DE102018007997A1 - Method and system for operating a wind turbine - Google Patents

Abstract

The invention relates to a method for operating a wind power plant which has at least two rotor blades, comprising the following steps: detecting sensor signals because they are suitable for characterizing blade loads on the rotor blades; determining an individual blade load, in particular an impact bending moment, for each rotor blade on the Basis of the sensor signals recorded on the respective rotor blade and a sensor-specific adjustment function of the respective rotor blade; control of the wind power installation, in particular the setting angle of the rotor blades, on the basis of the determined individual blade loads; at least one of the sensor-individual adjustment functions of the rotor blades being adapted such that a Difference between the individual blade loads of different rotor blades is at least reduced.

Description

The invention relates to a method for operating a wind energy plant which has at least two rotor blades, sensor signals which are suitable for characterizing a blade load on the at least two rotor blades being detected, a blade load being determined on the basis of the detected sensor signals and a calibration function, and the wind turbine, in particular the setting angle of the at least two rotor blades, is controlled on the basis of the blade load determined.

In the case of control strategies for minimizing loads on a wind energy installation, in particular a control of an adjustment angle of the rotor blades is used. In this case, the rotor blades are preferably turned (pitched) individually during the revolution, so that a total mechanical load that is introduced via a rotor hub, a rotor shaft and a nacelle onto a tower of the wind energy installation can be minimized. In this case, we speak of an individual pitch control.

It is particularly advantageous here if an individual blade load, in particular a blade bending moment, can be provided for control purposes for each rotor blade. In this case, we speak of a rotor blade feedback control (blade feedback control).

In order to carry out load measurements, strain sensors, for example fiber Bragg sensors or strain gauges, are usually used as sensors, which are connected in such a way that only bending strains, but not normal forces due to temperature expansion or centrifugal forces, are taken into account.

In general, sensors for measuring blade loads cannot be attached exactly to the place where, according to theoretical considerations, they should be attached. Furthermore, sensors can change their properties over time, so that calibration of the sensors is necessary.

The calibration of the sensors is usually carried out statically against the gravitational bending moment from the known mass and the known center of gravity of the distance of the rotor blade from the measuring point with the rotor blade in a horizontal position. To determine a possible offset of the sensors for measuring the blade bending moments, the rotor blade is preferably placed vertically. The impact bending moment of the rotor blade, which is essentially perpendicular to a reference plane of the rotor blade, or the swivel bending moment, which is essentially parallel to a reference plane of the rotor blade, can be addressed by rotating the blade setting angle by 90 °.

When operating the wind turbine, imbalances occur on the rotor. The unbalances result on the one hand from unequal mass distributions in the rotor, in particular on the rotor blades of the wind turbine, for example due to manufacturing-related or wear-related unequal masses of the rotor blades, fluid deposits on the rotor blades, for example due to clogged condensate drain openings or ice deposits. On the other hand, aerodynamic imbalances occur, that is to say different blade loads on the different rotor blades, which are caused by the wind flow on different rotor blades at the same time. These can occur, for example, due to different adjustment angles of the rotor blades due to production, differences in the respective blade twist, damage to flow elements or even wear. Furthermore, there are different wind conditions, both with regard to wind speed and with respect to wind direction, at different positions which the rotor blades pass during the revolution. One reason for this is the altitude wind gradient, because the wind speed generally increases with altitude.

The unbalance and the asymmetries in the flow against the individual rotor blades both lead to rotational frequency load components ( 1P ), which occur with the simple rotational frequency of the rotor and arise, for example, from a mass imbalance of the rotor or structural differences between the rotor blades, as well as to blade frequency load components ( 3P ), which occur at the speed of the rotor multiplied by the number of rotor blades and arise, for example, due to asymmetry in the flow against the rotor. Due to the control of the wind turbine by means of a rotor blade feedback control, the controller tries to compensate, in particular the 1p load components, by cyclical, rotational frequency adjustment of the individual setting angles of the rotor blades, which causes considerable, permanent blade adjustment activity.

It is an object of the invention to improve control of a wind turbine. In particular, it is an object of the invention to control a wind turbine with little wear, in particular wear-optimized.

This task is solved by the teaching of the independent claims. Beneficial Refinements are claimed in the subclaims.

A first aspect of the invention relates to a method for operating a wind power plant which has at least two rotor blades, comprising the following steps: detecting sensor signals because they are suitable for characterizing blade loads on the rotor blades;
Determining an individual blade load, in particular an impact bending moment, for each rotor blade on the basis of the sensor signals detected on the respective rotor blade and a sensor-specific adaptation function of the respective rotor blade;
Controlling, in particular rules, of the wind power installation, in particular the setting angle of the rotor blades, on the basis of the determined individual blade loads;
wherein at least one of the sensor-specific adjustment functions of the rotor blades is adapted in such a way that a difference between the individual blade loads of different rotor blades is at least reduced.

• eine Schnittstelle, eingerichtet zum Erfassen von Sensorsignalen, die geeignet sind, Blattbelastungen an den Rotorblättern zu charakterisieren;
• Auswertungsmittel, eingerichtet zum Ermitteln einer individuellen Blattbelastung, insbesondere eines Schlagbiegemoments, für jedes Rotorblatt auf der Grundlage der an den jeweiligen Rotorblatt erfassten Sensorsignale und einer sensorindividuellen Angleichfunktion des jeweiligen Rotorblatts und zum Anpassen wenigstens einer der sensorindividuellen Angleichfunktionen der Rotorblätter in der Weise, dass ein Unterschied zwischen den individuellen Blattbelastungen verschiedener Rotorblätter wenigstens verringert wird;
• Steuermittel der Windenergieanlage, eingerichtet zum Steuern der Windenergieanlage, insbesondere des Einstellwinkels der Rotorblätter, auf der Grundlage der ermittelten individuellen Blattbelastungen.

Accordingly, a second aspect of the invention relates to a system for operating a wind power plant which has at least two rotor blades, comprising:

• an interface set up for detecting sensor signals which are suitable for characterizing blade loads on the rotor blades;
• Evaluation means set up to determine an individual blade load, in particular an impact bending moment, for each rotor blade on the basis of the sensor signals recorded on the respective rotor blade and a sensor-specific adjustment function of the respective rotor blade and for adjusting at least one of the sensor-individual adjustment functions of the rotor blades in such a way that a difference is at least reduced between the individual blade loads of different rotor blades;
• Control means of the wind turbine, set up to control the wind turbine, in particular the setting angle of the rotor blades, on the basis of the determined individual blade loads.

Further aspects of the invention relate to a corresponding computer program, a computer-readable medium and a wind turbine.

The features and advantages explained below in relation to the first aspect of the invention apply accordingly to the further aspects of the invention and vice versa.

A setting angle in the sense of the invention is preferably an angle measured about a longitudinal axis of the rotor blade between a reference plane of a rotor blade and a rotor plane swept by the longitudinal axis of the rotor when the rotor rotates, which plane is preferably perpendicular to the rotor axis or, in the case of rotors with a cone angle, represents a conical surface area. The setting angle is preferably defined as 0 ° when the rotor blade is in the operating position, i. H. in operation with an optimal high-speed number, which delivers the maximum power to the rotor shaft. As an alternative or in addition, the setting angle can also be defined as an angle between a rotor blade chord, which preferably extends at least essentially from a rotor blade leading edge to a rotor blade trailing edge, on a predefined profile cut of the rotor blade and the above-mentioned rotor plane or conical outer surface.

An impact bending moment in the sense of the invention preferably occurs perpendicular to the reference plane of the rotor blade. This is preferably perpendicular to the rotor plane when the setting angle of the rotor blade is 0 °.

Controlling a wind power plant in the sense of the present invention is preferably specifying setpoints of operating parameters of the wind power plant when the rotor is rotating in spin mode, in which the rotor blades are preferably in a flag position, or in production mode. The control is preferably a regulation.
A matching function in the sense of the invention is preferably an assignment rule between sensor signals or blade loads as input variables and a blade load adjusted as an output variable between the rotor blades.

A calibration function in the sense of the invention is preferably an assignment rule between sensor signals or blade loads as input variables and an adjusted blade load as output variable.

A gain in the sense of the invention is preferably a proportional part of an adjustment function and / or a calibration function.

A means in the sense of the present invention can be designed in terms of hardware and / or software, in particular one that is data or signal-connected, preferably digital, processing, in particular microprocessor unit (CPU) and / or a, preferably connected to a memory and / or bus system or have several programs or program modules. The CPU can be designed to process commands that are implemented as a program stored in a memory system, to acquire input signals from a data bus and / or to output signals to deliver a data bus. A storage system can have one or more, in particular different, storage media, in particular optical, magnetic, solid-state and / or other non-volatile media. The program can be designed in such a way that it embodies or is capable of executing the methods described here, so that the CPU can execute the steps of such methods and thus in particular can control and / or monitor a wind energy installation.

A time constant in the sense of the invention preferably characterizes a characteristic time period or decay time. The time constant can preferably also be specified as a time period.

Sensor-specific in the sense of the invention preferably means individually for each sensor device and / or individually for each rotor blade. Each rotor blade usually has a sensor device.

The invention is based in particular on the approach of at least reducing periodic, in particular rotational frequency, changes in a setting parameter which is used to control the wind energy installation.

According to the invention, this is achieved in that the determined individual blade loads, which are used to control the wind energy installation, in particular the, preferably individual, setting angle of the rotor blades, are matched to one another. The individual blade load is preferably an impact bending moment of each rotor blade. The adjustment of the values of the individual blade load is accomplished according to the invention by an adjustment function which is taken into account when determining the individual blade load on the basis of the detected sensor signals.

By adjusting the individual blade loads of different rotor blades to each other, the circulating frequency behavior of the determined individual blade load can be reduced. The circulating frequency behavior of the control parameter or the control parameters of the wind power plant, into which the individual blade load is an input parameter, is correspondingly reduced. In this way, the periodic variation of the control parameter during an orbital cycle is reduced; ideally, static periodic variations, which occur in particular in orbital frequency, are even completely eliminated by adapting the individual blade loads to one another. Correspondingly, the adjustment according to the invention of the sensor-specific leaf loads can also be referred to as homogenization of the individual leaf loads.

The invention allows loads on the adjustment system of the setting angle of the rotor blades, which are caused by control interventions due to rotational frequency load components or rotational frequency measurement errors in the load components, to be reduced.
In addition, the influence of measurement errors on the control of the wind turbine can be limited with the system according to the invention. For example, the deviation of a leaf load, which is caused by a sensor drift or a sensor that is slowly becoming detached, would at least partially be compensated for by reducing the difference between the individual leaf loads.

In a further advantageous embodiment of the method according to the invention, the time constant A chosen such that the difference between the individual blade loads over a predefined period of time, over a predefined number of determined blade loads or over a predefined number of rotor revolutions decreases to a value smaller and / or equal to a limit value ε, the limit value ε being in particular zero . As a result, smooth control of the wind energy installation is achieved. The limit value ε is preferably defined by an absolute value or a quotient. The limit value ε is more preferably approximately 63%.

The time constant A is advantageously chosen so that the differences between the loads on the individual blades resulting from the dynamic wind turbulence are retained in order to use them as the basis for controlling the individual blade pitch angles, the static differences in the loads on the individual blades, which lead to periodic 1p Effects, but are compensated for by the adjustment functions.

This advantageous embodiment thus has the effect of filtering out the quasi-static circulating-frequency effects in order not to place unnecessary strain on the blade adjustment systems, but to allow the dynamic circulating-frequency effects, since they are the basis for the load-reducing single-sheet adjustment.

In an advantageous embodiment of the method according to the invention, a predefined number of values of the determined individual blade loads are adjusted on the basis of a predefined time constant in order to adapt the adjustment function A , which preferably defines a predefined period of time, whereby an older value, in particular an oldest value, is no longer taken into account as soon as a younger value, in particular a youngest value, is taken into account. Younger values are the most recently determined values. Of the the most recent value is therefore the last determined value.

Older values of the leaf loads are taken into account in order to reduce interference effects by deviating the measured values, so-called single outliers. A permanent deviation of the blade loads determined from the sensor signals, which makes an adjustment necessary, is taken into account by the gradually increasing influence of the younger values.

In a further advantageous embodiment of the method according to the invention, younger values of the individual blade loads are weighted higher than older values of the individual blade loads, preferably on the basis of a predefined time constant, in order to adapt the adjustment function A , which preferably defines a predefined time period.

The higher weighting of younger values of the individual blade loads of the rotor blade compared to older values ensures that the adaptation responds adequately to the latest changes to the sensors or the rotor blades, which necessitates the adjustment to be made.

In a further advantageous embodiment of the method according to the invention, the at least one of the sensor-specific adaptation functions is continuously adapted, in particular before, during or after a respective control process of a multiplicity of control processes. This has the advantage that a correction of circulating-frequency effects can be continued continuously and new circulating-frequency effects can be taken into account from the beginning by circulating-frequency load components.

In a further advantageous embodiment, the method according to the invention further comprises the step of determining at least one adjustment parameter of at least one of the sensor-specific adjustment functions on the basis of the difference between the individual blade loads on the rotor blades. Such an adjustment parameter, which is preferably in the form of a factor, in particular a gain, enables simple adjustment of the individual sheet load.
Accordingly, in an advantageous embodiment of the system according to the invention, the evaluation means is set up to determine at least one adjustment parameter of at least one of the sensor-specific adjustment functions on the basis of the difference between the individual blade loads on the rotor blades.

In a further advantageous embodiment of the method according to the invention, when adapting the at least one of the sensor-specific adaptation functions, values of the determined blade load are weighted in such a way that new values cause a change in the determined individual blade load of a rotor blade which does not exceed a predetermined limit value ζ. This also enables a smoothed control behavior of the wind power installation to be achieved. The limit value ζ is preferably defined by an absolute value or a quotient.

In a further advantageous embodiment of the method according to the invention, the weighting is carried out in such a way that a change in the determined individual blade load on a rotor blade beyond the predetermined limit value erfordert requires a large number of values, in particular at least one third of the values taken into account for adaptation. This also results in a particularly smooth control behavior of the wind energy installation. In particular, jumps in the specification of the setpoint for the setting parameters of the wind turbine can be avoided in this way. Such an adjustment parameter is, for example, an individual adjustment angle for a rotor blade.

In a further advantageous embodiment, a delay function, in particular a function of a PT1 element, is mapped when the values are taken into account.

A PT1 element in the sense of the invention is preferably a transfer function with proportional transfer behavior with first-order delay. The time response of a PT1 element with respect to an input value over time is preferably an exponential function.

K gibt hierbei den anfänglichen Wert an, T ist eine vordefinierte Zeitkonstante des PT1-Glieds. Der zuletzt berücksichtigte Wert, d.h. der jüngste Wert, liegt bei t=0. Die Wichtung der Werte nimmt ab, je älter die Werte sind.A weighting A The adjustment processes are initially K and gradually decrease over time, the longer the adjustment processes are in the past. This course can be described mathematically by the following relationship: $$\text{A}\left(\text{t}\right)=\text{K}·{\text{e}}^{-\text{t}/\text{T}}$$

K specifies the initial value, T is a predefined time constant of the PT1 element. The last value taken into account, ie the most recent value, is t = 0. The weighting of the values decreases the older the values are.

The use of a delay function brings about a particularly advantageous smoothing of the control behavior of the wind energy installation.

In a further advantageous embodiment of the method according to the invention, the at least one of the sensor-specific adaptation functions is only adapted if predefined criteria, in particular with regard to the operating conditions and / or an operating state of the wind power installation, are met when the sensor signals are detected. By taking predefined criteria into account when adapting the individual blade load, it can be ensured that only those circulation-frequency effects of the control of the wind energy installation that are relevant for the working area of the wind energy installation are suppressed. The work area can be limited as required. In this way, it can in particular be prevented that an adaptation outside the operating area counteracts an adaptation in the operating area.

In a further advantageous embodiment of the method according to the invention, when adapting the at least one of the sensor-specific adaptation functions, only those values are taken into account in the individual blade load in which, when the sensor signals are recorded, the setting angle of the rotor blades essentially corresponds to the performance-optimized operating position with an optimal high-speed number, especially if a collective setting angle the rotor blades are below a predefined setting angle limit, in particular less than about 4 °.

Further preferably, when adapting the at least one of the sensor-specific adaptation functions, only those values of the individual blade load are taken into account at which the rotor speed when the sensor signals are detected is above a predefined speed limit value, in particular above approximately 70% of the nominal speed.

Further preferably, when adapting the at least one of the sensor-specific adaptation functions, only those values of the individual blade load are taken into account at which the power when the sensor signals are detected is above a predefined power limit, in particular above about 40% of the nominal power.

The criteria mentioned have the effect that essentially values in the upper load range of the wind power installation are taken into account for determining the adjustment parameter. This ensures a stable and convergent alignment process.

• Abgleichen des Werts des Angleichparameters mit jeweils einem vordefinierten Grenzwert η, insbesondere 0,85 < η < 1,15, bevorzugt 0,92 < η < 1,08, besonders bevorzugt 0,94 < η < 1,06; und
• Ausgeben einer Statusmitteilung, insbesondere einer Warnmeldung oder Fehlermeldung, wenn der Angleichparameter den jeweils vordefinierten Grenzwert η erreicht und/oder über- oder unterschreitet.

In a further advantageous embodiment, the method according to the invention also has the following work steps:

• Comparing the value of the adjustment parameter with a predefined limit value η, in particular 0.85 <η <1.15, preferably 0.92 <η <1.08, particularly preferably 0.94 <η <1.06; and
• Output of a status message, in particular a warning message or error message, if the adjustment parameter reaches and / or exceeds or falls below the respectively predefined limit value η.

Accordingly, in an advantageous embodiment of the system according to the invention, the evaluation means is set up to value the adjustment parameter with a predefined limit value η, in particular 0.85 <η <1.15, preferably 0.92 <η <1.08, particularly preferably 0, 94 <η <1.06, and issue a status message, in particular a warning or error message, when the adjustment parameter reaches and / or exceeds or falls below the respectively predefined limit value η.

If there is a divergent behavior when adjusting the individual blade load, this is, for example, a sign that one of the rotor blades has damage which is increasing and / or spreading, or that one of the sensors or sensor devices on the rotor blades is defective, or also that an adjustment system of the wind turbine is defective. In these cases it is advantageous if a status message is issued. The operation of the wind turbine is preferably discontinued at the same time. In this case, the wind turbine must be checked to ensure that it continues to function properly.

In a further advantageous embodiment of the method according to the invention, the control comprises determining an individual setpoint angle setpoint for each rotor blade and the setpoint angle of each rotor blade is set on the basis of this individual setpoint angle setpoint. Controlling a wind turbine with individual setting angles is advantageous since special features of the individual rotor blades can be taken into account in the control. In particular, the sensor-specific blade loads are used as part of a rotor blade feedback control.

In a further advantageous embodiment of the method according to the invention, all sensor-specific adaptation functions are adapted, wherein an adaptation function for determining the blade load per rotor blade is taken into account. In this case, the method according to the invention preferably also has the step of determining an average leaf load on the basis of the individual leaf load of all Rotor blades, the sensor-specific adjustment functions being adjusted on the basis of the average blade load. The adjustment parameters of the sensor-specific adjustment functions are preferably determined on the basis of a difference between the individual blade loads on the rotor blades and the average blade load. In this case, the average blade load serves as a reference point for adjusting the individual blade loads. Such a common reference point is advantageous, in particular if there are more than two rotor blades.

Accordingly, in an advantageous embodiment of the system according to the invention, the evaluation means is set up to determine an average blade load on the basis of the individual blade load of all rotor blades, the sensor-specific adjustment functions being adapted on the basis of the average blade load.

In an advantageous embodiment of the method according to the invention, the individual sheet load is additionally determined on the basis of a calibration function, the calibration function being adapted by means of calibration processes based on the determined sheet load, with recent calibration processes depending on a predefined time constant B , which preferably characterizes a predefined period of time, are weighted higher than older calibration processes, the time constant B with respect to the calibration function is preferably less than the time constant A in terms of the matching functions, in particular 1.5 to 2.5 times smaller. The different time constants ensure that the inventive adjustment of the individual sheet load is faster than a recalibration of a sensor device or a sensor. If the calibration is complete and there is no fundamental fault, such as B . a defective sensor device or rotor blade properties deviating from the design goal, the adjustment according to the invention of differences in the individual blade load and the adjustment parameters of each sensor-specific adjustment function should be superfluous 1 be.

• 1 eine Windenergieanlage mit einem System zum Betreiben der Windenergieanlage; und
• 2 ein Verfahren zum Betreiben der Windenergieanlage nach 1.

Further advantages and features result from the exemplary embodiments which are described with reference to the figures. Here at least partially shows schematically:

• 1 a wind turbine with a system for operating the wind turbine; and
• 2nd a method for operating the wind turbine 1 .

1 shows a wind turbine 1 with a tower 6 , on which a tower head in the form of a machine gondola 5 is appropriate. The machine gondola 5 is about a vertical axis of rotation A rotatable to the respective wind direction WR , indicated by an arrow, to be able to be tracked.

At the machine gondola 5 is a rotor 2nd around a rotor axis R rotatably mounted. This rotor has three rotor blades 3a , 3b , 3c by means of which a rotor hub 4th is set into rotation when the wind acts. The rotor 2nd is preferably coupled to a generator device (not shown) via a transmission (not shown) in order to generate electrical energy.

The rotor blades 3a , 3b , 3c are each about a rotor blade longitudinal axis E pivotable, the degree of pivoting being indicated by an angle of incidence.

The wind turbine 1 also has a system 10th to operate the wind turbine with several components in the 1 represented by the curly bracket.

The system 10th preferably also has sensor devices 30a , 30b on which each on one of the rotor blades 3a , 3b , 3c are arranged to the blade load on the respective rotor blade 3a , 3b , 3c capture.

Sensor signals S (not in 1 are shown) with sensor data in the system 10th recorded and by means of an interface 20th and an evaluation means 40 passed on.

This evaluation tool 40 is set up based on the sensor signals S Leaf loads M1 , M2 , M3 (also not in 1 shown) on the rotor blades 3a , 3b , 3c to investigate.

A tax resource 50 of the system 10th is set up on the basis of the determined blade loads M1 , M2 , M3 of the rotor blades 3a , 3b , 3c the wind turbine 1 , in particular the setting angle of the rotor blades 3a , 3b , 3c about the respective rotor blade longitudinal axis E to adjust. In this way, safe, wear- and / or yield-optimized operation of the wind power plant is achieved 1 guaranteed.

The arrangement of the elements shown 20th , 40 and 50 is to be understood as an example; in an advantageous embodiment, partial functions or the entire functionality of the elements can also be used 20th and 40 adjacent to the control means 50 be arranged in the rotor hub, or partial functions of the control means 50 adjacent to the elements 20th and 40 in the machine gondola.

One of the system according to the invention 10th procedure carried out 100 to operate the Wind turbine 1 is in 2nd represented by a block diagram. The system accordingly 10th Means or modules, in particular the evaluation means 50 , which are implemented in terms of hardware or software technology and set up to implement such a method 100 computer-aided execution.

Preferably includes a process for controlling the wind turbine 1 based on the leaf load M1 , M2 , M3 of the rotor blades 3a , 3b , 3c three basic steps.

So in a first step 101 Sensor signals S the sensor devices 30a , 30b detected. These sensor signals S characterize one on the rotor blades 3a , 3b , 3c adjacent leaf load M1 , M2 , M3 , especially an impact bending moment. Based on these sensor signals S become the impact bending moments M1 , M2 , M3 on the rotor blades 3a , 3b , 3c by means of an assignment rule in a second step 102 certainly. Finally, the pitch angle of the rotor blades 3a , 3b , 3c individually based on the impact bending moments determined in each case M1 , M2 , M3 and any other control parameters of the wind turbine 1 in a third step 103 determined.

As stated at the beginning, static rotational frequency effects are a problem in the control of wind energy plants with a blade feedback control, because they lead to increased loads and wear in the blade adjustment systems of the wind energy plant.

For this reason, in the process 100 in addition to the sensor signals S an adjustment function when calculating the respective individual blade load M1 , M2 , M3 considered. This adjustment function is used in particular as a calibration term in an assignment rule between the sensor signals S as the input variable and the individual sheet loads M1 , M2 , M3 considered as an output variable. There are preferably several adjustment functions, the adjustment functions being sensor-specific or individual for each rotor blade 3a , 3b , 3c are. Has a rotor 2nd three rotor blades 3a , 3b , 3c on, so are in the evaluation means 40 preferably also stored three different matching functions. At least depending on a number Z of the rotor blades should at least Z-1 Sensor-specific adjustment functions in a wind turbine 1 are used.

The matching functions or matching terms in particular cause differences between the values of the individual sheet loads M1 , M2 , M3 different rotor blades 3a , 3b , 3c at least reduced or even fully adjusted over a sufficiently long averaging time. This effect is referred to below as matching.

The matching terms preferably have at least one matching parameter AP on. In the case of a linear adjustment term, this adjustment parameter is a so-called gain parameter or gain.

wobei m ein Umrechnungsfaktor von jenem Teil der Sensorsignale S, welche das erste Rotorblatt 3a betreffen, zur Blattbelastung M ist, Gain der Angleichparameter und in diesem Fall gleichzeitig der Angleichterm ist und C ein konstanter Offset ist. Bei einem optimal kalibrierten System unter optimalen Betriebsbedingungen, beispielsweise ohne Toleranzen in der Rotorblattproduktion und der Rotorblattjustierung, bei dem an allen Rotorblättern stets der gleiche Wert der Biegemomente anliegt, wäre der Gain 1.The assignment rule preferably has for determining the individual sheet bending moment M1 , in particular the impact bending moment, of a first rotor blade 3a the following form: $$\text{M}1\left(\text{S}\right)=\text{Gain}·\text{m}·\text{S}+\text{C.},$$

where m is a conversion factor from that part of the sensor signals S which is the first rotor blade 3a relate to the sheet load M, Gain is the adjustment parameter and in this case the adjustment term and C is a constant offset. In the case of an optimally calibrated system under optimal operating conditions, for example without tolerances in the rotor blade production and the rotor blade adjustment, in which the same value of the bending moments is always present on all rotor blades, the gain would be 1 .

The matching term is preferably adjusted continuously. In particular, the at least one adjustment parameter AP in this case continuously on the basis of the sensor signals recorded for controlling the wind energy installation S in one step 105a adjusted. This is particularly advantageous since the sensor device 40a , 40b can change their properties over time (sensor drift). It is also possible that the rotor blades 3a , 3b , 3c or also a device for adjusting the setting angle of the rotor blades 3a , 3b , 3c damaged, which can result in new mass-related imbalances, for example. Aerodynamic imbalances can also change, for example if aerodynamic attachments on the rotor blades, such as spoilers or vortex generators, fall off. All of these changes can result in new orbital effects. These can be adjusted or adjusted by the at least one adjustment parameter AP again at least be reduced.

To determine the at least one adjustment parameter AP is in step 105a preferably a difference between the individual leaf loads M1 , M2 , M3 at least two rotor blades 3a , 3b , 3c the wind turbine 1 , in particular the individual impact bending moment. The respective impact bending moments M1 , M2 , M3 are determined with the adjustment term valid at that moment. Then the adjustment term, in particular its adjustment parameter AP , chosen so that the difference between the values of the individual impact bending moments is reduced or, viewed over a sufficiently long averaging time, is even completely canceled. If only one sensor-specific adjustment function is used, only one of the impact bending moments is preferably used M1 to the other of the impact bending moments M2 aligned.

Preferably, especially in wind turbines 1 with more than two rotor blades 3a , 3b , 3c , an adjustment of the impact bending moments M1 , M2 , M3 of the rotor blades 3a , 3b , 3c by means of an average M carried out, which in an additional step 104 is calculated as in 2nd is shown in dashed lines. The individual impact bending moments M1 , M2 , M3 are then each with the average impact bending moment M compared and the respective adjustment term, in particular the respective adjustment parameter AP , in the step 105a chosen in such a way that the differences between the respective impact bending moments M1 , M2 , M3 are reduced, in particular eliminated.

Regardless of whether one or more adjustment functions when adjusting the values of the individual blade loads M1 , M2 , M3 are used to determine the respective adjustment parameter AP of a respective rotor blade 3a , 3b , 3c preferably a delay function, in particular a function which exhibits a behavior of a PT1 element, is used.

The delay function preferably has the property that in each case a predefined number of values of the determined individual impact bending moments M1 , M2 , M3 this rotor blade 3a , 3b , 3c be taken into account. Alternatively, it is not the number that is predefined, but a time constant A , which preferably defines a predefined time period. When determining the adjustment parameter AP then all values of the sheet bending moments determined in this period M1 , M2 , M3 of the respective rotor blade 3a , 3b , 3c used.

Such a delay function preferably has a behavior of a FIFO data memory or a ring memory. Depending on the predefined number or predefined time period, older values of the respective impact bending moment are used M1 , M2 , M3 when determining the adjustment parameter AP then no longer taken into account as soon as a recent value of the same impact bending moment M1 , M2 , M3 which was currently determined when determining the adjustment parameter AP is taken into account. Furthermore, the delay function preferably makes younger values of the respective impact bending moment in the calculation of the adjustment parameter AP weighted higher than older values. The weighting between younger and older values also depends on the time constant A from.

Both measures result in a statistical averaging of the values of the adjustment parameter, which in particular reduce the significance of short-term operating point deviations or measurement deviations.

As a time constant A For example, a time period of approximately 1200 seconds can be set.

Furthermore, it can be provided in the delay function that when adapting the at least one of the sensor-specific adaptation functions, the values of the determined bending moments are weighted in such a way that newly considered values bring about a change in the respective adaptation parameter which does not exceed a predetermined limit value ζ. A change beyond this predetermined limit value erfordert preferably requires a large number of newly considered values of the impact bending moment M1 , M2 , M3 , in particular at least one third of the values taken into account for the adjustment of an adjustment term.

By using delay functions when adjusting the impact bending moments M1 , M2 , M3 different rotor blades 3a , 3b , 3c statistically high-quality results can be achieved. Furthermore, the influence of short-term incorrect measurements or effects, such as a temporary ice build-up, on the future calculation of the impact bending moment M can be kept low.

Preferably, when adapting the sensor-specific adaptation function or when determining the adaptation parameter AP such values of the respective impact bending moment M1 , M2 , M3 takes into account which during the acquisition of the sensor signals S , which correspond to the determined individual sheet loads M1 , M2 , M3 are based on predefined criteria, in particular with regard to the operating conditions and / or an operating state of the wind energy installation 1 , are fulfilled. The predefined criteria serve as a boundary condition for taking relevant measurements into account. Exemplary criteria are the setting angle of the rotor blades 3a , 3b , 3c , a rotor speed of the rotor 2nd or a service provided by the wind turbine 1 .

Preferably used to calculate the adjustment parameter AP only such values of the respective impact bending moment M1 , M2 , M3 of the respective rotor blade 3a , 3b , 3c takes into account a collective setting angle below a predefined setting angle limit of less than 4 °. Furthermore, the rotor speed should preferably be above a predefined speed limit of 70% of the nominal speed, and the power should be above a predefined power limit of approximately 40% of the nominal power. The criteria mentioned cause the determination of the adjustment parameter AP essentially values in the upper load range of the wind turbine 1 be taken into account. Adjustment processes of the control of the wind turbine 1 , in particular a change in the setting angle of the rotor blades 3a , 3b , 3c , have on the impact bending moments M1 , M2 , M3 Little influence in this operating range, so that the values of the respective impact bending moment taken into account for setting the adjustment parameter M1 , M2 , M3 are essentially the same over a certain period of time or are at least of the same order of magnitude.

A period of time of the presence of these boundary conditions in the method is preferred 100 measured by means of a time counter. To values over a period, which is determined by the time constant A can be defined, a real time period, which is determined by this time constant A is specified to be significantly longer than this period. For example, a PT1 element with a set time constant A = 1200 seconds can determine a real time period from several days in the past to determine the adjustment parameter if, on average, measurements can only be carried out twice a day for a period of a few seconds under the specified boundary conditions . The time constant is preferred A chosen in such a way that about 10 to 30 rotor revolutions are taken into account when determining the adjustment parameter. In this way, temporal and local fluctuations in wind speed and wind direction can be averaged out.

Employees doing repairs to the wind turbine 1 can make the time counter and adjustment terms, especially the adjustment parameters AP , preferably reset. This is done in particular when there are changes to the setting parameters or structural changes to the wind turbine 1 were made.

After a restart, in particular after a reset to a default state, the wind turbine 1 preferably smaller time constants come A for use as in ongoing operations. This allows a relatively quick, albeit statistically less precise, adjustment of the adjustment function after a restart. Current operation is preferably defined as operation after a period after the restart, which corresponds to one to ten times the value of the time constant at restart. A pair of values for the time constants A can preferably be _small = 120 seconds and A _large = 1200 seconds. The switchover is preferably carried out automatically when the time recorded by the time counter exceeds a multiple of the smaller time constant A _{small of} the PT1 link setting.

Preferably in the process 100 for the at least one adjustment parameter AP Limit values η specified. In one step 106 becomes the at least one adjustment parameter AP which in step 105a was determined, compared with this limit value η. Is the value of the adjustment parameter AP outside the limit value η or limit value range, then preferably a warning or error message in one step 106 spent. In this way, increasing imbalance on the rotor 2nd or a defective sensor device 30a , 30b be reliably recognized.

Measures to protect the wind turbine can then be made on the basis of these messages 1 to be hit. For example, the operation of the wind turbine 1 be stopped in the event of an error message. Alternatively, the wind turbine can continue to be operated with a reduction in power and / or speed with reduced loads.

The described method preferably comes 100 together with a method for calibrating the wind turbine 1 for use. A calibration is preferably taken into account by means of a calibration function. In this case, an assignment rule between the sensor signals S as the input variable and the individual impact bending moments M1 , M2 , M3 of the rotor blades 3a , 3b , 3c as an output variable in addition to the at least one adjustment term calibration terms, which can be generally referred to as a calibration function. These calibration terms can, in particular, cause sensor errors and specific properties of the respective rotor blade 3a , 3b , 3c be corrected so that a determined value of the respective blade load M1 , M2 , M3 , in particular the impact bending moment, is as close as possible to the actual value. The calibration terms preferably have correction parameters, similar to the matching terms KP on.

As in 2nd Such a correction parameter is shown in dashed lines KP based on one based on the sensor signals S determined respective Leaf loads M1 , M2 , M3 as well as position data P and reference data R in one step 105b certainly. The reference data R are preferably simulated target values of the individual blade loads or determined by means of reference measurements M1 , M2 , M3 , the position data P give a position of the rotor 2nd at. By comparing the values of the reference data R with the corresponding values of the determined individual leaf loads M1 , M2 , M3 can the respective correction parameters KP in relation to each rotor blade 3a , 3b , 3c be determined.

Even when determining the correction parameters KP In this case, a delay function, in particular a function that maps the behavior of a PT1 element, is preferably used. As described above in relation to the adjustment, this delay function also has a time constant B on.

With regard to the calibration function, there are also criteria as to which values of the individual sheet bending moments are used for calibration.

Because when calibrating, compared to the adjustment of the individual blade load described above M1 , M2 , M3 , the position of the rotor 2nd An important meaning is that the required real time period, which must be taken into account when using a delay function until the calibration is complete, is generally significantly longer than the real time period, which must be taken into account when aligning, because position data in relation to the rotor 2nd preferably not taken into account. This can occur even if the time constant B with regard to the delay function during calibration is chosen to be significantly shorter than the time constant A when aligning.

In general, adjustment is faster than calibration, even though the time constant A the adjustment function is usually greater than the time constant B the calibration function. However, because the adjustment process is carried out more often than the calibration process, the calibration is still slower. The boundary conditions for calibration processes are in fact less frequently fulfilled than the boundary conditions for the adjustment processes and the respective time constant is always "stopped" if the boundary conditions are not fulfilled.
The adjustment is preferably 63% complete after 1200 seconds. Calibration takes longer.

The determined values of the at least one adjustment parameter AP and possibly the correction parameter KP are used to determine the respective blade load M1 , M2 , M3 in the work step 102 used in the adjustment function and in the calibration function. The values of the respective blade load determined in this way M1 , M2 , M3 then, as in 2nd shown, on the one hand for controlling the wind turbine 1 in work step 103 are used, and on the other hand are used in turn to adapt the adaptation terms and / or the calibration terms in the work steps 105a , 105b spent. In this way it is possible to adjust the individual leaf load M1 , M2 , M3 and if necessary a calibration of the respective blade load M1 , M2 , M3 done iteratively in several steps.

It is pointed out that the exemplary embodiments described above are only examples which in no way limit the scope of protection, the application and the structure. Rather, the person skilled in the art is given a guide for the implementation of at least one embodiment by the preceding exemplary embodiments, it being possible for various changes, in particular with regard to the function and arrangement of the described components, to be carried out without leaving the scope of protection which results from the claims and gives these equivalent combinations of features.

Reference list

1

Wind turbine

2nd

rotor

3a, 3b, 3c

Rotor blade

4th

Rotor hub

5

gondola

6

tower

10th

system

20th

interface

30a, 30b

Sensor device

40

Means of evaluation

50

Tax funds

100

method

101-106

Procedural steps

R

Rotor axis

E

Longitudinal rotor blade axis

A

Axis of rotation

WR

Wind direction

S

Sensor signal

M1, M2, M3

Sensor-specific blade load

R

Reference data

P

Position data

Claims (17)

Method (100) for operating a wind power plant (1) which has at least two rotor blades (3a, 3b, 3c), comprising the following working steps: - Detection of sensor signals (S) which are suitable for characterizing blade loads (M1, M2, M3) on the rotor blades (3a, 3b, 3c); - Determining an individual blade load (M1, M2, M3), in particular an impact bending moment, for each rotor blade (3a, 3b, 3c) on the basis of the sensor signals (S) recorded on the respective rotor blade (3a, 3b, 3c) and a sensor-specific one Alignment function of the respective rotor blade (3a, 3b, 3c); and - Controlling the wind turbine (1), in particular the setting angle of the rotor blades (3a, 3b, 3c), on the basis of the determined individual blade loads (M1, M2, M3); wherein at least one of the sensor-specific adjustment functions of the rotor blades (3a, 3b, 3c) is adapted in such a way that a difference between the individual blade loads (M1, M2, M3) of different rotor blades (3a, 3b, 3c) is at least reduced. Method (100) according to Claim 1 , wherein a predefined number of values of the determined individual sheet loads (M1, M2, M3) are taken into account for adapting the adjustment function on the basis of a predefined time constant A, which preferably defines a predefined time period, preferably not an older value, in particular an oldest value more is taken into account as soon as a younger value, in particular a youngest value, is taken into account. Method (100) according to Claim 2 , the time constant A being chosen such that the difference between the individual blade loads (M1, M2, M3) over a predefined period of time, over a predefined number of determined blade loads (M1, M2, M3) or over a predefined number of rotor revolutions reduced to a value less than and / or equal to a limit value ε, the limit value ε being in particular zero. Method (100) according to one of the preceding claims, wherein to adapt the adjustment function, younger values of the individual blade loads (M1, M2, M3) are weighted higher than older values of the individual blade loads (M1, M2, M3), preferably using a predefined time constant A , which preferably defines a predefined time period. Method (100) according to one of the preceding claims, wherein the at least one of the sensor-specific adaptation functions is continuously adapted, in particular before, during or after a respective control process of a plurality of control processes. Method (100) according to one of the preceding claims, further comprising the following work step: - Determining (105) at least one adjustment parameter (AP) of the at least one of the sensor-specific adjustment functions on the basis of the difference between the individual blade loads (M1, M2, M3) of the different rotor blades (3a, 3b, 3c). Method (100) according to one of the Claims 2 to 6 , wherein a delay function, in particular a function of a PT1 element, is mapped when the values are taken into account. Method (100) according to one of the preceding claims, wherein the at least one of the sensor-specific adaptation functions is only adapted if predefined criteria, in particular with regard to at least one operating condition and / or an operating state of the wind power plant (1 ) are fulfilled. Method (100) according to one of the Claims 6 to 8th , further comprising the following working steps: - Adjustment (105) of the value of the adjustment parameter (AP), each with a predefined limit value η, in particular 0.85 <η <1.15, preferably 0.92 <η <1.08, particularly preferably 0.94 <η <1.06; and - outputting (106) a status message, in particular a warning message or an error message, when the adjustment parameter (AP) reaches and / or exceeds or falls below the respectively predefined limit value η. Method (100) according to one of the preceding claims, wherein the control comprises determining an individual setpoint angle value for each rotor blade (3a, 3b, 3c) and wherein the setpoint angle of each rotor blade (3a, 3b, 3c) based on this individual setpoint angle Setpoint is set. Method (100) according to one of the preceding claims, wherein all sensor-specific adaptation functions are adapted, further comprising the following work step: - determining (104) an average sheet load ( M ) on the basis of the individual blade loads (M1, M2, M3) of all rotor blades (3a, 3b, 3c); the sensor-specific adjustment functions based on the average leaf load ( M ) be adjusted. Method (100) according to one of the Claims 6 to 11 , wherein respective adjustment parameters (AP) of the sensor-specific adjustment functions based on a difference between the individual blade loads (M1, M2, M3) of the rotor blades (3a, 3b, 3c) and the average blade load ( M ) be determined. Method (100) according to one of the Claims 2 to 12th , wherein the determination of the individual leaf loads (M1, M2, M3) is additionally carried out on the basis of a calibration function, the calibration function being adapted by means of calibration processes on the basis of the determined leaf load, recent calibration processes depending on a predefined time constant B, which is preferably a predefined time period are defined, weighted higher than older calibration processes, wherein the time constant B in relation to the calibration function is preferably smaller than a time constant A in relation to the adjustment functions, in particular 1.5 to 2.5 times larger. Computer program comprising instructions which, when executed by a computer, cause the computer to carry out the steps of a method according to one of the Claims 1 to 13 to execute. Computer-readable medium on which a computer program is based Claim 14 is saved. System (10) for operating a wind turbine (1) which has at least two rotor blades (3a, 3b, 3c), comprising: - An interface (30) set up for detecting sensor signals (S) which are suitable for characterizing blade loads (M1, M2, M3) on the rotor blades (3a, 3b, 3c); - Evaluation means (50), set up to determine an individual blade load (M1, M2, M3), in particular an impact bending moment, for each rotor blade (3a, 3b, 3c) on the basis of those recorded on the respective rotor blade (3a, 3b, 3c) Sensor signals (S) and a sensor-specific adaptation function of the respective rotor blade (3a, 3b, 3c) and for adapting at least one of the sensor-specific adaptation functions of the rotor blades (3a, 3b, 3c) in such a way that there is a difference between the individual blade loads (M1, M2 , M3) of different rotor blades (3a, 3b, 3c) is at least reduced; - Control means (60) of the wind turbine (1), set up to control the wind turbine (1), in particular the setting angle of the rotor blades (3a, 3b, 3c), on the basis of the determined individual blade loads (M1, M2, M3). A wind turbine (1) with a system according to Claim 16 .

Images