Aerodynamic Unbalance Load Control Method for Wind Turbines Based on Robust Control

Abstract

The present invention provides a method for controlling the aerodynamic unbalance load of a wind turbine based on robust control. The method includes: obtaining the measured spindle load of the nacelle to be measured; transforming the measured spindle load of the nacelle by nacelle coordinate transformation into the effective input of a robust independent pitch controller; establishing a multiplicative perturbation model of the aerodynamic unbalance of the wind turbine; calculating the robust independent pitch controller; and obtaining the input pitch angle of the wind turbine control system by inverse nacelle coordinate transformation. The present invention uses the measured bending moment based on the nacelle as the feedback input of the controller, solves the problem of installing sensors in rotating components, and enables the control cabinet and sensors to be located inside the nacelle; the proposed control strategy is still based on the Coleman transformation and only uses one controller to complete the aerodynamic unbalance control; a robust independent pitch control strategy with multiple inputs and multiple outputs and strong robustness is adopted to solve the problems of incomplete decoupling of traditional controllers and system nonlinearity.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation, and particularly relates to a method for controlling the aerodynamic unbalanced load of a wind turbine based on robust control. Background Art

[0002] Restricted by the manufacturing precision and installation level of wind turbines, uneven initial installation angles and mass distributions of blades will cause imbalance between the blades, increase the load fluctuation at the top of the tower, and result in obvious differences in the flapping loads of each blade. This will not only reduce the performance of the wind turbine, but also increase the fatigue load of the unit and the power generation cost.

[0003] The premise of the traditional independent pitch design of wind turbines is that the three blades are symmetric, and it is difficult to eliminate the impeller unbalanced load due to their completely identical characteristics.

[0004] In view of the above problems, it is necessary to propose a method for controlling the aerodynamic unbalanced load of a wind turbine based on robust control, which is reasonable in design and can effectively solve the above problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art, and provides a method for controlling the aerodynamic unbalanced load of a wind turbine based on robust control.

[0006] The present invention provides a method for controlling the aerodynamic unbalanced load of a wind turbine based on robust control, and the method includes:

[0007] Obtain the measured main shaft load of the nacelle to be measured;

[0008] Transform the measured main shaft load of the nacelle into an effective input of a robust independent pitch controller by using nacelle coordinate transformation;

[0009] Establish a multiplicative perturbation model of the aerodynamic unbalance of the wind turbine;

[0010] Calculate the robust independent pitch controller;

[0011] Obtain the input pitch angle of the wind turbine control system by using the inverse nacelle coordinate transformation.

[0012] Optionally, the step of transforming the measured main shaft load of the nacelle into an effective input of a robust independent pitch controller by using nacelle coordinate transformation includes:

[0013] The measured main shaft load of the nacelle includes the bending moment M in the y direction of the nacelle main shaft y and the bending moment M in the z direction of the nacelle main shaft z ,

[0014] The transformation equation of the coordinate transformation is:

[0015]

[0016] Among them, is the azimuth angle of the wind turbine rotor, M y2 is the bending moment obtained through coordinate transformation, M y3 is the bending moment obtained through coordinate transformation.

[0017] Optionally, establishing the multiplicative perturbation model of the aerodynamic imbalance of the wind turbine generator set includes:

[0018] Using the input multiplicative perturbation model to represent the uncertainty of the unit model as a combination of a linear deterministic model and a multiplicative perturbation factor;

[0019] Regarding the harmonic components in the bending moment M y in the y - direction of the nacelle main shaft and the bending moment M z in the z - direction of the nacelle main shaft as the uncertainty part of the unit load model, and obtaining the input multiplicative perturbation model of the overturning moment, yaw moment, and pitch angle in the fixed coordinate system. Among them, in the multiplicative perturbation model, the actual control object is:

[0020] Among them,

[0021] is the actual control object, P is the nominal control object (state - space model), Δ(s) is the scale factor, and W(s) is the weight factor.

[0022] Optionally, the structure of the scale factor Δ(s) is:

[0023]

[0024] The structure of the weight factor W(s) is:

[0025]

[0026] Among them, Δd(s) is the component of the scale factor on the d - axis, Δq(s) is the component of the scale factor on the q - axis, Wd(s) is the component of the weight factor on the d - axis, and Wq(s) is the component of the weight factor on the q - axis.

[0027] Optionally, before calculating the robust independent pitch controller, the method further includes:

[0028] Using the μ - synthesis method to solve, and equivalently representing the robust independent pitch controller K as:

[0029] K(s) = [K y (s) K r (s)],

[0030] Among them, K is the robust independent pitch controller, K y is the feedback part of the robust independent pitch controller, K r is the feedforward part of the robust independent pitch controller;

[0031] According to the feedback part K of the robust independent pitch controller y and the feedforward part K of the robust independent pitch controller r , design the input weighting function W of the robust independent pitch controller u and the output weighting function W p , where

[0032]

[0033] where, e p is the weighted output of the overturning moment and the yaw moment, e u is the weighted output of the pitch angle, S is the sensitivity function of the system, T is the complementary sensitivity function of the system, r is the reference input, n is the measurement noise of the sensor, I is the identity matrix, and M is the reference model.

[0034] Optionally, the sensitivity function S of the system is:

[0035] S = (I + PK) -1 ,

[0036] where, K is the robust independent pitch controller, I is the identity matrix, and P is the nominal control object (state space model);

[0037] The complementary sensitivity function T of the system is:

[0038] S = (I + PK) -1 PK y ,

[0039] where, K is the robust independent pitch controller, I is the identity matrix, P is the nominal control object (state space model), and Ky is the feedback part of the robust independent pitch controller.

[0040] Optionally, the performance target of the robust independent pitch controller needs to be satisfied:

[0041]

[0042] Optionally, before calculating the robust independent pitch controller, the method further includes:

[0043] Using the structured singular value μ to synthesize an independent pitch control loop;

[0044] To solve the robust independent pitch controller K, the module structure Δp is defined as:

[0045]

[0046] Δ is the uncertainty module of the system, and ΔF is the fictitious module;

[0047] To meet the design objective of the robust independent pitch controller K that is stable, for each frequency ω ∈ [0, ∞], the structured singular value must satisfy:

[0048]

[0049] where F L is the lower linear fractional transformation, P is the nominal control object (state - space model), K is the robust independent pitch controller, j is the imaginary symbol, and ω is the frequency.

[0050] Optionally, before calculating the robust independent pitch controller, the method further includes:

[0051] Designing a reference model M for the robust independent pitch controller, and the reference model M is:

[0052]

[0053] where T is the time constant and ξ is the damping ratio.

[0054] Optionally, the step of obtaining the input pitch angle of the wind turbine control system by using the inverse transformation of nacelle coordinates includes:

[0055] Obtaining the blade pitch angles β1, β2, β3 according to the inverse transformation of nacelle coordinates, and adding them to the pitch angle β of the collective pitch controller c to obtain the input pitch angle of the unit control system; where the inverse transformation equation of nacelle coordinates is:

[0056]

[0057] where β i (i = 1, 2, 3) is the blade pitch angle, is the azimuth angle of the wind turbine, where β d is the pitch angle of the d - axis in the d - q coordinate system, and β q is the pitch angle of the q - axis in the d - q coordinate system.

[0058] A method for controlling the aerodynamic unbalance load of a wind turbine based on robust control according to an embodiment of the present invention is used to suppress the aerodynamic unbalance load of the wind turbine rotor. In this control method, the measured spindle load of the nacelle to be measured is transformed into an effective input of a robust independent pitch controller by using nacelle coordinate transformation. Different from the traditional pitch controller based on the measurement of the blade root bending moment, the present invention uses the measured moment based on the nacelle as the feedback input of the controller, solves the problem of installing sensors in rotating components, makes the control cabinet and sensors both located inside the nacelle, reduces the difficulty of sensor layout, and at the same time reduces the number of sensors, reduces the risk of sensor failure, and improves the reliability of measurement; different from the method of improving the Coleman coordinate transformation adopted by the existing aerodynamic unbalance load control method, the proposed control strategy is still based on the Coleman transformation and only uses one controller to complete the aerodynamic unbalance control; different from the traditional multiple single-input single-output PI independent pitch controllers, the present invention adopts a multi-input multi-output and strongly robust robust independent pitch control strategy, which solves the problems of incomplete decoupling and system nonlinearity of the traditional controller. Using this control method can suppress the aerodynamic unbalance load of the wind turbine caused by the blade installation process level, and can also reduce the power generation cost caused by the increase of the fatigue load of the wind turbine. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 Schematic diagram of the flow of a method for controlling the aerodynamic unbalance load of a wind turbine based on robust control according to an embodiment of the present invention;

[0060] Figure 2 Schematic diagram of the structure of an improved robust independent pitch control strategy for impeller imbalance according to another embodiment of the present invention;

[0061] Figure 3 Schematic diagram of the structure of the uncertainty model of a wind turbine according to another embodiment of the present invention;

[0062] Figure 4 Schematic diagram of the control structure diagram of a robust independent pitch controller according to another embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0064] As Figure 1 shown, the present invention provides a method S100 for controlling the aerodynamic unbalance load of a wind turbine based on robust control. The monitoring method S100 includes:

[0065] S110. Obtain the measured spindle load of the nacelle to be measured.

[0066] Specifically, in this embodiment, the measured spindle load of the nacelle to be measured is the bending moment M in the y direction of the spindle of the nacelle to be measured y and the bending moment M in the z direction of the spindle z .

[0067] S120. Transform the measured spindle load of the nacelle by nacelle coordinate transformation into the effective input of the robust independent pitch controller.

[0068] Specifically, as Figure 2 shown, first, it is necessary to perform coordinate transformation on the measured nacelle load measurement to provide effective input for the robust independent pitch controller. Different from the Coleman transformation in the common independent pitch control strategy, in the present invention, the nacelle coordinate transformation in Equation (1) is used to transform the bending moments M y and M z in the y and z directions of the nacelle spindle into the feedback inputs M y2 and M y3 of the robust pitch controller. The transformation equation of the coordinate transformation is:

[0069]

[0070] where is the wind turbine azimuth angle, M y2 is the bending moment obtained by coordinate transformation, and M y3 is the bending moment obtained by coordinate transformation.

[0071] The robust independent pitch controller formed by this coordinate system does not depend on the measurement of the rotating variable, and its feedback input is based on the nacelle spindle, reducing the difficulty of sensor arrangement. At the same time, the number of sensors is reduced from 3 to 2, reducing the risk of sensor failure and improving the reliability of measurement.

[0072] S130. Establish the multiplicative perturbation model of the aerodynamic imbalance of the wind turbine.

[0073] Specifically, regard the harmonic components in the bending moment M y in the y direction of the nacelle spindle and the bending moment M z in the z direction of the nacelle spindle as the uncertain part of the unit load model, and obtain the input multiplicative perturbation model of the overturning moment, yaw moment and pitch angle in the fixed coordinate system in Equation (2). Among them, in the multiplicative perturbation model, the actual control object is:

[0074]

[0075] is the actual control object, P is the nominal control object (state space model), Δ(s) is the scale factor, and W(s) is the weight factor.

[0076] Optionally, the structure of the scale factor Δ(s) is as follows:

[0077]

[0078] The structure of the weight factor W(s) is as follows:

[0079]

[0080] where Δd(s) is the component of the scale factor on the d-axis, Δq(s) is the component of the scale factor on the q-axis, Wd(s) is the component of the weight factor on the d-axis, and Wq(s) is the component of the weight factor on the q-axis.

[0081] Based on the linearized state-space model, the nominal model and the weight factor are calculated by selecting an appropriate perturbation range according to formula (2), and the multiplicative perturbation model of the aerodynamic imbalance of the unit is established.

[0082] As Figure 3 shown, the system model considering the unbalanced load of the wind turbine can be expressed as a combination of a deterministic model and an uncertain model. Then, based on the multiplicative perturbation model, a two-degree-of-freedom robust independent pitch controller is designed. The robust independent pitch controller takes the measured bending moment and the reference bending moment as inputs and the pitch angle as the output. The control structure of the robust independent pitch controller is as Figure 4 shown. As Figure 4 shown, G is the generalized model, including the model and the interconnection structure between the model and the robust independent pitch controller. The interconnection structure includes a weighting function to facilitate further loop shaping.

[0083] S140. Calculate the robust independent pitch controller.

[0084] Before solving the controller K, it is necessary to design the weight functions W p and W u and the reference model M so that the independent pitch controller has the desired closed-loop dynamic performance.

[0085] Specifically, the μ-synthesis method is used to solve, and the robust independent pitch controller K is equivalently expressed as:

[0086] K(s) = [K y (s) K r (s)], (5)

[0087] where K is the robust independent pitch controller, K y is the feedback part of the robust independent pitch controller, and K r is the feedforward part of the robust independent pitch controller;

[0088] According to the feedback part K of the robust independent pitch controller y and the feedforward part K of the robust independent pitch controller r , design the input weighting function W u and the output weighting function W p of the robust independent pitch controller, where

[0089]

[0090] where, e p is the weighted output of the overturning moment and the yaw moment, e u is the weighted output of the pitch angle, S is the sensitivity function of the system, T is the complementary sensitivity function of the system, r is the reference input, n is the measurement noise of the sensor, I is the identity matrix, and M is the reference model.

[0091] The weight functions W p and W u are also transfer functions, and their component weights are different in different frequency domains. The weight function W u will penalize the output of the robust independent pitch controller, aiming to limit the action amount of the pitch actuator. At the same time, high-frequency control actions that reach the rate limit of the pitch actuator should also be avoided. Therefore, when selecting W u , it is necessary to ensure high gain at frequencies above the actuator bandwidth and low gain at frequencies below the actuator bandwidth. The weight function W p will weight the control output error. At a given frequency, high gain can reduce the sensitivity at that frequency, thereby generating a high controller gain to improve the disturbance rejection ability.

[0092] Exemplarily, the sensitivity function S of the system is:

[0093] S = (I + PK) -1 , (7)

[0094] where, K is the robust independent pitch controller, I is the identity matrix, and P is the nominal controlled object (state space model);

[0095] The complementary sensitivity function T of the system is:

[0096] S = (I + PK) -1 PK y , (8)

[0097] where, K is the robust independent pitch controller, I is the identity matrix, P is the nominal controlled object (state space model), and Ky is the feedback part of the robust independent pitch controller.

[0098] Exemplarily, the performance objective of the robust independent pitch controller needs to be satisfied:

[0099]

[0100] Exemplarily, before calculating the robust independent pitch controller, the method further includes:

[0101] Using the structured singular value μ synthesis to form an independent pitch control loop;

[0102] To solve the robust independent pitch controller K, define the module structure Δp as:

[0103]

[0104] Δ is the uncertainty module of the system, and ΔF is the fictitious module;

[0105] To meet the design goal of the stable robust independent pitch controller K, for each frequency ω ∈ [0, ∞], the structured singular value needs to satisfy:

[0106]

[0107] where, F L is the lower linear fractional transformation, P is the nominal control object (state - space model), K is the robust independent pitch controller, j is the imaginary symbol, and ω is the frequency.

[0108] Exemplarily, before calculating the robust independent pitch controller, the method further includes:

[0109] Design a reference model M for the robust independent pitch controller, and the reference model M is:

[0110]

[0111] where, T is the time constant and ξ is the damping ratio.

[0112] The reference model M connects the reference signal and the output signal, improving the performance and robust stability of the robust independent pitch controller. Considering the response characteristics of the wind turbine and the adjustment ability of the pitch system comprehensively, the desired closed - loop system dynamic response is achieved by designing a stable model M.

[0113] Select the same coefficients T and ξ of the transfer function in the reference model to make the two channels have similar dynamic characteristics, and the robust independent pitch controller K can be calculated by solving the mixed - sensitivity problem through the D - K iteration algorithm.

[0114] S150. Adopt the inverse transformation of the nacelle coordinates to obtain the input pitch angle of the wind turbine control system.

[0115] Specifically, using the inverse coordinate transformation of formula (13), the blade pitch angles β1, β2, β3 are obtained, and are superimposed with the pitch angle β of the centralized pitch controller to obtain the input pitch angle of the unit control system; wherein, the nacelle coordinate inverse transformation equation is: c Among them, β

[0116]

[0117] where β i (i = 1, 2, 3) is the blade pitch angle, is the wind turbine azimuth angle, where β d is the pitch angle of the d-axis in the d-q coordinate system, and β q is the pitch angle of the q-axis in the d-q coordinate system.

[0118] Different from the method of improving the Coleman coordinate transformation adopted by the existing aerodynamic unbalance load control method, the proposed control strategy is still based on the Coleman transformation, and only one controller is used to complete the aerodynamic unbalance control.

[0119] It can be understood that the above embodiments are merely exemplary embodiments adopted to illustrate the principle of the present invention. However, the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also regarded as the protection scope of the present invention.

Claims

1. A method for controlling the aerodynamic unbalanced load of a wind turbine based on robust control, characterized in that, The method includes: Obtaining the measured main shaft load of the nacelle to be measured; Converting the measured main shaft load of the nacelle into an effective input of the robust independent pitch controller by using nacelle coordinate transformation, specifically including: The measured spindle load of the nacelle includes the bending moment M in the y direction of the spindle of the nacelle y and the bending moment M in the z direction z , The transformation equation of the coordinate transformation is: Among them, is the azimuth angle of the wind turbine rotor, M y2 is the bending moment obtained through coordinate transformation, M y3 is the bending moment obtained through coordinate transformation; Establishing the multiplicative perturbation model of the aerodynamic imbalance of the wind turbine, specifically including: Representing the uncertainty of the unit model as a combination of a linear deterministic model and a multiplicative perturbation factor by using the input multiplicative perturbation model; The bending moment M in the y-direction of the main shaft of the nacelle y and the bending moment M in the z-direction of the main shaft of the nacelle z The harmonic components in are regarded as the uncertain part of the unit load model, and an input product perturbation model of the overturning moment, yaw moment, and pitch angle in the fixed coordinate system is obtained. Among them, in the product perturbation model, the actual controlled object is: Among them, The actual control object is \(P_a\), the nominal control object is \(P\), \(\Delta(s)\) is the scale factor, and \(W(s)\) is the weight factor; Calculating the robust independent pitch controller; Obtaining the input pitch angle of the wind turbine control system by using the inverse nacelle coordinate transformation, specifically including: According to the inverse transformation of the nacelle coordinates, the blade pitch angles β1, β2, and β3 are obtained and superimposed with the pitch angle β of the centralized pitch controller to obtain the input pitch angle of the unit control system; among them, the inverse transformation equation of the nacelle coordinates is: c superimposed to obtain the input pitch angle of the unit control system; wherein, the inverse transformation equation of the nacelle coordinates is: where, β i (i = 1, 2, 3) is the blade pitch angle, is the wind turbine azimuth angle, where β d is the pitch angle of the d-axis in the d-q coordinate system, and β q is the pitch angle of the q-axis in the d-q coordinate system.

2. The control method according to claim 1, characterized in that The structure of the scale factor Δ(s) is: The structure of the weight factor W(s) is: Where, Δd(s) is the component of the scale factor on the d-axis, Δq(s) is the component of the scale factor on the q-axis, Wd(s) is the component of the weight factor on the d-axis, and Wq(s) is the component of the weight factor on the q-axis.

3. The control method according to claim 1, wherein Before calculating the robust independent pitch controller, the method further includes: Solving by using the μ-synthesis method and equivalently representing the robust independent pitch controller K as: K(s) = [K y (s) K r (s)], Among them, K is the robust independent pitch controller, and K y is the feedback part of the robust independent pitch controller, and K r is the feedforward part of the robust independent pitch controller; According to the feedback part K of the robust independent pitch controller y and the feedforward part K of the robust independent pitch controller r , design the input weighting function W of the robust independent pitch controller u and the output weighting function W p , where where, e p is the weighted output of the overturning moment and yaw moment, e u is the weighted output of the pitch angle, S is the sensitivity function of the system, T is the complementary sensitivity function of the system, r is the reference input, n is the measurement noise of the sensor, I is the identity matrix, and M is the reference model.

4. The control method according to claim 3, wherein The sensitivity function S of the system is: S = (I + PK) -1 , Where, K is the robust independent pitch controller, I is the identity matrix, and P is the nominal controlled object; The complementary sensitivity function T of the system is: T = (I + PK) -1 PK y , Where, K is the robust independent pitch controller, I is the identity matrix, P is the nominal controlled object, and Ky is the feedback part of the robust independent pitch controller.

5. The control method according to claim 3, wherein The performance objective of the robust independent pitch controller needs to satisfy:

6. The control method according to claim 1, characterized in that Before calculating the robust independent pitch controller, the method further includes: Using the structured singular value μ-synthesis to form an independent pitch control loop; To solve the robust independent pitch controller K, define the module structure Δp as: Δ is the uncertainty module of the system, and ΔF is the fictitious module; To meet the design objective of the stable robust independent pitch controller K, for each frequency ω∈[0,∞], the structured singular value needs to satisfy: where F L is a lower linear fractional transformation, P is a nominal controlled object (state space model), K is the robust independent pitch controller, j is the imaginary symbol, and ω is the frequency.

7. The control method according to claim 6, characterized in that Before calculating the robust independent pitch controller, the method further includes: Designing the reference model M of the robust independent pitch controller, and the reference model M is: Where, T is the time constant and ξ is the damping ratio.

Images