WIND TURBINE CONTROL METHOD USING AN ESTIMATION OF THE INCIDENT WIND SPEED

Abstract

The invention is a method for controlling a wind turbine to optimize the energy production in which control was an estimation of the incident wind speed Vw to obtain an optimum control (COM) θsp,Tesp. The estimation of wind speed Vw is achieved by accounting for the dynamics of the system (MOD DYN) from the measurement of rotor speed Ωr the torque imposed on the generator Te and of orientation θ of the turbine blades.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

Reference is made to French Application Serial No. 12/02.601, filed on Oct. 1, 2012 and PCT/FR2013,052040, filed on Sep. 4, 2013, which applications are incorporated herein by reference in their entirety.

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to renewable energy and more particularly to control of wind turbines.

2. Description of the Prior Art

A wind turbine allows the kinetic energy from the wind to be converted into electrical or mechanical energy. For conversion of the wind energy to electrical energy, the wind turbine is made up of the following elements:

• • A tower allowing a rotor to be positioned at a sufficient height to enable motion thereof (necessary for horizontal-axis wind turbines) or allowing the rotor to be positioned at a height enabling it to be driven by a stronger and more regular wind than at ground level. The tower generally houses part of the electric and electronic components (modulator, control, multiplier, generator, etc.).
  • A nacelle mounted at the top of the tower, housing mechanical, pneumatic and some electric and electronic components necessary to operate the machine. The nacelle can rotate to adjust the machine to the correct wind direction;
  • A rotor fastened to the nacelle, comprising blades (generally three) and the nose of the wind turbine. The rotor is driven by wind energy and it is connected by a mechanical shaft, directly or indirectly (via a gearbox and mechanical shaft system), to an electric machine (electric generator) that converts the energy recovered to electrical energy and
  • A transmission having two shafts (mechanical shaft of the rotor and mechanical shaft of the electric machine) connected by a transmission (gearbox).

Since the beginning of the 1990s, there has been renewed interest in wind power, in particular in the European Union where the annual growth rate is about 20%. This growth is attributed to the benefit of carbon-emission-free electricity generation. In order to sustain this growth, the energy yield of wind turbines still has to be improved. The prospect of wind power production increase requires developing effective production tools and advanced control tools in order to improve the performance of the machines. Wind turbines are designed to produce electricity at the lowest possible cost. They are therefore generally built to reach their maximum performance at a wind speed of approximately 15 m/s. It is in fact unnecessary to design wind turbines that maximize their yield at higher wind speeds, which are not common. In the case of wind speeds above 15 m/s, it is necessary to not capture part of the additional energy contained in the wind to avoid damage to the wind turbine. All wind turbines are therefore designed with a power regulation system.

For this power regulation, controllers have been designed for variable-speed wind turbines. The purpose of the controllers is to maximize the electrical power which is recovered, to minimize the rotor speed fluctuations and to minimize the fatigue and the extreme moments of the structure (blades, tower and platform).

Linear controllers have been widely used for power control which control the blade pitch angle (orientation of the blades). These include techniques using PI (proportional integral) and PID (proportional integral derivative) controllers, LQ (linear quadratic) control techniques and strategies based on robust linear controls.

However, the performance of these linear controllers is limited by the highly non-linear characteristics of the wind turbine. First strategies based on non-linear controls have been developed, such a strategy is for example described in the document: Boukhezzar B., Lupu L., Siguerdidjane H., Hand M. “Multivariable Control Strategy for Variable Speed, Variable Pitch Wind Turbines” Renewable Energy, 32 (2007) 1273-1287.

None of these strategies however uses the incident wind speed, which is a fundamental element for the aerodynamic phenomena that govern the wind turbine. To take this component into account, initial work was performed with a measurement of the wind speed. This work shows that the productivity of a wind turbine and the life thereof can be significantly increased through innovative strategies using the wind speed.

This technique unfortunately requires a sensor that is expensive and not very accurate. To take the incident wind speed into account without a sensor, an estimation of this speed can be performed to use this data in the control. Further work has been conducted to this end using a Kalman filter, which is described in the document: Boukhezzar B., Siguerdidjane H., “Nonlinear Control of Variable Speed Wind Turbine Without Wind Speed Measurement” IEEE Control and Decision Conference (2005). This method is not sufficiently accurate because the wind reconstruction is poorly representative. Indeed, according to this method, the wind is not structured, it is considered as white noise, which is not the case experimentally.

SUMMARY OF THE INVENTION

The invention relates to a method for controlling a wind turbine in order to optimize the energy produced with control accounting for an estimation of the incident wind speed. Estimation of the wind speed is achieved by accounting for the dynamics of the system, from the measurement of the rotor speed, the torque imposed on the generator and of the orientation of the turbine blades for an accurate estimation.

The invention relates to a method for controlling a wind turbine including a rotor to which at least one blade is attached, and an electric machine connected to the rotor, wherein a pitch angle θ of the at least one blade and an electrical recovery torque T_e of the electric machine are known. The method comprises carrying out the following stages:

• • a) constructing a dynamic model of the rotor by applying the fundamental principle of dynamics to the rotor wherein the model relates the incident wind speed at the wind turbine V_w to a rotating speed of the rotor Ω_r , to the pitch angle θ of the blade and to the electrical recovery torque T_e;
  • b) measuring the rotating speed of the rotor Ω_r;
  • c) determining the incident wind speed V_w from use of the rotor dynamics model, of the measured rotating speed of the rotor Ω_r, of the pitch angle θ of the blade and of the electrical torque T_e; and
  • d) controlling the pitch angle θ of the blade and/or the electrical recovery torque T_e as a function of the incident wind speed V_w to optimize the production of energy by said wind turbine.

According to the invention, the rotor dynamics model is expressed by the relationship:

$J_r\frac{{\Omega }_r}{t}=T_{\mathrm{aero}}-T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e,$

with J_r being the inertia of the rotor, T_aero being the aerodynamic torque applied to the rotor that depends on the incident wind speed V_w , the rotating speed of the rotor Ω_r and the pitch angle θ, T_l (Ω_r) being friction and load torque on the rotor that depends on the rotating speed of the rotor Ω_r, and N being the transmission ratio between the rotor and the electric machine.

Advantageously, the aerodynamic torque applied to the rotor is expressed by a formula:

$T_{\mathrm{aero}}=0.5\mathrm{\rho \Pi }R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{V_w}\right)V_w^2,$

with R_b being the radius of the rotor, ρ being the air density, c_q being a parameter determined by mapping the rotor as a function of the pitch angle θ and of ratio

$\frac{R_b{\Omega }_r}{V_w}.$

According to an embodiment of the invention, when the incident wind speed V_w is considered as a harmonic perturbation, the incident wind speed V_w can be written with a formula:

$V_w\left(t\right)=\sum _{k=-p}^pc_k{}^{\left({\mathrm{\omega }}_kt\right)}$

with p being the number of harmonics considered, ω_k being the frequency of harmonic k, c_k being the coefficient of harmonic k, and the incident wind speed V_w is determined by determining parameters ω_k and c_k by use of the rotor dynamics model, then the incident wind speed is deduced therefrom.

Preferably, the frequency of a harmonic k is determined using a formula of the type:

${\omega }_k=3\frac{k}{p}.$

Advantageously, coefficients c_k are determined by solving a system of equations of the form:

$\hspace{1em}\{\begin{array}{c}{J}_r\frac{{\Omega }_r}{t}=0.5\mathrm{\rho \Pi }R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}}\right){\left(\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}\right)}^2-T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e-J_rL_{\Omega }\left({\Omega }_r-\Omega \right)\\ \frac{c_k}{t}=-L_k{}^{\left(-{\omega }_kt\right)}\left({\Omega }_r-\Omega \right)\mathrm{for}\mathrm{any}k\in \left[-p,p\right]\end{array}$

with L_Ω being a gain that controls the convergence rate of the estimation of the rotor rotating speed, L_k being a gain that controls the convergence rate of the harmonic decomposition, and Ω representing the measured rotor speed.

Preferably, the gain L_Ω is substantially 1 and the gain L_k is determined with a formula:

$L_k=\frac{10}{1+{\omega }_k^2}.$

Furthermore, the pitch angle θ can be controlled by use of the following stages:

• • i) determining an aerodynamic torque setpoint T_aero^sp an electric machine torque setpoint R_e^sp which allows maximizing the power recovery, from the incident wind speed V_w, measurements of the rotor rotating speed Ω_r and the rotating speed of the electric machine shaft Ω_e;
  • ii) modifying at least one of the setpoint values by subtracting a term proportional to a difference between measured rotor speed Ω_r and measured electric machine speed Ω_e;
  • iii) determining a pitch angle θ^sp for the blade permitting achieving the aerodynamic torque setpoint R_aero^sp; and
  • iv) orienting the blade according to the pitch angle θ^sp.

Preferably, at least one of the setpoint values is modified by carrying out the following stages:

• • (1) determining a torque T_res on the transmission resulting from the aerodynamic torque T_aero^sp and electric machine torque T_e^sp setpoints;
  • (2) determining a resulting torque setpoint T_ressp by subtracting from the resulting torque T_res a term proportional to the difference between measured rotor speed Ω_r and measured electric machine speed Ω_e; and
  • (3) modifying the aerodynamic torque setpoint T_aero^sp by distributing the resulting torque setpoint T_ressp between an aerodynamic torque T_aero^sp and an electric machine torque T_e^sp.

According to an embodiment of the invention, the resulting torque setpoint T_res^sp is written as follows: T_res^sp= T_res−k{dot over (γ)}_tr with k being strictly positive calibration parameters and {dot over (γ)}_tr being speed of a gear transmission torsion, equal to a difference of rotor speed Ω_r and electric machine speed Ω_e brought into the same axis:

${\stackrel{.}{\gamma }}_{\mathrm{tr}}={\Omega }_r-\frac{1}{N}{\Omega }_e,$

where N is a gear ratio between the rotor and the electric machine.

Furthermore, the pitch angle of the blade can be determined by inverting an aerodynamic torque model and using the incident wind speed V_w and measured rotor speed Ω_r.

Advantageously, the proportional term is determined using a transmission dynamics model.

BRIEF DESCRIPTION OF THE DRAWINGS

Other features and advantages of the method according to the invention will be clear from reading the description hereafter of embodiments given by way of non limitative example, with reference to the accompanying figures wherein:

FIG. 1 illustrates the stages of the method according to the invention;

FIG. 2 illustrates the stages of the method according to an embodiment of the invention;

FIG. 3 shows a map of the rotor giving coefficient c_q as a function of pitch angle θ and of ratio

$\frac{R_b{\Omega }_r}{V_w};$

and

FIG. 4 illustrates the wind turbine control stages according to an embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

The invention relates to a method for controlling an onshore or offshore horizontal-axis wind turbine, wherein the pitch angle of the blades and/or the electrical recovery torque of the electric machine are controlled to optimize energy recovery.

Notations

In the description hereafter, the following notations are used:

• • θ is the pitch angle of the blades which is the angle between the blades and a reference such as the ground (horizontal plane, perpendicular to the tower of the wind turbine) and the pitch angle can be controlled by the method according to the invention;
  • T_e is the electrical recovery torque of the electric machine, which can be controlled by the method according to the invention;
  • V_w is the incident wind speed at the blades of the wind turbine which is estimated by the method according to the invention;
  • Ω_r is the rotating speed of the wind turbine rotor which can be measured;
  • Ω_e is the rotating speed of the electric machine shaft which can be measured;
  • J_r is the inertia of the wind turbine rotor which is a known value;
  • J_g is the inertia of the electric machine;
  • T_aero is the aerodynamic torque applied to the rotor which is the rotational force applied to the rotor under the effect of wind;
  • T_l is the friction and load torque on the rotor which can be determined by a second-order polynomial of the rotor rotating speed Ω_r;
  • N is the gear ratio between the rotor and the electric machine;
  • R_b is the rotor radius;
  • ρ is the air density. The density varies as a function of temperature and atmospheric pressure. It is generally around 1.2 kg/m³;
  • c_q is the coefficient given by a map of the rotor as a function of pitch angle θ and ratio

$\frac{R_b{\Omega }_r}{V_w}$

FIG. 3 is an example of a map used for determining the coefficient;

• • c_k is the coefficient of harmonic k of the harmonic decomposition of incident wind speed V_w;
  • ω_k is the frequency of harmonic k of the harmonic decomposition of incident wind speed V_w;
  • p is the number of harmonics considered for the harmonic decomposition of incident wind speed V_w;
  • L_Ω is the gain controlling the convergence rate of the rotor speed estimation;
  • L_k is the gain controlling the convergence rate of the harmonic decomposition of incident wind speed V_w;
  • T_res is the resulting torque on the transmission between the rotor and the electric machine;
  • {dot over (γ)}_tr is the torsion speed of the transmission between the rotor and the electric machine.

These notations, when followed by superscript -^sp, represent the setpoints associated with the quantities considered.

FIG. 1 shows the method according to the invention. The stages of the method according to the invention are as follows:

• • 1) Measurement of rotor rotating speed Ω_r
  • 2) Construction of a rotor dynamics model (MOD DYN)
  • 3) Wind turbine control (COM)

Stage 1)—Measurement of the Rotor Rotating Speed

Rotor rotating speed Ω_r is measured during operation of the wind turbine, notably by a rotation sensor.

Stage 2)—Construction of a Rotor Dynamics Model (MOD DYN)

The actual pitch angle θ of the blades is first determined, as well as electrical recovery torque T_e , by measuring (with a sensor for example) or as a function of the control applied to the actuators to modify these parameters.

A rotor dynamics model is then constructed. A rotor dynamics model is understood to be a model representing the dynamic phenomena applied to the rotor. This model is obtained by applying the fundamental principle of dynamics to the rotor. The model allows relating the incident wind speed V_w to a rotating speed of the rotor Ω_r, to pitch angle θ of the blade and to electrical recovery torque T_e of the electric machine.

This rotor dynamics model is then applied with known data: θ and T_e, and at the measured value Ω_r, which thus allows to determine incident wind speed V_w.

FIG. 2 shows an embodiment of the method according to the invention. For this embodiment, the rotor dynamics model (MOD DYN) is constructed by use of a model of the aerodynamic torque (MOD AERO), of the fundamental dynamics principle (PFD) applied to the rotor and of a wind model (MOD VENT).

Aerodynamic Torque Model (MOD AERO)

According to this embodiment of the invention, aerodynamic torque T_aero is modelled by a model describing the wind power contained in a cylinder, multiplied by a factor describing the fact that a wind turbine only allows recovery of part of this power. Aerodynamic torque T_aero is thus modelled as a function of incident wind speed V_w pitch angle θ and rotor speed Ω_r. Such a model can be expressed under steady state conditions as

$T_{\mathrm{aero}}=0.5\mathrm{\rho \Pi }R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{V_w}\right)V_w^2.$

The parameter c_q can be determined by mapping the rotor. An example of mapping parameter c_q is shown in FIG. 3. This map shows the value of parameter c_q as a function of ratio

$\frac{R_b{\Omega }_r}{V_w}$

for various pitch angles (a curve for each θ). This type of map is well known. Ratio

$\frac{R_b{\Omega }_r}{V_w}$

is denoted by TSR in FIG. 3.

Aerodynamic torque T_aero can therefore be written as a function of quantities related to the wind turbine (ρ,R_b), of the known value (θ) and of the incident wind speed to be estimated (V_w).

Fundamental Dynamics Principle (PFD)

By writing the fundamental dynamics principle applied to the rotor concerning the moments on the axis of rotation thereof, a relation is obtained of the type:

$J_r\frac{{\Omega }_r}{t}=T_{\mathrm{aero}}-T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e.$

In this relation, the aerodynamic torque T_aero determined with the aerodynamic torque model described above is used. Furthermore, the friction and load torque on the rotor T_l can be conventionally determined by a second-order polynomial of rotating speed Ω_r of the rotor.

By combining the two models, a relation between incident wind speed V_w and the known or measured quantities such as pitch angle θ of the blades, electrical recovery torque T_e and rotating speed Ω_r of the rotor can be established.

Incident Wind Modelling (MOD VENT)

The last modelling stage considers the wind as a harmonic perturbation. This means an assumption that the wind can be is written in the form:

$V_w\left(t\right)=\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}$

with p being the number of harmonics considered for the harmonic decomposition of the wind, p=50 can be selected for example.

Moreover, for the frequency of the harmonics,

${\omega }_k=3\frac{k}{p}$

in Hz (with k>0) can be chosen.

By combining the three models presented above, a dynamic model of the form as follows can be obtained:

$\hspace{1em}\{\begin{array}{cc}{J}_r\frac{{\Omega }_r}{t}=0.5\rho \Pi R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}}\right){\left(\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}\right)}^2& -T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e\\ \frac{c_k}{t}=0& \mathrm{for}\mathrm{any}k\in \left[-p,p\right]\end{array}$

From this system of equations, an observer can be constructed allowing determination of coefficients c_k of the harmonic decomposition of the incident wind. The observer can be written in the form:

$\hspace{1em}\{\begin{array}{cc}{J}_r\frac{{\Omega }_r}{t}=0.5\rho \Pi R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}}\right){\left(\sum _{k=-p}^pc_k{}^{\left({\omega }_kt\right)}\right)}^2& -T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e-J_rL_{\Omega }\left({\Omega }_r-\Omega \right)\\ \frac{c_k}{t}=-L_k{}^{\left(-{\omega }_kt\right)}\left({\Omega }_r-\Omega \right)& \mathrm{for}\mathrm{any}k\in \left[-p,p\right]\end{array}$

where Ω represents the measured rotor speed, L_Ω is necessarily a positive gain controlling the convergence rate of the rotor speed estimation. The gain can be selected equal to 1 for example, and {L_k}_k∈[−p,p] is a gain controlling the convergence rate of the harmonic decomposition. These gains must be positive, and can be selected equal to

$\frac{10}{1+{\omega }_k^2}$

for example.

The latter system of equations represents an adaptive type non-linear estimator allowing estimation of coefficients c_k of the harmonic decomposition of the incident wind signal V_w.

This method reconstructs the excitation of wind V_w through coefficients c_k . Reconstructed wind V_w is given by the relation as follows:

$V_w\left(t\right)=\sum _{k=-p}^pc_k{}^{\left({\mathrm{\omega }}_kt\right)}.$

Stage 3)—Wind Turbine Control (COM)

Depending on incident wind speed V_w, the wind turbine is controlled so as to optimize the energy recovered. According to the invention, pitch angle θ of the blades and/or electrical recovery torque T_e can be controlled as a function of incident wind V_w.

According to an embodiment of the invention, pitch angle θ of the blades and/or electrical recovery torque T_e are determined by mapping the wind turbine as a function of incident wind V_w.

Alternatively, according to an embodiment of the invention illustrated in FIG. 4, pitch angle θ of the blades can be controlled by the following stages:

• • 1—Determining the pitch allowing the recovered power to be optimized
    • i—Generating an electrical torque setpoint T_e^sp
    • ii—Generating an aerodynamic torque setpoint T_aero^sp
    • iii—Determining a pitch position θ
  • 2—Determining the torque resulting from the torque setpoints T_esp and T_aero^sp
  • 3—Generating a resulting torque setpoint (T_res^sp) that decreases fatigue and extreme moments of the transmission
  • 4—Distributing the resulting torque setpoint (T_res^sp) between the aerodynamic and electrical torques
  • 5—Determining a pitch position allowing this aerodynamic torque to be achieved
  • 6—Orienting the blades at the determined pitch angle.

1—Determining the Pitch Allowing the Recovered Power to be Optimized

One goal of the method according to the invention is to maximize the energy production of an onshore or offshore horizontal-axis wind turbine (blade perpendicular to the wind) while limiting extreme moments and fatigue of the mechanical structure.

To maximize the energy production of a wind turbine, the pitch angle θ of the blades allowing maximizing the recovered power P_aero as a function of incident wind speed V_w determined in the rotor dynamics model construction stage is sought.

According to an embodiment, a model of the recoverable power is used to define this angle. This power P_aero can be written as follows:

P _aero =T _aero*Ω_r

Angle θ allowing P_aero to be maximized is thus sought. The following stages are therefore carried out:

• • i—Generating an electric machine torque setpoint T_e^sp
  • ii—Generating an aerodynamic torque setpoint T_aero^sp
  • iii—Determining a pitch position θ

i—Generating an Electric Machine Torque Setpoint T_e^sp

An electric machine torque setpoint T_e^sp is first determined. This setpoint value is obtained through mapping as a function of the speed of the electric machine.

According to the invention, aerodynamic torque T_aero is modelled by the aerodynamic model as described in the dynamic model construction part.

$T_{\mathrm{aero}}=0.5\mathrm{\rho \Pi }R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{V_w}\right)V_w^2$

Thus, to determine the torque setpoint for the electric machine as a function of the speed of the electric machine, the aerodynamic power recovered for each wind speed is optimized.

$T_e^{\mathrm{sp}}=\arg \left({\max }_{\theta ,V_w}\frac{0.5}{N}\rho \Pi R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_g}{{\mathrm{NV}}_w}\right)V_w^2\right)$

This allows having torque setpoint T_e that depends on the rotating speed of the electric machine shaft: T_e^sp=f(Ω_e).

However, in relation to this reference curve, two limitations are applied:

• • a zero torque for low electric machine speeds so that the wind turbine speed can be increased
  • a maximum torque to limit the power of the electric machine.

Thus, there are three regions in the curve T_e^sp=f(Ω_e):

• • Region 1: zero torque
  • Region 2: optimum torque
  • Region 3: torque limited by the maximum power.

ii—Generating an Aerodynamic Torque Setpoint T_aero^sp

The purpose is to generate an aerodynamic torque setpoint T_aero^sp allowing achieving the rotor rotating speed setpoint Ω_r^sp. The rotor dynamics model is therefore used.

$J_r\frac{{\Omega }_r}{t}=T_{\mathrm{aero}}-T_l\left({\Omega }_r\right)-{\mathrm{NT}}_e\left({\Omega }_e\right)$

Thus, the control strategy used is a dynamic control strategy that anticipates the setpoint variation and corrects with two terms which are a proportional term and an integral term. The strategy is written with a relation of the form:

$T_{\mathrm{aero}}^{\mathrm{sp}}=T_l\left({\Omega }_r\right)+{\mathrm{NT}}_e\left({\Omega }_r\right)+J_r\frac{{\Omega }_r^{\mathrm{sp}}}{t}-k_p\left({\Omega }_r-{\Omega }_r^{\mathrm{sp}}\right)-k_i\int \left({\Omega }_r-{\Omega }_r^{\mathrm{sp}}\right)$

where kp and ki are two real parameters to be calibrated so as to guarantee convergence of the speed to the setpoint thereof.

iii—Determining a pitch position θ

From this aerodynamic torque setpoint T_aero^sp, a pitch angle θ is determined for the blades to satisfy this aerodynamic torque request T_aero^sp. The aerodynamic torque model is therefore used with the incident wind speed V_w determined in the rotor dynamics model construction stage, the measured rotor speed Ω_r^sp and the torque setpoint T_aero^sp. By inverting the model (using a Newton algorithm for example), a pitch setpoint θ is obtained:

$\stackrel{\_}{\theta }=\arg \left({\min }_{\theta }{\left(T_{\mathrm{aero}}^{\mathrm{sp}}-0.5\rho \Pi R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{V_w}\right)V_w^2\right)}^2\right)$

Thus, with this control law, the convergence to the reference rotor speed allowing the recovered power to be maximized is guaranteed.

2—Determining the Torque Resulting from the Torque Setpoints T_e^sp and T_aero^sp

From setpoints T_e^sp and T_aero^sp, torque T_res resulting from these two torques and relating to the transmission is determined. Therefore, this torque is modeled by us of the formula as follows:

${\stackrel{\_}{T}}_{\mathrm{res}}=\frac{J_g}{J_r+J_g}\left(T_{\mathrm{aero}}^{\mathrm{sp}}-T_l\right)+\frac{J_r}{J_r+J_g}{\mathrm{NT}}_e^{\mathrm{sp}}$

where J_r and J_g are the inertias of the rotor and of the electric machine respectively.

3—Generating a Resulting Torque Setpoint (T_res^sp) that Decreases the Fatigue and the Moments of the Transmission

It is desired to modify this resulting torque T_res in order to minimize the impact on the transmission and thus to increase the life thereof. Therefore, it is desirable to decrease the torsion speed variations of the transmission. Thus, an attempt is made to compensate for the torque with terms proportional to the difference between the speed of the rotor and of the electric machine. The dynamics of the mechanical structure (transmission dynamics) can be written in form of two coupled second-order systems.

$\{\begin{array}{c}\frac{J_rJ_g}{J_r+J_g}{\ddot{\gamma }}_{\mathrm{tr}}=-c_d{\gamma }_{\mathrm{tr}}-k_d{\stackrel{.}{\gamma }}_{\mathrm{tr}}+\frac{J_g}{J_r+J_g}\left(T_{\mathrm{aero}}-T_l\right)+\frac{J_r}{J_r+J_g}{\mathrm{NT}}_e\\ J_g{\stackrel{.}{\Omega }}_e=c_d{\gamma }_{\mathrm{tr}}+k_d{\stackrel{.}{\gamma }}_{\mathrm{tr}}+N_{\mathrm{gb}}T_e\end{array}\hspace{1em}$

where:

• • γ_tr, {dot over (γ)}_tr and {umlaut over (γ)}_tr are the angle, the speed and the acceleration of the torsion of the shaft respectively. It should be noted that the transmission torsion speed is the difference between the speed of the rotor and the generator brought into the same axis, that is

${\stackrel{.}{\gamma }}_{\mathrm{tr}}={\Omega }_r-\frac{1}{N}{\Omega }_e$

• • k_d is the structural transmission damping
  • c_d is the transmission stiffness
  • N_gb is the gearbox ratio, that is the ratio of the generator speed to the rotor speed.

Thus, the control strategy is designed to generate a resulting torque different from T_res to minimize the fatigue and the extreme moments of the transmission. Accordingly, the relationship is obtained:

T _res ^sp =T _res −k{dot over (γ)} _tr

with k being strictly positive calibration parameters. These parameters can be determined experimentally. All the parameters k can be considered equal to 1 for example.

4—Distributing the Resulting Torque Setpoint (T_res^sp) Between the Aerodynamic and Electrical Torques

This resulting torque setpoint T_res^sp is then distributed between aerodynamic torque T_aero and torque T_e of the electric machine. Distribution is achieved according to the operating zones. In a zone 2, where the aerodynamic torque is limiting, a torque reserve exists. In this case, the torque modification influences the torque of the electric machine and not the aerodynamic torque.

Thus, in this case, the relationship:

$\{\begin{array}{c}{T}_{\mathrm{aero}}^{\mathrm{strat}}=T_{\mathrm{aero}}^{\mathrm{sp}}\\ T_e^{\mathrm{strat}}=T_e^{\mathrm{sp}}-k\frac{J_r+J_g}{{\mathrm{NJ}}_r}{\stackrel{.}{\gamma }}_{\mathrm{tr}}\end{array}\hspace{1em}$

Similarly, in a zone 3 where the torque of the electric machine is limiting, the torque modification influences the aerodynamic torque. Therefore, the relationship is obtained:

$\{\begin{array}{c}{T}_{\mathrm{aero}}^{\mathrm{strat}}=T_{\mathrm{aero}}^{\mathrm{sp}}-k\frac{J_r+J_g}{J_g}{\stackrel{.}{\gamma }}_{\mathrm{tr}}\\ T_e^{\mathrm{strat}}=T_e^{\mathrm{sp}}\end{array}\hspace{1em}$

5—Determining a Pitch Position Allowing this Aerodynamic Torque to be Achieved

From this aerodynamic torque setpoint T_aero^strat, a pitch angle θ for the blades to satisfy this aerodynamic torque request T_aero^strat. Therefore, the aerodynamic torque model with the incident wind speed V_w determined in the rotor dynamics model construction stage, the measured rotor speed Ω_r^sp and the torque setpoint T_aero^strat is used. By inverting the model (using a Newton algorithm for example), a pitch setpoint θ is obtained:

$\stackrel{\_}{\theta }=\arg \left({\min }_{\theta }{\left(T_{\mathrm{aero}}^{\mathrm{strat}}-0.5\rho \Pi R_b^3c_q\left(\theta ,\frac{R_b{\Omega }_r}{V_w}\right)V_w^2\right)}^2\right)$

Thus, with this control law, the convergence to the reference rotor speed allowing the recovered power to be maximized, while minimizing the mechanical impact (fatigue and extreme moment) on the transmission is guaranteed.

6—Orienting the Blades at the Determined Pitch Angle

To optimize the electric power recovered by the wind turbine, the blades are oriented at the pitch angle calculated in the previous stage.

Claims

1-12. (canceled)

13. A method for controlling a wind turbine, comprising a rotor to which at least one blade is attached and an electric machine connected to the rotor, wherein a pitch angle of the at least one blade and an electrical recovery torque of the electric machine are known, comprising: a) constructing a model of dynamics of the rotor using a fundamental principle of dynamics to describe the rotor, the model relating incident wind speed at the wind turbine to a rotating speed of the rotor, to the pitch angle of the blade and to the electrical recovery torque;b) measuring the rotating speed of said rotor;c) determining an incident wind speed by using the model of dynamics of the rotor, of the measured rotating speed of the rotor, of the pitch angle of the at least one blade and of the electrical torque; andd) controlling the pitch angle of the blade and/or the electrical recovery torque as a function of the incident wind speed to optimize production of energy by the wind turbine.

14. A method as claimed in claim 13, wherein the model of dynamics of the rotor is expressed as:

15. A method as claimed in claim 14, wherein the aerodynamic torque applied to the rotor is expressed by the relationship:

16. A method as claimed in claim 13 wherein, the incident wind speed Vw is considered to be a harmonic perturbation, and is expressed by the relationship:

17. A method as claimed in claim 14 wherein, the incident wind speed Vw is considered to be a harmonic perturbation, and is expressed by the relationship:

18. A method as claimed in claim 15 wherein, the incident wind speed Vw is considered to be a harmonic perturbation, and is expressed by the relationship:

19. A method as claimed in claim 16, wherein the frequency of harmonic k is determined by a relationship:

20. A method as claimed in claim 17, wherein the frequency of harmonic k is determined by a relationship:

21. A method as claimed in claim 18, wherein the frequency of harmonic k is determined by a relationship:

22. A method as claimed in claim 16, wherein the coefficient ck is determined by solving a system of equations expressed as:

23. A method as claimed in claim 17, wherein the coefficient ck is determined by solving a system of equations expressed as:

24. A method as claimed in claim 18, wherein the coefficient ck is determined by solving a system of equations expressed as:

25. A method as claimed in claim 19, wherein the coefficient ck is determined by solving a system of equations expressed as:

26. A method as claimed in claim 20, wherein the coefficient ck is determined by solving a system of equations expressed as:

27. A method as claimed in claim 21, wherein the coefficient ck is determined by solving a system of equations expressed as:

28. A method as claimed in claim 22, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

29. A method as claimed in claim 23, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

30. A method as claimed in claim 24, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

31. A method as claimed in claim 25, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

32. A method as claimed in claim 26, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

33. A method as claimed in claim 27, wherein the gain LΩ is substantially 1 and the gain Lk is determined by a relationship:

34. A method as claimed in claim 13, wherein the pitch angle θ is controlled by a process comprising: i) determining an aerodynamic torque setpoint and an electric machine torque setpoint allowing maximizing the power recovery, from the incident wind speed, measurements of the rotor rotating speed and the rotating speed of the electric machine shaft;ii) modifying at least one of the setpoint values by subtracting a term proportional to a difference between measured rotor speed and measured electric machine speed;iii) determining a pitch angle for the at least one blade allowing achieving the aerodynamic torque setpoint; andiv) orienting the blade according to the pitch angle.

35. A method as claimed in claim 34, wherein at least one of the setpoint values is modified by steps comprising: (1) determining a torque on the transmission resulting from the aerodynamic torque and electric machine torque setpoints;(2) determining a resulting torque setpoint by subtracting from the resulting torque a term proportional to a difference between measured rotor speed and measured electric machine speed; and(3) modifying the aerodynamic torque setpoint by distributing the resulting torque setpoint between an aerodynamic torque and an electric machine torque.

36. A method as claimed in claim 35, wherein the resulting torque setpoint is expressed as: Tressp= Tres−k{dot over (γ)}tr with k being strictly positive calibration parameters and {dot over (γ)}tr being a speed of a gear transmission torsion, equal to a difference of rotor speed Ωr and electric machine speed Ωe brought into a same axis:

37. A method as claimed in claim 34, wherein the pitch angle of the at least one blade is determined by inverting an aerodynamic torque model and using the incident wind speed and measured rotor speed.

38. A method as claimed in claim 34, wherein a proportional term is determined using a model of transmission dynamics.