A data-driven independent pitch control method and system

Abstract

The application discloses a kind of data-driven independent variable pitch control method and system, pitch moment, yaw moment and low-pass filtered wind wheel speed signal are stored as output vector, and pitch angle closed loop compensation is stored as input vector, and equivalent identification parameter matrix is constructed;Using the input vector of k-1 time and the output vector of k time to solve the data-driven control quantity under the fixed hub coordinate system, and then the pitch angle feedback control quantity is obtained;Using the wind wheel speed signal combined with azimuth angle signal after band-pass filtering to calculate the pitch angle feedforward compensation;The total pitch angle control quantity obtained by adding the pitch angle feedback control quantity, realizes independent variable pitch control.The application comprehensively considers the mutual influence between blade root bending moment and wind wheel speed, and the control method is significantly better than the unified variable pitch based on speed and the independent variable pitch based on blade root bending moment designed respectively.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wind power generation, and particularly relates to a data-driven independent variable pitch control method and system. BACKGROUND

[0002] With the increasing demand for clean energy and the increasing support of the country for the wind power industry, large-scale is the inevitable direction of the development of wind turbine, and the development is accelerating. However, the large-scale of wind turbine will greatly increase the unevenness of the wind flow into the wind wheel, especially the wind shear effect, the tower shadow effect and the turbulence, which will make the wind load on the unit complex and unevenly distributed, and further affect the safe and reliable operation of the wind turbine and the fatigue life of each component. Therefore, reducing the load of the wind turbine has great significance for further promoting the large-scale of the wind turbine, reducing the cost of per kilowatt-hour of electricity and enhancing the wind power bidding capacity.

[0003] Independent variable pitch control is the mainstream active load control technology of wind turbine. The most traditional independent variable pitch control technology uses multi-blade coordinate transformation to convert the blade load in the rotating coordinate system to the fixed coordinate system. The transformed load is divided into pitch and yaw torque components, and then two classic PID controllers are designed to control the pitch and yaw torque, thereby realizing the suppression of blade fatigue load. Using multi-blade coordinate transformation can only partially decouple the pitch and yaw torque components, and designing PID controllers separately cannot consider the coupling relationship. The pitch and yaw torque control loops will interfere with each other. The existing technology also uses fuzzy PID, neural network PID and particle swarm optimization PID to replace the classic PID to realize nonlinear control, but it still does not consider the coupling relationship between the pitch and yaw torque. In addition, a small amount of existing technology applies multivariable control technology to realize the coupling effect of pitch and yaw torque, such as model predictive control, linear quadratic optimal control and linear variable parameter control. However, these technologies are based on simplified mechanism models or linearized models extracted based on small perturbation assumption at a typical steady-state operating point, and the model fidelity needs to be evaluated. The reliability of the controller designed for these simplified mechanism models and linearized models is not high and the applicable range of working conditions is narrow.

[0004] Therefore, in order to improve the load reduction effect and reliability of variable pitch control, it is necessary to design an independent variable pitch control strategy that considers the coupling relationship between pitch and yaw torque and is widely applicable in a wide range of working conditions. SUMMARY

[0005] The technical problem to be solved by the application is to provide a data-driven independent variable pitch control method and system to reduce the power fluctuation and fatigue load of the floating wind turbine above the rated wind speed, so as to solve the technical problem of complex and strong coupling of the load of the floating wind turbine.

[0006] The application adopts the following technical solutions:

[0007] A data-driven independent variable pitch control method comprises the following steps:

[0008] S1, obtaining an azimuth angle signal of a wind turbine blade, a blade root bending moment signal and a wind wheel speed signal;

[0009] S2, performing low-pass filtering on the wind wheel speed signal obtained in step S1 by using a first-order inertia link to obtain a filtered wind wheel speed signal;

[0010] S3, transforming the blade root bending moment signal obtained in step S1 into a hub coordinate system by using an inverse Kalman coordinate transformation to obtain a pitch moment and a yaw moment;

[0011] S4, storing the wind wheel speed signal obtained in step S2, the pitch moment and the yaw moment obtained in step S3 as an output vector, and storing a pitch angle closed-loop compensation quantity as an input vector;

[0012] S5, constructing an equivalent identification parameter matrix based on the input vector and the output vector obtained in step S4;

[0013] S6, solving a data-driven control quantity in a fixed hub coordinate system based on the input vector at time k-1 and the output vector at time k obtained in step S4 by using the equivalent identification parameter matrix obtained in step S5;

[0014] S7, converting the data-driven control quantity in the fixed hub coordinate system obtained in step S6 into a rotating blade coordinate system by using a Kalman coordinate transformation to obtain a pitch angle feedback control quantity;

[0015] S8, performing band-pass filtering on the wind wheel speed signal obtained in step S2 to obtain a band-pass filtered wind wheel speed signal, and calculating a pitch angle feedforward compensation quantity in combination with the azimuth angle signal obtained in step S1;

[0016] S9, adding the pitch angle feedback control quantity obtained in step S7 and the pitch angle feedforward compensation quantity obtained in step S8 to obtain a total pitch angle control quantity, and realizing independent variable pitch control according to the total pitch angle control quantity.

[0017] Specifically, in step S3, the pitch moment e and the yaw moment f are respectively:

[0018]

[0019]

[0020] wherein b1, b2 and b3 are blade root bending moment signals, and a1, a2 and a3 are blade azimuth angle signals.

[0021] Specifically, in step S4, the output vector y and the input vector u are respectively:

[0022] y = [d, e, f]

[0023] u = [g1, g2, g3]

[0024] wherein d is a wind wheel speed signal, e is a pitch moment, f is a yaw moment, g1, g2, g3 are pitch angle closed-loop compensation quantities.

[0025] Specifically, in step S5, the equivalent identification parameter matrix Φ(k) is:

[0026] Φ(k) = [Δy(k) Δu T (k-1) + hΦ(k-1)] [Δu(k-1) Δu T (k-1) + h] -1

[0027] wherein k is a time step, Δy(k) is a difference between the output vector at the current time and the output vector at the last time, Δu T (k-1) is a transpose of a difference between the input vector at the k-1 time and the output vector at the k-2 time, h is a normal number, and Φ(k-1) is a 3*3 dynamic time-varying parameter matrix at the k-1 time.

[0028] Specifically, in step S6, the data-driven control quantity u(k) in the fixed wheel hub coordinate system is:

[0029] u(k) = u(k-1) - [Φ(k)IΦ(k) + K + L] -1 Φ(k)I[y(k) - y d (k+1)] - [K + Φ(k)IΦ(k) + L] -1 Lu(k-1)

[0030] wherein y d (k+1) is a set value of the output vector at the k+1 time; I, K and L are all positive definite quadratic form constant matrices, u(k-1) is the input vector at the k-1 time, y(k) is the output vector at the k time, and Φ(k) is the equivalent identification parameter matrix.

[0031] Specifically, in step S7, the pitch angle feedback control quantities m1, m2 and m3 are respectively:

[0032] m1 = u1(k) + u2(k)cosa1 + u3(k)sina1

[0033] m2 = u1(k) + u2(k)cosa2 + u3(k)sina2

[0034] m3 = u1(k) + u2(k)cosa3 + u3(k)sina3

[0035] Wherein, a1, a2, a3 are blade azimuth angle signals, u1(k), u2(k), u3(k) are three-dimensional elements of data-driven control quantity u(k) respectively.

[0036] Specifically, in step S8, the pitch angle feedforward compensation quantity o1, o2, o3 are respectively:

[0037] o1 = p*n*cosa1

[0038] o2 = p*n*cosa2

[0039] o3 = p*n*cosa3

[0040] Wherein, p is a positive proportionality constant, a1, a2, a3 are blade azimuth angle signals, n is a wind wheel speed signal filtered by a band-pass filter.

[0041] Further, the frequency of the band-pass filter is 3 times the wind wheel rotation frequency.

[0042] Specifically, in step S9, the total pitch angle control quantity q1, q2, q3 are respectively:

[0043] q1 = m1 + o1

[0044] q2 = m2 + o2

[0045] q3 = m3 + o3

[0046] Wherein, o1, o2, o3 are pitch angle feedforward compensation quantities, m1, m2, m3 are pitch angle feedback control quantities respectively.

[0047] In a second aspect, the embodiment of the present application provides a data-driven independent variable pitch control system, comprising:

[0048] The acquisition module obtains the azimuth angle signal of the wind turbine blade, the blade root bending moment signal and the wind wheel speed signal;

[0049] The transformation module performs low-pass filtering on the wind wheel speed signal obtained by the acquisition module by using a first-order inertia link to obtain a filtered wind wheel speed signal; the blade root bending moment signal obtained by the acquisition module is transformed into a hub coordinate system by using inverse Kalman coordinate transformation to obtain a pitch moment and a yaw moment; the wind wheel speed signal, the pitch moment and the yaw moment are stored as an output vector, and the pitch angle closed-loop compensation quantity is stored as an input vector;

[0050] A matrix module constructs an equivalent identification parameter matrix based on the input vector and the output vector obtained by the transformation module, and solves the data-driven control quantity in the fixed hub coordinate system by using the input vector at the k-1 moment and the output vector at the k moment;

[0051] A transformation module converts the data-driven control quantity in the fixed hub coordinate system obtained by the matrix module into the rotating blade coordinate system by using the Kalman coordinate transformation to obtain the pitch angle feedback control quantity;

[0052] A control module performs band-pass filtering on the wind wheel rotating speed signal obtained by the acquisition module to obtain a band-pass filtered wind wheel rotating speed signal, and calculates a pitch angle feedforward compensation quantity in combination with the azimuth angle signal; the total pitch angle control quantity obtained by adding the pitch angle feedback control quantity obtained by the transformation module and the pitch angle feedforward compensation quantity is used to realize independent pitch control according to the total pitch angle control quantity.

[0053] Compared with the prior art, the present application has at least the following beneficial effects:

[0054] A data-driven independent pitch control method comprises two control links, namely an azimuth angle feedforward control link and a data-driven feedback control link; the feedforward control link generates a pitch angle increment related to the azimuth angle, which is used to eliminate the periodic load caused by the known wind shear effect; the data-driven feedback control link simultaneously realizes multivariable closed-loop control of the blade root bending moment and the wind wheel rotating speed, first uses inverse Kalman coordinate transformation to transform the blade root bending moment into the pitch moment and the yaw moment in the fixed hub coordinate system, and performs low-pass filtering on the rotating speed signal to reduce the influence of high-frequency disturbance on the pitch control, then stores and updates the pitch angle closed-loop compensation quantity, the rotating speed, the pitch moment and the yaw moment in the data storage unit, constructs an online data model, designs a data-driven control quantity based on the data model, and converts the control quantity into the pitch angle feedback control quantity in the blade coordinate system through Kalman coordinate transformation, adds the pitch angle feedback control quantity and the azimuth angle feedforward compensation quantity, and sends the total pitch angle control quantity obtained to the pitch angle actuator; the present application combines the model-based disturbance feedforward control with the closed-loop control based on data driving, which not only ensures the response speed of the pitch control to the periodic load caused by the wind shear effect and the tower shadow effect, but also can online identify the time-varying relationship between the internal variables of the wind power system, and the controller designed based on the data model can avoid the control error caused by the simplified mechanism model; in addition, the present application comprehensively considers the mutual influence between the blade root bending moment and the wind wheel rotating speed, and the control idea is significantly superior to the unified pitch based on rotating speed and the independent pitch based on the blade root bending moment.

[0055] Further, through the coordinate transformation related to the azimuth angle, the blade root bending moment in the rotating blade coordinate system is converted into the pitch moment e and the yaw moment f in the fixed hub coordinate system, which can reduce the influence of the time-varying azimuth angle on the controller design and improve the identification and control accuracy of the system.

[0056] Further, the wind rotor speed d, the pitch moment e and the yaw moment f are selected to form the output vector, the coupling relationship of the wind rotor speed, the pitch moment and the yaw moment is comprehensively considered, and the multi-objective control is facilitated; the closed-loop compensation amount of the pitch angle in the fixed hub coordinate system is selected as the input vector, and the influence of the time-varying azimuth angle on the system parameter identification is reduced.

[0057] Further, the identification iteration process of the parameter matrix Φ(k) comprehensively considers the identification accuracy and the time-varying nature of the parameters as the performance index, limits the variation rate of the parameter matrix under the premise of ensuring the identification accuracy, and avoids the identification error caused by the possible matrix singularity problem in the parameter identification process.

[0058] Further, the iteration process of the data-driven control amount u(k) in the fixed hub coordinate system comprehensively considers the control accuracy, the pitch speed and the pitch amplitude performance index, avoids the problems of excessive additional inertia load, excessive pitch motor current and serious heating caused by excessive pitch speed, avoids the system instability caused by excessive pitch amplitude, and introduces a quadratic matrix to plan the weights of each control error vector, input vector and input variation rate vector, which greatly increases the flexibility of the weight distribution of each index.

[0059] Further, the Kalman coordinate transformation is used to convert the data-driven control amount in the fixed hub coordinate system into the pitch angle feedback control amount m1, m2 and m3 in the rotating blade coordinate system, which is beneficial to the independent completion of the respective pitch instructions by the single pitch actuator.

[0060] Further, the pitch angle feedforward compensation amount o1, o2 and o3 is calculated according to the azimuth angle signals a1, a2 and a3 obtained from the band-pass filtered wind rotor speed signal n and S1, and the size of the band-pass filtered speed signal n can represent the periodic load amplitude introduced by the wind shear effect, and the pitch angle feedforward compensation amount o1, o2 and o3 calculated by the band-pass filtered speed signal n, the proportional coefficient p and the cosine of the azimuth angle can eliminate the periodic load caused by the wind shear effect to a certain extent.

[0061] Further, the band-pass frequency of the band-pass filter is 3 times the rotor rotation frequency, the periodic load caused by the wind shear effect is related to the azimuth angle, and it introduces a load of 3 times the rotor rotation frequency in the rotor speed, and the size of the band-pass filtered speed signal n can represent the periodic load amplitude introduced by the wind shear effect.

[0062] Further, the feedforward control amount and the feedback control amount are added to obtain a total pitch angle control amount, the feedforward compensation amount responds to the load disturbance introduced by the known wind shear, has fast response speed and good stability, and the pitch angle feedback control amount is calculated through data-driven closed-loop control, and has high control precision.

[0063] It can be understood that the beneficial effects of the second aspect described above can be referred to the related description in the first aspect described above, which will not be repeated here.

[0064] In summary, the model-based disturbance feedforward control and the data-driven closed-loop control are combined in the application, the response speed of the pitch control to the periodic load caused by the wind shear effect and the tower shadow effect is ensured, and the time-varying relationship of the internal variables of the wind power system can be identified online, so that the flexibility of the weight distribution of each index is greatly increased.

[0065] The technical solutions of the application will be further described in detail below with the help of the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 The method flowchart of the application. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the application will be clearly and completely described below with the help of the accompanying drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0068] In the description of the application, it should be understood that the terms "include" and "contain" indicate the existence of described features, whole, steps, operations, elements and / or components, but do not exclude the existence or addition of one or more other features, whole, steps, operations, elements, components and / or sets thereof.

[0069] It should also be understood that the terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in the specification and the appended claims of the application, unless otherwise clear from the context, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0070] It should be further understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, and that the term "or" as used herein refers to and encompasses any one of the associated items and that the term "at least one of" as used herein refers to and encompasses any one of the associated items or combination of one or more of the associated items.

[0071] It should be understood that, although the terms first, second, third, etc. can be used herein to describe various ranges, etc., these ranges should not be limited by these terms. These terms are only used to distinguish one range from another. For example, a first range could be termed a second range without departing from the scope of the embodiments.

[0072] The word "if" as used herein means "when" or "upon" or "in response to a determination" or "in response to a detection," depending on the context. Similarly, the phrase "if it is determined" or "if [a stated condition or event] is detected" can mean "when it is determined" or "in response to a determination" or "when [a stated condition or event] is detected" or "in response to a detection [of a stated condition or event]," depending on the context.

[0073] Various structural diagrams according to the disclosed embodiments of the present application are shown in the accompanying drawings. These diagrams are not drawn to scale, in which certain details are shown in a somewhat exaggerated manner for purposes of clarity and understanding, and certain details can be omitted. The shapes and relative sizes of the various regions, layers, and their relative positions shown in the drawings are merely exemplary, and in actuality can be deviated due to manufacturing tolerances or technical limitations, and regions / layers with different shapes, sizes, and relative positions can be additionally designed by those skilled in the art according to actual needs.

[0074] The application provides a data-driven independent variable pitch control method, adopts an azimuth angle feedforward control loop and a data-driven feedback control loop, performs low-pass filtering on a wind wheel rotating speed signal by using a first-order inertia link to obtain a filtered wind wheel rotating speed signal, transforms blade root bending moment signals to a hub coordinate system by using inverse Kalman coordinate transformation to obtain a pitching moment and a yawing moment, stores the wind wheel rotating speed signal, the pitching moment and the yawing moment as an output vector, stores a pitch angle closed loop compensation quantity as an input vector, constructs an equivalent identification parameter matrix based on the input vector and the output vector, solves a data-driven control quantity in a fixed hub coordinate system by using the input vector at k-1 time and the output vector at k time, converts the data-driven control quantity in the fixed hub coordinate system to a rotating blade coordinate system by using Kalman coordinate transformation to obtain a pitch angle feedback control quantity, calculates a pitch angle feedforward compensation quantity by using a band-pass filtered wind wheel rotating speed signal combined with an azimuth angle signal, and adds the pitch angle feedback control quantity to obtain a total pitch angle control quantity, so that the independent variable pitch control is realized.

[0075] Please refer to Figure 1 The application provides a data-driven independent variable pitch control method, which comprises the following steps:

[0076] S1, detecting three blade azimuth angle signals a1, a2 and a3, three blade root bending moment signals b1, b2 and b3, and a wind wheel rotating speed signal c;

[0077] S2, performing low-pass filtering on the rotating speed signal c obtained in the step S1 by using a first-order inertia link to reduce the influence of high-frequency disturbance on variable pitch control, and obtaining a filtered rotating speed signal d;

[0078] S3, in order to reduce the influence of time-varying azimuth angle on controller design and improve system identification and control accuracy, transforming the blade root bending moment signals b1, b2 and b3 obtained in the step S1 to a hub coordinate system by using inverse Kalman coordinate transformation to obtain a pitching moment e and a yawing moment f;

[0079] The pitching moment e and the yawing moment f are respectively:

[0080]

[0081]

[0082] S4, storing the wind wheel rotating speed signal d obtained in the step S2, the pitching moment e and the yawing moment f obtained in the step S3 as an output vector y by using a data storage unit, and storing pitch angle closed loop compensation quantities g1, g2 and g3 as an input vector u;

[0083] The output vector y and the input vector u are respectively:

[0084] y=[d,e,f]

[0085] u = [g1, g2, g3]

[0086] S5, constructing a discrete online data model, i.e. an equivalent identification parameter matrix Φ(k), based on the input vector u and the output vector y obtained in step S4;

[0087] For the convenience of identification algorithm and controller design, the structure form of the data model is assumed as the following linear time-varying model:

[0088]

[0089] wherein k is a time step, and Φ(k) is a 3*3 dynamic time-varying parameter matrix; is an estimated value of the output vector.

[0090] The establishment of the online data model is equivalent to the identification of the parameter matrix Φ(k); the identification objective function comprehensively considers the identification accuracy and the time-varying property of the parameters as performance indexes, and the mathematical expression thereof is:

[0091]

[0092] wherein the vector represents the identification accuracy index; the norm represents the time-varying property index of the to-be-identified parameter Φ(k); h is a normal number, which is a penalty factor of the time-varying property of the to-be-identified parameter matrix Φ(k). The penalty factor limits the change rate of the parameter matrix on one hand, and can avoid the identification error caused by the singular problem of the matrix possibly occurring in the parameter identification process on the other hand; ||·||F Fro is the frobenius norm.

[0093] The extreme value solution of the identification objective function is obtained as follows:

[0094] Φ(k) = [Δy(k) Δu T (k-1) + h Φ(k-1)] [Δu(k-1) Δu T (k-1) + h] -1

[0095] S6, using the parameter matrix Φ(k) obtained in step S5, the input vector u(k-1) at the k-1 time and the output vector y(k) at the k time obtained in step S4, to solve the data-driven control quantity in the fixed hub coordinate system;

[0096] The control accuracy, the variable pitch speed and the variable pitch amplitude performance indexes are comprehensively considered in the control objective function, a quadratic matrix is introduced to plan the weight values of each control error vector, input vector and input change rate vector, and the mathematical expression thereof is:

[0097] Jc (u(k)) = [y d (k+1)-y(k+1)] T I[y d (k+1)-y(k+1)]+Δu T (k)KΔu(k)+u T (k)Lu(k)

[0098] wherein y d (k+1) is the output vector set value at k+1 time; I, K and L are all positive definite quadratic form constant matrix, the first term in the control objective function represents control accuracy, the second term represents variable pitch speed, the third term represents variable pitch amplitude, the second term and the third term can also optimize variable pitch energy consumption.

[0099] Therefore, the superiority of the control objective function is that: ①the control accuracy, variable pitch speed and variable pitch amplitude indicators are comprehensively considered, which avoids the problems of excessive variable pitch speed leading to additional inertia load, excessive variable pitch motor current and serious heating, and excessive variable pitch amplitude leading to system instability; ②the quadratic matrix is introduced to weight plan the control error vector, input vector and input change rate vector, which greatly increases the flexibility of the weight distribution of each indicator.

[0100] The extremum of the control objective function is solved by using the vector derivation rule, and the data-driven control quantity in the fixed hub coordinate system is obtained:

[0101] u(k) = u(k-1) - [Φ(k)IΦ(k) + K + L] -1 Φ(k)I[y(k) - y d (k+1)] - [K + Φ(k)IΦ(k) + L] -1 Lu(k-1)

[0102] Let u(k) = [u1(k), u2(k), u3(k)] T

[0103] The advantages of the data-driven model of step S5 and the data-driven control law of step S6 are that the data-driven model and the data-driven control law are calculated based on the online data of the wind power system and do not contain any physical information related to the structure and parameters of the wind power system. Any uncertainty factor of the wind power system can be included in the time-varying parameter matrix Φ(k), so the accuracy of the data-driven model is higher than that of the simplified mechanism model in principle, and the accuracy of the data-driven control law based on the data-driven model is also higher than that of the controller based on the simplified mechanism model.

[0104] S7, using the Kalman coordinate transformation to drive the data in the fixed hub coordinate system obtained in step S6 to obtain the pitch angle feedback control quantity u(k)=[u1(k), u2(k), u3(k)] T Convert to the rotating blade coordinate system to obtain the pitch angle feedback control quantity m1, m2, m3;

[0105] The pitch angle feedback control quantity m1, m2, m3 is respectively:

[0106] m1=u1(k)+u2(k)cosa1+u3(k)sina1

[0107] m2=u1(k)+u2(k)cosa2+u3(k)sina2

[0108] m3=u1(k)+u2(k)cosa3+u3(k)sina3

[0109] S8, band-pass filter the wind rotor speed signal c obtained in S2, the band-pass frequency of the band-pass filter is 3 times the wind rotor rotation frequency, and the wind rotor speed signal after band-pass filtering is denoted as n, and the pitch angle feedforward compensation quantity o1, o2, o3 is calculated according to the wind rotor speed signal n after band-pass filtering and the azimuth angle signals a1, a2, a3 obtained in S1;

[0110] The pitch angle feedforward compensation quantity o1, o2, o3 is respectively:

[0111] o1=p*n*cosa1

[0112] o2=p*n*cosa2

[0113] o3=p*n*cosa3

[0114] Wherein, p is a positive proportionality constant.

[0115] The periodic load caused by the wind shear effect is related to the azimuth angle, which introduces a load of 1 times the wind rotor rotation frequency in the blade and a load of 3 times the wind rotor rotation frequency in the wind rotor speed. The size of the speed signal n after band-pass filtering can represent the amplitude of the periodic load introduced by the wind shear effect, and the pitch angle feedforward compensation quantity o1, o2, o3 calculated by the speed signal n after band-pass filtering, the proportionality coefficient p and the cosine of the azimuth angle can eliminate the periodic load caused by the wind shear effect to some extent.

[0116] S9, add the pitch angle feedback control quantity m1, m2, m3 obtained in step S7 to the azimuth angle feedforward compensation quantity o1, o2, o3 obtained in step S8, and send the total pitch angle control quantity q1, q2, q3 obtained to the pitch angle actuator.

[0117] q1=m1+o1

[0118] q2 = m2 + o2

[0119] q3 = m3 + o3

[0120] The feedforward compensation amount o1, o2, o3 responds to the load disturbance introduced by the known wind shear, has fast response speed and good stability, but the control precision is insufficient; the pitch angle feedback control amount m1, m2, m3 is calculated through data-driven closed-loop control, and has high control precision, but the control amount is generated after the error occurs, and the response is slow. The present application combines feedforward compensation with data-driven closed-loop control, while ensuring the control precision and dynamic response speed of the variable pitch.

[0121] The pitch angle actuator limits the change range and rate of the total pitch angle set value q1, q2, q3 obtained in step S9, and drives the variable pitch motor to realize blade pitching to the specified angle.

[0122] In another embodiment of the present application, a data-driven independent variable pitch control system is provided, which can be used to implement the above-mentioned data-driven independent variable pitch control method. Specifically, the data-driven independent variable pitch control system comprises an acquisition module, a transformation module, a matrix module, a conversion module and a control module.

[0123] The acquisition module obtains the azimuth angle signal of the wind turbine blade, the blade root bending moment signal and the wind rotor speed signal.

[0124] The transformation module performs low-pass filtering on the wind rotor speed signal obtained by the acquisition module using a first-order inertia link to obtain a filtered wind rotor speed signal; the blade root bending moment signal obtained by the acquisition module is transformed into the hub coordinate system using inverse Kalman coordinate transformation to obtain the pitch moment and the yaw moment; the wind rotor speed signal, the pitch moment and the yaw moment are stored as an output vector, and the pitch angle closed-loop compensation amount is stored as an input vector;

[0125] The matrix module constructs an equivalent identification parameter matrix based on the input vector and the output vector obtained by the transformation module, and solves the data-driven control amount in the fixed hub coordinate system by using the input vector at time k-1 and the output vector at time k;

[0126] The conversion module converts the data-driven control amount in the fixed hub coordinate system obtained by the matrix module into the rotating blade coordinate system using Kalman coordinate transformation to obtain the pitch angle feedback control amount;

[0127] The control module performs band-pass filtering on the wind wheel rotating speed signal obtained by the acquisition module to obtain a band-pass filtered wind wheel rotating speed signal, and calculates a pitch angle feedforward compensation amount in combination with the azimuth angle signal; the total pitch angle control amount obtained by adding the pitch angle feedback control amount obtained by the conversion module and the pitch angle feedforward compensation amount; and the independent variable pitch control is realized according to the total pitch angle control amount.

[0128] In summary, the data-driven independent variable pitch control method and system combine the model-based disturbance feedforward control with the data-driven closed-loop control, which not only ensures the response speed of the variable pitch control to the periodic load caused by the wind shear effect and the tower shadow effect, but also identifies the time-varying relationship of the internal variables of the wind power system online, and the controller designed based on the data model can avoid the control error caused by the simplified mechanism model. The feedback control loop based on data-driven comprehensively considers the mutual influence between the blade root bending moment and the wind wheel rotating speed, and the multivariable closed-loop control. The iterative process of the data-driven control amount comprehensively considers the control accuracy, the variable pitch speed and the variable pitch amplitude performance indicators, avoids the problems of excessive additional inertia load, excessive variable pitch motor current and serious heating caused by excessive variable pitch speed, and avoids the system instability caused by excessive variable pitch amplitude. Meanwhile, the quadratic matrix is introduced to plan the weights of each control error vector, input vector and input change rate vector, which greatly increases the flexibility of the weight distribution of each indicator.

[0129] Those skilled in the art will understand that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0130] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The functions specified in one flow or multiple flows and / or blocks

[0131] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed methods can be implemented on a computer or other programmable data processing apparatus. Figure 1

[0132] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions that are executed on the computer or other programmable apparatus provide steps for implementing the functions specified in the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed methods can be implemented on a computer or other programmable data processing apparatus. Figure 1

[0133] The above merely illustrates the technical idea of the present application, and cannot be used to limit the protection scope of the present application. Any modification made according to the technical idea of the present application, on the basis of the technical scheme, falls within the protection scope of the claims of the present application.​​

Claims

1. A data-driven independent pitch control method, characterized in that, Includes the following steps: S1. Acquire the azimuth angle signal, blade root bending moment signal, and wind turbine speed signal of the wind turbine blades; S2. Use a first-order inertial element to perform low-pass filtering on the wind turbine speed signal obtained in step S1 to obtain the filtered wind turbine speed signal. S3. Use inverse Kalman coordinate transformation to transform the blade root bending moment signal obtained in step S1 to the hub coordinate system to obtain the pitching moment and yaw moment. S4. Store the rotor speed signal obtained in step S2, the pitch moment and yaw moment obtained in step S3 as an output vector, and store the pitch angle closed-loop compensation amount as an input vector. and input vector They are respectively: in, This is the wind turbine speed signal. For pitching moment, For yaw moment, This is the closed-loop compensation amount for the pitch angle; S5. Construct an equivalent identification parameter matrix based on the input and output vectors obtained in step S4. for: in, k For time step, This is the difference between the output vector at the current time step and the output vector at the previous time step. For the first Time-input vector and The transpose of the difference between the output vectors at time step 1. For positive integers, This is the 3*3 dynamic time-varying parameter matrix at time k-1; S6. Using the equivalent identification parameter matrix obtained in step S5, solve the data drive control quantity under the fixed hub coordinate system based on the input vector at time k-1 and the output vector at time k obtained in step S4. S7. Using Kalman coordinate transformation, the data-driven control quantity obtained in step S6 in the fixed hub coordinate system is converted to the rotating blade coordinate system to obtain the pitch angle feedback control quantity. , , They are respectively: in, This is the blade azimuth angle signal. , , They are data-driven control quantities 3D elements; S8. Bandpass filter is applied to the wind turbine speed signal obtained in step S2 to obtain the bandpass filtered wind turbine speed signal. Combined with the azimuth angle signal obtained in step S1, the pitch angle feedforward compensation amount is calculated. S9. The total pitch angle control quantity is obtained by adding the pitch angle feedback control quantity obtained in step S7 to the pitch angle feedforward compensation quantity obtained in step S8. Independent pitch control is then implemented based on the total pitch angle control quantity.

2. The data-driven independent pitch control method according to claim 1, characterized in that, In step S3, the pitching moment e and the yaw moment f are respectively: in, This is the leaf root bending moment signal. This is the blade azimuth angle signal.

3. The data-driven independent pitch control method according to claim 1, characterized in that, In step S6, the data-driven control quantity in the fixed hub coordinate system for: in, for k The output vector setting value at time +1; , and All are positive definite quadratic constant matrices. Let k be the input vector at time k-1. Let k be the output vector at time k. This is the equivalent identification parameter matrix.

4. The data-driven independent pitch control method according to claim 1, characterized in that, In step S8, the pitch angle feedforward compensation amounts o1, o2, and o3 are respectively: Where p is a direct proportionality constant. For the blade azimuth angle information, This is the wind turbine speed signal after bandpass filtering.

5. The data-driven independent pitch control method according to claim 4, characterized in that, The bandpass filter frequency is three times the wind turbine rotation frequency.

6. The data-driven independent pitch control method according to claim 1, characterized in that, In step S9, the total pitch angle control values ​​q1, q2, and q3 are respectively: Where o1, o2, and o3 are the pitch angle feedforward compensation values, respectively. , , These are the pitch angle feedback control variables.

7. A data-driven independent pitch control system, characterized in that, include: The acquisition module acquires the azimuth angle signal, blade root bending moment signal, and rotor speed signal of the wind turbine blades. The transformation module uses a first-order inertial element to perform low-pass filtering on the turbine rotation speed signal obtained from the acquisition module, obtaining the filtered turbine rotation speed signal. It then uses inverse Kalman coordinate transformation to transform the blade root bending moment signal obtained from the acquisition module to the hub coordinate system, obtaining the pitching moment and yaw moment. The turbine rotation speed signal, pitching moment, and yaw moment are stored as output vectors, while the pitch angle closed-loop compensation is stored as an input vector. The output vector... and input vector They are respectively: in, This is the wind turbine speed signal. For pitching moment, For yaw moment, This is the closed-loop compensation amount for the pitch angle; The matrix module constructs an equivalent identification parameter matrix based on the input and output vectors obtained from the transformation module. It then uses the input vector at time k-1 and the output vector at time k to solve for the data-driven control quantity in a fixed hub coordinate system. for: in, k For time step, This is the difference between the output vector at the current time step and the output vector at the previous time step. For the first Time-input vector and The transpose of the difference between the output vectors at time step 1. For positive integers, This is the 3*3 dynamic time-varying parameter matrix at time k-1; The transformation module uses Kalman coordinate transformation to convert the data-driven control quantity obtained from the matrix module in the fixed hub coordinate system into the rotating blade coordinate system, thus obtaining the pitch angle feedback control quantity. , , They are respectively: in, This is the blade azimuth angle signal. , , They are data-driven control quantities 3D elements; The control module performs bandpass filtering on the wind turbine speed signal obtained from the acquisition module to obtain the bandpass-filtered wind turbine speed signal. Combined with the azimuth angle signal, it calculates the pitch angle feedforward compensation amount. The pitch angle feedback control amount obtained from the conversion module is added to the pitch angle feedforward compensation amount to obtain the total pitch angle control amount. Independent pitch control is realized based on the total pitch angle control amount.

Images