomposing the current normalized unbalance vector into an aerodynamic unbalance component and a mass unbalance component includes: based on the aerodynamic unbalance sensitivity matrix, solving for the blade pitch offset combination that minimizes the residual vector of the current normalized unbalance vector, and taking the vector corresponding to the blade pitch offset combination as the aerodynamic unbalance component, and taking the remaining vector after subtracting the aerodynamic unbalance component from the current normalized unbalance vector as the mass unbalance component.

9. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, In step S7, when the ratio of the amplitude of the aerodynamic imbalance component to the amplitude of the current normalized imbalance vector is greater than the first proportional threshold, the blade pitch offset correction amount is determined first based on the aerodynamic imbalance component; when the ratio of the amplitude of the mass imbalance component to the amplitude of the current normalized imbalance vector is greater than the second proportional threshold, the counterweight and counterweight orientation are determined based on the mass imbalance component.

10. The method for balancing and correcting the rotor of a wind turbine generator according to claim 1, characterized in that, After completing the balance correction, the vibration signals and operating parameters of the wind turbine are collected again during operation. The corrected normalized unbalance vector is recalculated. When the corrected normalized unbalance vector is greater than the preset balance threshold, blade pitch offset correction and / or counterweight correction are performed based on the corrected normalized unbalance vector.

Images