METHOD FOR CONTROLLING A WIND POWER PLANT

Description

[0001] The present invention relates to a method for controlling a wind turbine and the present invention relates to a corresponding wind turbine.

[0002] Wind turbines are known for generating electrical power from wind. To do this, an aerodynamic rotor is driven by the wind to rotate at a suitable rotor speed, thereby rotating the wind turbine's generator so that it can deliver power. The wind turbine operates at an operating point that depends primarily on the wind speed and is characterized by the rotor speed and the output power.

[0003] Such operating points are usually and essentially stationary operating points, in which the generator speed and the output power are essentially constant. Typically, such operating points are essentially specified by a speed-power characteristic curve. Such a speed-power characteristic curve specifies power values, i.e., the power to be delivered by the wind turbine or generator, for speed values, namely the rotor speed.

[0004] The control principle of such a speed-power characteristic works by detecting a speed, and adjusting the corresponding power according to the speed-power characteristic. If the wind speed increases—to give an illustrative example—the speed also increases, so that a higher power is adjusted according to the speed-power characteristic. The higher power setting counteracts the acceleration caused by the increase in wind speed. The power is then increased according to the speed-power characteristic until the speed no longer increases, whereby a stable operating point is found.

[0005] In this respect, these operating points represent stationary and thus stable operating points in which the wind turbine can be operated well.

[0006] However, if the wind speed increases above a nominal wind speed, the rotor speed reaches its nominal speed and the power output reaches its nominal power. The speed should then not increase any further, and neither should the power output. A further increase in speed can be counteracted by adjusting the blade angle of the rotor blades and thus their angle of attack. The rotor blades are then gradually turned out of the wind depending on the wind speed. Their blade angle therefore increases. This range, in which the wind speed is so high that an increase in speed and power must be prevented by adjusting the rotor blades in order to operate the wind turbine at the nominal speed and power, is also referred to as the full load range or full load operation.Even at full load, the wind turbine essentially operates at stationary and therefore stable operating points. Slight fluctuations in wind speed can naturally lead to slight fluctuations in the operating point.

[0007] During partial load operation, i.e., when the wind speed has not yet reached the nominal wind speed, the blade angle is usually constant, typically in the range of 0 to 10 degrees, and particularly approximately 2 to 7 degrees. The wind turbine is designed so that the operating points during partial load operation lead to aerodynamically optimal operation. In particular, a constant tip speed ratio is often used, at least in a core area. This design of the wind turbine is realized by the speed-power characteristic curve in the wind turbine.

[0008] However, it may be necessary to deviate from these operating points, particularly the aerodynamically optimal operating points, but also the operating points during full-load operation, due to an external specification. Such an external specification may be the specification of a new, particularly reduced, rotor speed value, but also (alternatively or additionally) the specification of a particularly reduced power value for the output power.

[0009] In principle, higher values can be specified for both speed and power. However, a wind turbine usually operates at a level where both speed and power are already as high as possible and / or as high as reasonable and / or optimal. However, if the wind turbine is currently operating at a reduced operating point, i.e., at a lower speed and / or lower power than possible and / or reasonable, specifying a higher speed and / or higher power is particularly appropriate.

[0010] Such a change in the operating point due to artificial specifications usually constitutes a disturbance and can place a strain on the turbine. Sudden changes, in particular, can lead to significant, particularly mechanical, stress on the wind turbine. Among other things, such a change in the operating point can trigger mechanical vibrations and, in the worst case, even intensify them. Tower vibrations can be particularly relevant here.

[0011] Such tower vibrations can be triggered particularly by a change in the operating point, which alters the shear load exerted by the wind on the aerodynamic rotor of the wind turbine, thereby changing the deflection of the tower in the area of the aerodynamic rotor, i.e., at the tower head. In other words, the tower head can swing back, which can lead to stress on the tower and thus the wind turbine as a whole.

[0012] Such vibrations can usually be counteracted or avoided by gently adjusting the operating point. However, it is often the case that changed operating parameters should be implemented as quickly as possible. Such changes to operating parameters are often linked to safety or other stability considerations. For example, a speed reduction may be required if an endangered bird approaches the wind turbine.

[0013] A power reduction may be a requirement of the electrical grid and therefore important for the stability of the electrical grid. For example, a rapid power reduction may be necessary in the event of an unexpected load shedding in the electrical grid. An emergency shutdown of a wind turbine may also be the reason for specifying a changed operating point.

[0014] In all these cases, which are only examples, a slow change in the operating point is undesirable, perhaps even impossible.

[0015] During the patent granting procedure, the following documents were considered, among others:

[0016] DE 10 2019 105296 A1,

[0017] Andreas Søndergaard Pedersen ET AL: "Safe Operation and Emergency Shutdown of Wind Turbines", 2012-05-31, XP055149694, and

[0018] US 2010 / 283247 A1.

[0019] The present invention is therefore based on the object of addressing at least one of the aforementioned problems. In particular, a solution is to be proposed that realizes a rapid change in an operating point, i.e., in particular, a rapid speed and / or power reduction, while simultaneously minimizing loads on the wind turbine, particularly mechanical loads on the tower. At the very least, an alternative to previously known solutions is to be proposed.

[0020] According to the invention, a method according to claim 1 is proposed. The method thus relates to the control of a wind turbine connected to an electrical supply grid, which has a rotor with adjustable blade angle rotor blades, is operable at a variable speed and is designed to generate turbine power from wind. The rotor can also be referred to as an aerodynamic rotor; it is driven by the wind via its rotor blades. The speed of the rotor, i.e. the rotor speed, is variable. It can be changed in particular by appropriate operating control, but also depends on the wind. The turbine power is the power generated and output by the wind turbine. Alternatively, the power output by the generator can also be considered.Although there may be differences between the power delivered by the generator and the power ultimately delivered by the wind turbine, these are not relevant for the present invention.

[0021] A blade angle control is provided to adjust the blade angle. Among other things, a corresponding control signal can be sent to an actuator so that corresponding pitch motors, also known as pitch drives, adjust the blade's pitch. In principle, the blade angle control can be provided entirely or partially for all rotor blades simultaneously. Particularly in cases where the rotor blades are not individually adjusted, the blade angle control can output three identical setpoints, namely one for each pitch drive of a rotor blade if the rotor has three rotor blades.

[0022] Any statements regarding blade angle control can also be understood to mean that this occurs individually and identically for each rotor blade. For example, if a specific blade angle is specified for the blade angle control, this is specified for each rotor blade.

[0023] A speed control is provided to regulate or limit the speed. This speed control can also process setpoints and control corresponding control elements and / or actuators or specify control values for them.

[0024] Power control is provided to limit the system's output. Power control converts a power setpoint into a power output. It usually has a limiting effect, since the output power cannot naturally be increased beyond the power available from the wind. Power control can be activated or acted upon by an inverter feeding into the electrical grid and / or by an active rectifier that controls the generator, particularly the generator current.

[0025] Furthermore, the method works in such a way that the wind turbine can be operated at a predeterminable operating point, whereby the operating point is characterized at least by the speed, i.e. the rotor speed, and the turbine power, i.e. in particular the generator power or output power. In partial load operation, this can particularly be the operating point resulting from a predefined speed-power characteristic curve. In full load operation, the operating point will be characterized by the rated speed and rated power. Even in full load operation, the operating point can therefore be specified by the speed and / or power being reduced compared to the rated speed or rated power. In exceptional cases, however, it is also possible for the speed and / or power to be higher than the rated speed or rated power, at least slightly and / or temporarily.

[0026] One step in the process is operating the wind turbine at its first operating point, which can also be referred to synonymously as the normal operating point, depending on the prevailing wind speed. In partial load operation, the speed and power of the normal operating point are then determined in particular from the speed-power curve. As a precaution, it should be noted that instead of a power specification, a torque specification is also possible to set such an operating point, although this may depend on the type of wind turbine. The power or torque specifications have the same effect in that they can be converted into one another at a fixed speed using the equation P = n m, which applies to the steady state. In this equation, P is the power (i.e. active power), n is the speed, and m is the torque.

[0027] In a further step, a check is carried out for a curtailment request, which requires a reduction in speed and / or system power. Such a curtailment request can be specified externally, for example, by a utility grid operator, or internally through a calculation rule and / or time-dependent, particularly depending on the time of day. Examples of curtailment requests are provided and explained below.

[0028] In particular, such a reduction request can consist of specifying a new value for the system output that is lower than the current value. Likewise, a reduced setpoint can be specified for the speed, which specifies a lower speed than the current speed.

[0029] If such a curtailment request exists, a new operating point is determined as the target operating point depending on the curtailment request, whereby the target operating point is characterized by a target speed and a target turbine power. Typically, the curtailment request only specifies a reduced speed value or only a reduced output power. When determining a new operating point as the target operating point, at least the other variable is determined. If a speed reduction is specified, from which a new speed value results as the target speed, an associated target turbine power is determined. If power curtailment with a corresponding power specification were to be specified as the target turbine power, an associated target speed is determined. This defines the target operating point, and the wind turbine is then to approach this target operating point.

[0030] The next step is to determine a desired blade angle as the target blade angle for the target operating point. Thus, such a blade angle is not regulated, but rather predefined. Typically, a blade angle is changed by a speed control system, which adjusts the blade angle depending on the deviation between the desired and actual speed. This is not provided here, but rather supplementary.

[0031] The proposal here is not to determine the target blade angle through a control process, essentially as the final result of a control process, but to specify it specifically from the outset. After receiving the control request, a target speed is specified as the new speed and a target turbine power as the new turbine power, and the target blade angle is calculated for this purpose. Starting from the current operating point, in this case the first operating point or normal operating point, further conclusions about the wind situation are also possible.

[0032] It is particularly important to note that while the current operating point is determined by the speed and turbine power, and possibly also by the set blade angle, the current wind speed essentially determines the current operating point. This means, in particular, that when a curtailment request is specified, the target turbine power can be directly specified as the target turbine power, but the associated target speed can vary greatly depending on the prevailing wind situation. Since the power extraction from the wind is artificially reduced due to the curtailment request, the rotor blades must also be adjusted; thus, a new blade angle, namely the target blade angle, must be specified.

[0033] Conversely, if a curtailment request specifies a reduced speed, a suitable value for the target turbine power and also a suitable target blade angle would first have to be found. In the case of a speed reduction, the desire would naturally be to reduce the power as little as possible, but in most cases it will still be necessary to reduce the turbine power and thus the target turbine power. The main reason for this may be the maximum generator torque, which means that at a low speed the power will also be lower. However, aerodynamic reasons may also play a role or be relevant, because - to put it simply - with a significantly reduced speed the optimal aerodynamic power can no longer be extracted from the wind.

[0034] It is further proposed that, in order to change the blade angle to the target blade angle, a pilot control blade angle or a pilot control adjustment rate, which describes an adjustment rate of the blade angle, is specified via a blade angle pilot control, wherein the blade angle pilot control directly passes the pilot control blade angle or the pilot control adjustment rate to the blade angle control for implementation.

[0035] The blade angle is therefore not regulated, but rather specified and adjusted. The setting can be made via an adjustment rate, i.e., by specifying the speed at which the blade angle is to be adjusted. The pilot control blade angle can also be specified as a curve, either directly or via a corresponding adjustment rate. If only a constant pilot control blade angle is specified, this can correspond to the target blade angle.

[0036] In other words, a blade angle is pre-controlled, meaning it is not, or at least not predominantly, adjusted by a blade control system. The blade angle to be pre-controlled, i.e., the pilot blade angle, is essentially simply passed to the blade angle control system for actual implementation via appropriate actuators.

[0037] This ensures that the corresponding blade angle can be controlled as quickly as possible without feedback from a control system delaying this process or causing oscillations. In particular, such feedback, which could induce oscillations, is avoided in every case, regardless of the initial operating point being changed to the target operating point.

[0038] In particular, the pilot blade angle corresponds to the target blade angle. However, it is possible to select a slightly lower value for the pilot blade angle. In this case, the target blade angle is not fully piloted, but for example only by 95%. The pilot blade angle can correspond to a value of 80 to 99% of the target blade angle, or to a value of 90 to 99% of the target blade angle. In this case, the basic principle of pilot control of the blade angle is not abandoned, but a different control, possibly even regulation to the target blade angle, can be used for the last few percent. This can compensate for small deviations that can arise with pure control without feedback. Such deviations can arise, for example, if the wind changed slightly during the adjustment process.

[0039] The pre-control of the pilot blade angle, in particular the target blade angle, can be implemented in such a way that this pilot blade angle is directly passed to the blade angle control for implementation or that the latter implements the blade angle directly by correspondingly controlling the actuators. This would then result in a jump in the blade angle. Since drives for blade pitch adjustment, in particular pitch motors, can reach their maximum speed very quickly, such a blade angle pre-control, in which the pilot blade angle is directly specified and implemented, could result in the adjustment drive or pitch motor rotating at this maximum speed until the target blade angle is reached, or until shortly before, e.g. up to a value in a range of 90 to 99% of the target blade angle or target blade angle.

[0040] Alternatively, to change operation, the pitch rate can be specified instead of the pilot blade angle, which accordingly forms a set pitch rate or a portion thereof. The pitch rate indicates a pitch change per unit of time and thus, in the physical sense, a rotational speed, namely, for the period of time required to reach the target blade angle. Such a pitch rate can also be understood as a temporal ramp. The pitch rate then ranges from the initial blade angle of the normal operating point to the target blade angle of the target operating point.

[0041] The method according to the invention therefore operates in particular such that, following a derating request, a new target operating point is determined, for which the target speed, the target turbine power, and the target operating point are determined, particularly as a step before the actual change to the operating point is initiated. To implement the change, a pilot blade angle is also determined in advance, which can, however, correspond to the target blade angle, and is pilot-controlled. The pilot blade angle, and in particular the target blade angle, are then implemented directly by the blade angle control without closed-loop control, i.e., without feedback.

[0042] Otherwise, if the change from the normal operating point to the target operating point is not carried out, the blade angle control or blade angle adjustment works in such a way that it is integrated into a speed control, i.e., a speed control loop. Depending on the deviation between the target and actual speed, a blade angle change is determined, or a blade angle adjustment rate, by which or according to which the rotor blade is adjusted. This directly changes the speed, which, via the feedback for comparing the target and actual speed values, leads to a changed speed control deviation and thus, again or continuously, changes the blade angle change or adjustment rate within this control loop.

[0043] Such integration of the blade pitch control into this speed loop is at least temporarily suspended by the blade angle feedforward control when the blade angle is pre-adjusted, i.e., when the pre-control blade angle is converted. It is generally suspended until the target blade angle is reached. However, if only minor deviations from the target blade angle remain and / or the blade angle feedforward control only results in minimal adjustment activity, the blade angle control integrated into the speed control loop can become active again or make a complementary, particularly additive, contribution.

[0044] According to one aspect, it is proposed that the pilot control specifies the pilot blade angle or the pilot control adjustment rate independently of a speed deviation as a deviation of a detected speed from a predetermined speed, and / or specifies it independently of the detected speed and independently of the detected system power, and additionally or alternatively the pilot control specifies the pilot blade angle or the pilot control adjustment rate independently of the speed control and independently of the power control. The pilot control and thus the adjustment of the blade angle therefore does not depend, at least not directly, on the speed or the system power. The adjustment of the blade angle will influence the speed and the system power, but this influence is not fed back in the sense of a control, because the pilot control does not attempt to adjust the detected speed or the detected system power to a setpoint.This is precisely what characterizes feedforward control: it specifies the blade angle, or rather the pitch rate, and this is used directly to adjust the blade angle. This avoids feedback loops that can lead to oscillations. However, this does not mean that the speed has to be completely ignored. For example, a wind estimator can be used to determine the wind speed, and the wind speed estimated in this way can be used to determine the blade angle to be fedforward controlled. However, the speed is not taken into account in the control system, i.e. not in the sense that a speed deviation or speed error is taken into account. It is also possible to consider the tip speed ratio, which is defined as the quotient of blade tip speed and wind speed. But this also does not take into account any speed deviation and therefore no speed control.

[0045] The fact that the feedforward control operates independently of the speed control and independently of the power control means in particular that the feedforward control is neither included in the speed control loop nor in the power control loop.

[0046] In particular, it is intended that the change in the blade angle from the initial blade angle to the target blade angle be determined entirely or at least predominantly by the pilot control. Thus, not only is the pilot blade angle determined by the pilot control, independent of the detected speed and power, and independent of the speed and power control, but also that no other control or regulation changes the blade angle at the same time.

[0047] However, it cannot be completely ruled out that further regulation or control also influences the blade angle to a subordinate extent. For example, a further adjustment by a control could be provided for the last 5%. In particular, the blade angle is changed by the feedforward control at least over 80% of the path from the initial blade angle to the target blade angle. In addition or alternatively, the blade angle is changed at least 80% by the feedforward control, in the sense that a set target blade angle or a set target pitch rate is specified by the feedforward control to at least 80% on average over the travel path from the initial blade angle to the target blade angle.

[0048] According to one aspect, it is proposed that an ideal blade angle be determined as the target blade angle, which leads to the target speed and target power under steady-state conditions at the target operating point. In this respect, the steady-state conditions relate in particular to a constant wind speed. The wind speed is always subject to at least slight fluctuations, so that the assumption of a constant wind speed is therefore an idealized consideration. This constant, i.e. in particular the idealized wind speed, can be assumed to be the one assigned to the first operating point. The ideal blade angle can therefore also be assigned to a wind speed.

[0049] In partial load operation, the prevailing, i.e. current, wind speed can be derived from the current speed and the currently generated turbine power. Such an operating point corresponds to a point on the speed-power curve, and this speed-power curve is assigned to a wind speed curve, so that a wind speed can be read off as the current wind speed for each point on the speed-power curve. Measuring the wind speed is, of course, also an option. Such a wind speed measurement is also possible, but can lead to inaccuracies depending on the conditions and the measuring device. In addition, to determine the ideal blade angle - put simply - an ideal wind speed is also sought. This can be easily read off from the current operating point.

[0050] If the wind turbine is operating at full load, the wind speed can be determined from the first blade angle, i.e., the blade angle at the first operating point. The design of the wind turbine can also be used for this purpose. However, empirical values can also be used, or, of course, the wind speed can also be measured.

[0051] It is also assumed that the wind speed of the first operating point will still be present when the target operating point is reached. For the target operating point, either a speed or a power is specified by the curtailment request.

[0052] For example, if the speed for the target operating point is specified, the maximum possible power can be sought, which in most cases is the power at which the nominal torque is present. To do this, a blade angle can be found at which this power can be extracted from the wind at the existing wind speed and specified speed. This is then the ideal blade angle, or the target blade angle.

[0053] It may also happen that other influences, such as aerodynamic boundary conditions, affect the maximum achievable power, thus further limiting it. To account for this, simulations can be conducted in advance in which wind speed, power, and rotational speed are varied. This allows a corresponding data field to be determined and stored. To then find a suitable power value with the corresponding blade angle for a given wind speed and specified rotational speed, corresponding values can be selected from the stored data and interpolated between them.

[0054] A similar situation arises when the derating request specifies a power reduction. In this case, a suitable speed is sought, which is usually not the maximum speed. Rather, a speed is sought that ensures the most aerodynamically stable operation. For this purpose, simulations can also be recorded in advance and stored in a data array, which can then be accessed to determine the speed and the associated blade angle in order to find a suitable value. Here, too, interpolation between two values is possible.

[0055] According to one aspect, it is proposed that the pilot control predetermine a fixed blade angle or a fixed temporal blade angle profile as the pilot control blade angle, or that the pilot control predetermine a fixed adjustment rate as the pilot control adjustment rate. Furthermore or alternatively, it is proposed that the pilot control blade angle or the pilot control adjustment rate be determined using a model that represents the behavior of the wind turbine. In particular, the target blade angle can be specified as the fixed blade angle.

[0056] This allows for particularly rapid adjustment to this target blade angle. However, it is also possible to specify a precise temporal progression of the blade angle. For example, a temporally continuous progression, particularly a linear progression, can be specified to move from the initial blade angle to the pilot blade angle, particularly the target blade angle. However, other progressions are also possible, in which the blade angle is initially adjusted faster and then more slowly, or vice versa, in a predetermined manner.

[0057] Likewise, a pitch rate can be specified, which ultimately reflects the change in blade angle over time. The integrated pitch rate thus leads to the pilot blade angle, specifically the target blade angle. By specifying the pitch rate, the actuator, especially the pitch motor or pitch drive, can be controlled more effectively.

[0058] To implement the feedforward control, it is particularly recommended to use a model that represents the behavior of the wind turbine. This allows possible operating points to be modeled, taking the prevailing wind speed into account, that can be considered for implementing the curtailment request. Based on this, a suitable target operating point can be found based on one or more boundary conditions. Or, if only one reasonable target operating point is considered for implementation, this can also be found.

[0059] When specifying a derating request that specifies a reduced speed value, one of the constraints may be to maximize the system's output. Limits to be observed, such as the maximum generator torque and, of course, the maximum permissible power, may be additional constraints.

[0060] If reduced power is specified as a curtailment requirement, one boundary condition may be to find the most aerodynamically stable operating point. In this case, the model particularly considers and compares operating points that extract an aerodynamic power from the wind that corresponds to the reduced turbine power specified by the curtailment requirement, plus power losses, which are neglected here for the purposes of explanation. This aerodynamic power depends on the rotor speed and the blade angle. A tip speed ratio can often be used here to consider this. In any case, operating points can be found through modeling that lead to the desired aerodynamic power, which is extracted from the wind.

[0061] The model can also be used to model the blade angle progression and / or the pitch rate. The model can take into account the states the wind turbine assumes on the way from the initial operating point to the target operating point. The pilot blade angle or pitch rate can be specified accordingly. For example, this model can prevent a stall, which could otherwise lead to the wind turbine coming to a halt.

[0062] In particular, it is proposed that the pilot blade angle and thus in particular the target blade angle or an adjustment rate for it, be determined by a model and not by a control loop in which the blade angle is adjusted until the otherwise desired operating point is reached.

[0063] According to one aspect, it is proposed that a target tip speed ratio be provided for the target operating point as the tip speed ratio characterizing the target operating point, and that the target blade angle be determined such that the target tip speed ratio is established at the target operating point under steady-state conditions. The tip speed ratio can be used to characterize the aerodynamic behavior of the rotor at the respective operating point, in this case the target operating point. The tip speed ratio, in this case the target tip speed ratio, can be derived from a model for the target operating point. It is also possible for the tip speed ratio to be specified from empirical values for the prevailing wind speed and the given curtailment requirement. The target blade angle is then determined such that this target tip speed ratio is established at the target operating point.

[0064] According to one aspect, it is proposed that a curtailment request be a speed request, as a request for a speed reduction, in particular to protect an approaching animal, such as a bird or a bat, wherein the speed request in particular specifies a concrete target speed in terms of altitude. In particular, this could involve detecting an approaching animal, in particular a bird or bat, and reducing the speed of the wind turbine depending on the animal's flight direction and flight speed to such an extent that the animal is not struck or injured by a rotor blade. Such a speed is then specified as a curtailment request. For example, a value of one revolution per minute (1 rpm) can be specified, to give a simple example. This speed mentioned as an example would then be the target speed, which could be specified in terms of altitude using this value.

[0065] It is also possible for the curtailment request to specify a power requirement as a requirement to reduce the system's output, in particular to support the electrical supply grid, whereby the power requirement in particular specifies a specific target output in terms of level. For example, a grid operator can request a power reduction and provide a percentage curtailment value for this purpose. Such a percentage curtailment value can be passed to a central farm control system, which then specifies a specific output value for each wind turbine in the farm, for example, 1 MW if the wind turbine previously generated and fed in more power. In any case, this power value, which represents a specific target output in terms of level, can be taken into account as a curtailment request in order to determine the target operating point based on it.

[0066] According to one aspect, it is proposed that in the event of a power request, a maximum power is specified as a curtailment request for the system power, the current system power, if it is greater than the maximum power, is reduced to the maximum power, and in order to change the speed to the target speed, the blade angle is adjusted to the target blade angle by means of the blade angle feedforward control, wherein the feedforward control specifies the blade angle or the adjustment rate as the manipulated variable.

[0067] It is therefore proposed that the power be reduced to the specified value as quickly as possible, in particular as quickly as physically possible. This can be achieved by ensuring that, upon receipt of the curtailment request, a wind turbine feed-in unit feeding into the electrical grid only feeds the power in accordance with the curtailment request.

[0068] This rapid power reduction can, in the short term, result in the wind turbine generating too much power from the wind, which cannot be fed into the grid. Such excess power can lead to a brief acceleration of the aerodynamic rotor, and / or it may be designed to route such excess power from an electrical intermediate circuit of the feed-in unit, or otherwise, through resistors to convert this excess power into heat. This can also be referred to as path chopping.

[0069] In any case, the wind turbine's operating point must be adjusted to the reduced power as quickly as possible. This specifically means reducing the speed and adjusting the blade angles to at least partially turn the rotor blades out of the wind. To do this, a new operating point is determined, namely the target operating point, and thus a target speed and a target blade angle are determined.

[0070] To implement this, i.e., to change the speed to the target speed, the blade angle is adjusted to the target blade angle using the blade angle feedforward control. In the simplest case, an adjustment rate of, for example, 5 degrees per second can be specified, at which the blade angle is adjusted to the target blade angle. Further options are described below.

[0071] According to one aspect, it is proposed that in the case of a speed request, the target speed is specified as a reduction request, the system power is adjusted depending on the specified target speed and, in order to change the speed to the target speed, the blade angle is adjusted to the target blade angle by means of the blade angle feedforward control.

[0072] The key difference between predefined speed control and predefined power control is that the target speed is fixed, while the turbine power can be used to reduce the speed, particularly for braking. In particular, the turbine power can initially be high in order to quickly reduce the speed. However, aerodynamic power, i.e., the power extracted from the wind, can also be specified for reduction, particularly negative power, in order to brake, which will be explained below.

[0073] In particular, it is proposed that, when a speed requirement is specified as a derating requirement, the turbine power or a generator torque be specified to brake the rotor in order to achieve the strongest possible rotor braking, together with a pre-controlled aerodynamic power. In this case, electrical braking can be applied in addition to the aerodynamic power. If the turbine power or generator torque is already at its maximum, the resulting braking effect can be maintained. Otherwise, the turbine power or generator torque can be increased as much as possible.

[0074] What both predefined speed control and predefined power control have in common is that, in order to change the speed to the target speed, the blade angle is adjusted to the target blade angle using the blade angle feedforward control. This blade angle or this blade angle adjustment rate is specified as a manipulated variable and passed to the blade angle control for implementation. How the blade angle can be determined for this purpose using the feedforward control is explained below.

[0075] According to one aspect, it is proposed that the blade angle feedforward control specifies the blade angle or the pitch rate as a control variable. This should preferably be model-based. When specifying the blade angle or the pitch rate, a model is used that simulates the behavior of the wind turbine. This blade angle or this pitch rate for the blade angle is passed on to the blade angle control system for implementation as a control variable. The blade angle control system then controls one or more actuators to adjust the respective rotor blade. In this way, all rotor blades are adjusted.

[0076] This means that the new operating point, namely the target operating point, can be reached quickly, as no longer too much power is generated or the speed has reached its cut-off value.

[0077] In particular, it is provided that the speed control specifies a blade angle or an adjustment rate of the blade angle as a function of a speed deviation as a manipulated variable and that the speed control is, at least temporarily, subordinate to the blade angle pre-control, so that the manipulated variable specified by the speed control is at least temporarily not passed on to the blade angle control or only with a weighting of less than 50%.

[0078] The speed control, which determines a blade angle or an adjustment rate of the blade angle depending on the speed deviation, can therefore also be provided outside of normal operation, i.e. outside of the first operating point, and thus also for controlling the target operating point. However, it is intended that the speed control is only subordinate to the feedforward control. In other words, the speed control is only used in a supplementary manner to control the target operating point. This can be achieved by temporarily, namely in particular for the majority of the time, i.e. at least 60% of the time, not applying the manipulated variable generated by the speed control, and thus the blade angle adjustment is carried out solely on the basis of the feedforward control during this time. In addition to or instead of this, it is also possible for the subordination of the speed control to the feedforward control to be achieved by weighting.For this purpose, the manipulated variable specified by the speed control can be weighted, i.e. multiplied by a value between 0 and 1. With a weighting of 50%, the manipulated variable is multiplied by 0.5.

[0079] In particular, it is proposed that the active control of the speed control and the blade angle feedforward control outputs the larger blade angle to be set or the larger blade angle to be set as the manipulated variable. The blade angle feedforward control is designed to adjust the blade angle as quickly as possible from the initial blade angle to the target blade angle. Thus, the feedforward control specifies a very large manipulated variable, which is usually larger than the manipulated variable that determines the speed control.

[0080] Especially towards the end, when the control variable specified by the feedforward control is smaller than the control variable specified by the speed control, a switch to speed control can be made. This is particularly useful when the blade angle has almost reached the target blade angle and it may then turn out that the target blade angle specified by the model does not exactly lead to the desired operation of the wind turbine at the target operating point.

[0081] If, at the target operating point reached—that is, at the target blade angle reached—more power is extracted from the wind than the electrical power generated by the turbine, i.e., the aerodynamic power is greater than expected, this leads to rotor acceleration. The speed control counteracts this, and therefore it is suggested that it take over blade adjustment toward the end.

[0082] If the blade angle feedforward control and the speed control operate in parallel, their two manipulated variables can be compared. If the manipulated variable of the speed control is larger in magnitude, the system switches to the speed control. Here, it was particularly recognized that the target operating point should be reached as quickly as possible, and this is usually achieved with large manipulated variables. Therefore, if the manipulated variable of the speed control is larger than that of the feedforward control, the speed control can replace the feedforward control and find the ideal operating point, which may differ from the precalculated target operating point.

[0083] According to one aspect, a temporal progression of the speed can be specified, thus specifying a temporal progression of a target speed. This can also be done model-based.

[0084] The blade angle precontrol, or a predetermined profile of the precontrol blade angle, is designed to adjust the blade angle, taking into account the reduced turbine power, so that this profile of the target speed is approximately achieved. Therefore, only a small deviation between the target speed and the actual speed is to be expected. As a result, even a speed control that aims to compensate for this small deviation only results in a small manipulated variable.

[0085] Only towards the end of the adjustment, i.e., near the target operating point and thus near the target blade angle, can larger deviations occur. This may be due to the fact that the pilot control manipulated variable becomes zero when the predetermined target blade angle is reached. However, if there are still slight deviations from the predetermined curve, the speed control will specify a manipulated variable that is not zero and thus greater than the manipulated variable of the blade angle pilot control. The speed control can then become active by being activated.

[0086] The speed curve can also be specified indirectly by specifying a speed change over time, i.e., a speed change rate. Depending on this speed change rate, a speed setpoint can be calculated at any time by multiplying it by the elapsed time.

[0087] According to one aspect, it is proposed that an aerodynamic braking variable be specified to change the rotational speed, wherein the aerodynamic braking variable is provided as rotor acceleration or aerodynamic power. Additionally or alternatively, it can be a negative acceleration or braking or deceleration. The braking variable can be embodied as a signal curve and therefore does not have to be a fixed value.

[0088] Aerodynamic power is the power that the rotor generates from the wind, i.e., the power it extracts from the wind. This aerodynamic power depends on the wind speed, the rotor speed, and the blade angle.

[0089] Furthermore, it is intended that the pilot blade angle is determined depending on the specified braking force in order to achieve the specified braking force. If the aerodynamic performance is specified as the braking force, the aerodynamic performance is piloted by specifying the blade angle so that the aerodynamic performance is achieved. The pilot-controlled blade angle ensures that the aerodynamic power is extracted from the wind.

[0090] Pre-controlling aerodynamic performance can therefore mean adjusting the rotor speed and blade angle while taking the prevailing wind into account. The rotor speed cannot be adjusted directly because, on the one hand, it changes slowly and, on the other hand, it is the result of the interaction between the rotor and the wind. However, this interaction can be influenced by the blade angle. In particular - to put it simply - given a given wind speed and rotor speed, the blade angle can be adjusted to achieve the desired aerodynamic performance. This is precisely what is done with pre-controlling aerodynamic performance. The rotor blade angle is adjusted to achieve the aerodynamic performance that is to be adjusted, i.e. pre-controlled.

[0091] The aerodynamic performance is controlled by specifying the blade angle via the blade angle feedforward control to achieve the desired aerodynamic performance. Thus, the blade angle is adjusted by the blade angle feedforward control and not by a control-related feedback. Therefore, the blade angle adjustment for the blade angle feedforward control must be planned in advance, either in anticipation of the actual rotor speed, or continuously adjusted based on a detected and / or estimated rotor speed.

[0092] In particular, it is proposed that, depending on the rotational speed, wind speed and specified aerodynamic performance, or alternatively a rotor acceleration, a blade angle is determined which leads to the specified aerodynamic performance.

[0093] The speed can be recorded and / or estimated, or the expected speed behavior can be predicted, which can be done through precalculation, simulation, or modulation. Based on this recorded / estimated or precalculated rotor speed, i.e., the precalculated rotor speed curve, the blade angle is adjusted accordingly to achieve the aerodynamic performance, or specified rotor acceleration. The blade angle essentially tracks the speed curve to always maintain the specified aerodynamic performance.

[0094] The blade angle can be determined based on a model, so that the corresponding blade angle is determined, or in particular calculated, using a model. Such a model simulates the behavior of the wind turbine, or at least takes it into account.

[0095] Because this is done via the feedforward control, any oscillations caused by control feedback are avoided.

[0096] According to one aspect, it is proposed that the presetting of the pilot blade angle be repeated continuously, or at least in short increments of less than one second. This always adjusts the pilot blade angle to the current operating point. In particular, the blade angle is intended to adjust the aerodynamic performance, which depends on the tip speed ratio and thus on the rotor speed. Therefore, the blade angle should be adjusted as closely as possible to the rotor speed. This can be measured, estimated, or predicted.

[0097] According to one aspect, it is proposed that in order to specify the pilot blade angle, in particular from the aerodynamic braking variable, a power coefficient to be set is determined, and depending on the power coefficient, the pilot blade angle is determined in order to achieve the power coefficient.

[0098] This power coefficient, which can also be referred to as the Cp value, indicates the relationship between the power extracted from the wind and the extractable power. This can be used to control the aerodynamic power, namely the extracted power, if the available power is known, which in turn can essentially be derived from the recorded wind speed. For an aerodynamic rotor of a wind turbine, there is usually a characteristic diagram that indicates a relationship between the tip-speed ratio, blade angle and power coefficient for this rotor. The characteristic diagram can be spanned as a curved surface in a coordinate system with these three quantities, by entering the corresponding Cp value for each pair of blade angle and tip-speed ratio, so that all of these Cp values form the curved surface.

[0099] Once the power coefficient to be set is determined, i.e., specified, the tip speed ratio can be calculated with a known wind speed and rotor speed. Using the power coefficient and the tip speed ratio, the corresponding blade angle, which must be set, can then be read off as the pilot blade angle.

[0100] According to one aspect, it is proposed that an ideal braking variable is specified to specify the aerodynamic braking variable, and the ideal braking variable is filtered by means of a filter to obtain the aerodynamic braking variable. The filter filters at least one natural vibration frequency of the wind turbine from the specified ideal braking variable, which is in particular a natural frequency of the tower, in particular the first natural frequency of the tower, to obtain the aerodynamic braking variable. The filter is particularly designed as a band-stop filter and / or notch filter.

[0101] The ideal braking value can thus be an ideal aerodynamic power or an ideal acceleration, namely negative acceleration, to decelerate the rotor, specifically as quickly as possible. Using aerodynamic power as an example, this can particularly be a jump in aerodynamic power from a maximum value, e.g., the rated power of the wind turbine, to a minimum value, e.g., the negative rated power of the wind turbine. Even in the case of a specified rotor acceleration, this can be a jump to a negative value.

[0102] However, it was recognized that such a jump can contain a frequency component with an excitation frequency particularly suitable for tower vibration. Therefore, it is proposed to filter out this frequency component from the ideal braking variable, or at least to reduce this frequency component. The result after filtering is the (no longer ideal) aerodynamic braking variable, which can then be used to determine the pilot blade angle in the manner already described.

[0103] To filter out, or at least reduce, the frequency component, a band-stop filter or notch filter is proposed, which is suitable for specifically filtering out a frequency component or a narrow frequency range. This allows a large part of the specified ideal braking value to be retained. The rotor speed can therefore still be reduced very quickly, but without exciting the relevant vibration, in particular without exciting the tower vibration.

[0104] In particular, it is proposed that the determination of the blade angle is carried out in a model-based manner and / or based on predetermined relationships between these three variables, depending on the rotational speed, wind speed and specified aerodynamic performance, in order to thereby adjust the aerodynamic performance.

[0105] Here, it was particularly recognized that oscillations can be avoided by avoiding control feedback. Furthermore, a particularly rapid adjustment of the initial operating point to the target operating point is possible because the feedforward control allows the adjustment possibilities to be fully exploited, especially when considering manipulated variable limitations.

[0106] Due to the lack of control feedback, the respective current operating point, as determined by rotor speed and blade angle, is calculated in advance. It is particularly important here that a corresponding course of this operating point is calculated in advance, since blade angle and rotor speed will change continuously from the first operating point to the target operating point. The prediction may contain inaccuracies that could be at least partially corrected by control. However, such control is largely avoided in order to avoid the disadvantages of control mentioned. Minor inaccuracies are accepted. However, it has been recognized that such inaccuracies are more acceptable than the disadvantages mentioned for control.

[0107] This prediction, in particular the prediction of the relationships between blade angle speed, wind speed, and specified aerodynamic performance, can be predetermined, i.e., precalculated. This precalculation can be model-based. Such a model-based precalculation means that a model of the wind turbine for the specified parameters is available, and this expected behavior is modeled or simulated in order to determine the appropriate value for the blade angle to be pre-controlled. Such a predetermination does not have to be done offline and before commissioning; it can also be performed during operation if the respective conditions are known, namely the current wind speed and the current curtailment requirement, i.e., in particular, the specified reduced speed.

[0108] Additionally or alternatively, it is proposed that the blade angle feedforward control specify the blade angle or the blade angle adjustment rate as a manipulated variable in order to pre-control the aerodynamic power, and that the speed control is, at least temporarily, subordinate to the blade angle feedforward control. Thus, the blade angle feedforward control and the speed control both output a blade angle or an adjustment rate, but these are not both used simultaneously to adjust the rotor blade. Rather, the adjustment is predominantly carried out by the value output by the blade angle feedforward control. However, particularly towards the end, the speed control can take over, or a weighted consideration can be given, as already explained above in connection with the specification of reduced power as a curtailment requirement.

[0109] In particular, it is proposed that of the speed control and the blade angle feedforward control, the one that outputs the larger manipulated variable in terms of magnitude, i.e. outputs the larger blade angle to be set or the larger adjustment rate to be set, be the active one. The effect of this has already been explained above, namely that the blade angle feedforward control usually outputs large manipulated variables because it is designed to get from the first operating point to the target operating point as quickly as possible. As soon as the target operating point has almost been reached, or from the feedforward control perspective has been almost or completely reached, the manipulated variable generated by the blade angle feedforward control decreases, ideally to zero. The manipulated variable of the blade angle feedforward control then falls below the manipulated variable of the speed control, so that the speed control takes over control of the operating point again, particularly towards the end, i.e. near the target operating point.

[0110] In particular, it is assumed that the change process from the initial operating point to the target operating point is so rapid that a substantially constant wind speed can be assumed for this process. In reality, the wind speed will of course not remain exactly constant, and the resulting slight deviations can also be compensated for by the speed control near the target operating point.

[0111] According to one aspect, it is proposed that the speed control outputs a target blade angle or a target adjustment rate as a function of a speed control deviation as the difference between the actual speed and a target speed. This target blade angle or this target adjustment rate can be passed to the blade angle control. It is further proposed that the speed control be changed or suspended, while the blade angle feedforward control directly specifies the blade angle or adjustment rate to be set. In this case, it is particularly important that the speed control is not activated during this time, or is only activated in a modified form. The algorithm underlying the speed control can continue to be executed unchanged, and an actual speed value can also be fed back, since this is already recorded and available in the overall operational management of the wind turbine.While the blade angle pre-control is active, the speed control should not be active.

[0112] If the speed control is changed, this can mean that it is activated via a weighting, in particular by a multiplication of a value less than 0.5, and / or that its parameters are changed. In particular, the parameters can be changed by reducing the controller gain or increasing a time constant of the speed control, so that the speed control does not work as hard or as quickly as usual. This could allow the speed control to still contribute to compensating for small deviations, but would prevent it from changing the blade angle feedforward control too drastically.

[0113] According to one aspect, it is proposed that a temporal profile of a target speed be specified to control the target operating point of the speed control and / or a temporal profile of a target power be specified to control the target operating point of the power control. This allows the progression from the first operating point to the target operating point to be controlled more precisely by guiding the speed profile or the power profile along the corresponding profiles.

[0114] According to one aspect, it is proposed that while the blade angle pre-control directly specifies the blade angle or adjustment rate to be set, the speed control is modified in its parameterization, in particular in at least one gain factor. In particular, such a gain factor is reduced in order to reduce the dominance of the speed control over the blade angle pre-control. Here, it was particularly recognized that when adjusting the first operating point to the target operating point, this adjustment process is paramount and should be implemented with the described advantages by the blade angle pre-control.

[0115] According to one aspect, it is proposed that a blade angle change from the first blade angle to the target blade angle can be described as a relative blade angle change from 0 to 100%, where 0% corresponds to the first blade angle and 100% to the target blade angle. The relative blade angle change can also be synonymously referred to as the relative blade angle range. For example, if the first blade angle is 10° and the target blade angle is 60°, to choose just one illustrative example, 0% corresponds to 10° and 100% to 60°. A change of one degree would then correspond to exactly 2% in this example.

[0116] It is further proposed that the blade angle adjustment from the initial blade angle to the target blade angle be aborted and / or taken over by the speed control using the blade angle feedforward control before the target blade angle is reached. Thus, the blade angle feedforward control is not fully active, but is replaced by the speed control before the target blade angle is reached.

[0117] In particular, it is proposed that the blade angle is adjusted using the blade angle feedforward control for at least 5% to 90%, in particular for 0% to 95%, of the blade angle change and / or that the speed control is changed or deactivated during this time. To stick with the example of an adjustment from 10° to 60°, the blade angle is adjusted using the blade angle feedforward control, i.e. for at least 12.5° to 55°, or from 10° to 57.5°. During this time, i.e. in this range specified by the percentage values, the speed control can be inactive and / or changed. Here, too, if it is inactive, its manipulated variable is not applied. If it is changed, its manipulated variable is applied at a lower level and / or the speed control parameters are changed.

[0118] According to one aspect, it is proposed that in order to control the target operating point, an aerodynamic braking power is determined with which the rotor is to be braked by the wind, and the temporal blade angle is determined by the blade angle feedforward control such that the rotor transfers the aerodynamic braking power to the wind, wherein in particular the aerodynamic braking power is selected in a range from 10% to 120% of the rated power of the wind turbine, in particular in the range from 50% to 100% of the rated power. Thus, it is particularly proposed that not only the aerodynamic power be controlled to zero, but even to a negative value, because now power is no longer absorbed from the wind, but is actually transferred to it. The rotor blades are therefore not only adjusted to the extent that they no longer absorb power from the wind, but their blade angle is adjusted even further so that they actually transfer power to the wind.The aerodynamic braking performance therefore corresponds to a negative aerodynamic performance.

[0119] The power delivered to the wind, and thus the aerodynamic braking power, can range from 10% to 120% of the wind turbine's rated power. It was particularly recognized that 10% aerodynamic braking power is a value that can make a noticeable contribution. The braking power is specifically designed up to 100% of the rated power, as the wind turbine is designed for such values, including the resulting mechanical loads. Since adjusting to the target operating point should be as short as possible, it was recognized that the aerodynamic braking power can even exceed the wind turbine's rated power, in particular up to 120% of it. A value of at least 50% is proposed so that the aerodynamic braking power can make a significant contribution to rotor braking.

[0120] According to one aspect, it is proposed that a wind speed is estimated to determine the target operating point, namely based on the rotational speed, the power and the first blade angle, i.e. the blade angle of the first operating point. The idea behind this is that the first operating point at which the wind turbine is located when the curtailment request is specified is a stationary operating point, i.e. in which the wind turbine is in steady-state operation. The values have therefore all stabilized and the wind speed can therefore be deduced from them. This is also based on the assumption that the wind speed will not change, or will not change significantly, until the target operating point is reached, so that the wind speed can be determined from the wind speed of the first operating point, which is assumed to also be the basis for the target operating point.The target operating point and all the calculations described therein, in particular model-based calculations and / or predeterminations or precalculations, can therefore be based on such a wind speed.

[0121] According to one aspect, it is proposed that the speed and / or the power be greater than zero at the target operating point. In particular, the presented method, at least according to a preferred embodiment, relates to the case where the curtailment request neither specifies the value zero for the power nor specifies the value zero for the rotor speed, but rather the wind turbine is still operated at the target operating point, at least at a speed other than zero and a power other than zero.

[0122] According to one aspect, it is proposed that a desired reduction time be specified, within which the target operating point is to be reached, and that the temporal blade angle is specified as a function of the desired reduction time, in particular using a model or the model underlying the previously described model-based determinations or calculations. In particular, the temporal blade angle profile can be determined by specifying the rotor acceleration, i.e., the rotor deceleration, as a function of the desired reduction time, or the rotor acceleration can be calculated directly therefrom.

[0123] According to one aspect, it is proposed that after the blade angle adjustment by means of the blade angle pre-control is completed, the speed control regulates or limits the speed to a target speed. If a reduced speed is specified by the reduction request, this is the target speed to which the speed control regulates.

[0124] If a reduced system output is specified by the curtailment request, the target speed can be determined depending on this specified reduced output and / or the wind speed can be measured or observed and, depending on this, the target speed can be determined in the case of a specified reduced system output.

[0125] According to one aspect, it is proposed that when a throttle-down request is present, a check is carried out to determine whether the target operating point is sufficiently far from the first operating point. To this end, it is further proposed that the blade angle adjustment be performed using the blade angle pre-control only if the distance from the target operating point is sufficient. A sufficient distance exists if the distance is above a predeterminable reference distance.

[0126] This ensures that the proposed blade angle feedforward control for changing the first operating point to the target operating point is only used if this promises an advantage or sufficient advantage over the variant of simply changing the first operating point to the target operating point by specifying the changed setpoints, in particular the changed speed setpoint, for the speed control. Speed control, in particular, is already suitable for making the change from the first operating point to the target operating point by specifying a new speed setpoint, which can correspond to the speed of the target operating point. However, it has been recognized that using blade angle feedforward control is advantageous and should be used for a significant adjustment; for a small adjustment, i.e. when the first operating point and the target operating point are close to one another, using blade angle feedforward control may be unnecessary.

[0127] In particular, it is proposed that a difference between the current speed and the target speed be considered as the distance and is considered sufficient if it is greater than a predefinable minimum speed difference, and / or a difference between the current turbine power and the target power be considered as the distance and is assessed as sufficient if this distance is greater than a predefinable minimum power difference. In addition or alternatively, it is proposed that a difference between the first blade angle and the target blade angle be considered as the distance and is assessed as sufficient if it is greater than a predefinable minimum angle difference. Preferably, the minimum angle difference is at least 10°, in particular at least 20°. The predefinable minimum power difference is in particular at least 20% of the rated power of the wind turbine, in particular at least 30% of the rated power of the wind turbine.The minimum speed difference is preferably at least 20% of the rated speed and in particular at least 40% of the rated speed.

[0128] According to one aspect, it is proposed that an approach of a flying animal, in particular a bird or a bat, leads to a speed requirement as a reduction requirement and the speed is determined depending on one criterion, several criteria or all criteria from the list, i.e. criterion list with the elements: a position of the flying animal, a type of flying animal, a speed of movement of the flying animal and a direction of movement of the flying animal.

[0129] In this case, it was particularly recognized that, to protect the flying animal, a reduction in rotor speed as quickly as possible may be necessary. Such a rapid reduction in rotor speed can be achieved through rotor blade pre-control and, if necessary, through other measures to reduce rotor speed as described above.

[0130] Additionally, it was recognized that it is important to not only check whether such an endangered animal is approaching, but also how it is doing so or how it is expected to do so. The position of the flying animal can be used to determine how close it is to the wind turbine and therefore how urgently it is necessary to reduce the speed, and thus how much it needs to be reduced. For this purpose, it is proposed to determine the target speed based on the position of the flying animal.

[0131] The type of flying animal can be used to determine how and how fast it is approaching the wind turbine. The type of flying animal can also be used to determine whether or not a change in its flight path is to be expected. This can also influence whether and how quickly the flying animal is approaching the wind turbine and thus a danger zone of the wind turbine, so that the target speed can be specified depending on the type of flying animal. The type of flying animal can specifically indicate whether it is a bat or a bird. In addition, the type of flying animal can be used to distinguish between different bat species and / or different bird species. The type of flying animal can be used to determine its behavior. A maximum and / or typical flight speed can be derived, and from this, it can be more accurately calculated how quickly it will approach the wind turbine.From the species it is also possible to determine whether a steady or discontinuous trajectory is to be expected.

[0132] From the movement speed of the flying animal, if this is also recorded, it can also be deduced how quickly this animal is approaching the wind turbine, and the target speed can be determined accordingly. It is therefore proposed to determine the target speed as a function of the movement speed of the flying animal.

[0133] The direction of movement of the flying animal can provide information about whether the flying animal is approaching the wind turbine and thus a danger zone, or whether it can be expected to fly past the wind turbine. Therefore, it is proposed to specify the target speed depending on the direction of movement of the flying animal.

[0134] Additionally or alternatively, it is proposed to continue monitoring the behavior of the flying animal and to specify a new target speed depending on the observed behavior of the flying animal. It was particularly recognized here that when a flying animal to be protected is detected approaching the wind turbine, the speed does not have to be immediately reduced to such a low value that any danger to the flying animal is excluded. Such a speed, which can be referred to as a safety speed and which can have very low values, such as 10% of the nominal speed or less, should be present when the flying animal is in the vicinity of the rotor of the wind turbine. This should ensure that the rotor speed can be reduced to this safety speed before the flying animal reaches the rotor area.

[0135] For this purpose, the rotor speed can be reduced slightly if a flying animal that needs to be protected approaches the wind turbine but is still far away. The rotor speed should then be reduced to a value that allows the rotor speed to be reduced to the safety speed if the flying animal continues to fly towards the wind turbine. However, if the flying animal does not continue its flight towards the wind turbine, the rotor speed does not need to be reduced further.

[0136] Put simply, the closer the flying animal comes to the wind turbine, the lower the rotor speed can be selected. The closer the flying animal comes to the wind turbine, the shorter the time it would take to reach the wind turbine. Accordingly, the time available to reduce the rotor speed to the safety speed is shorter. Accordingly, a speed can be selected as the target speed that is low enough to allow for a timely reduction to the safety speed, should this be necessary due to the subsequent behavior of the flying animal.

[0137] Different target speeds can be specified in stages depending on the subsequent behavior of the flying animal, in particular depending on the distance of the flying animal from the wind turbine. For example, in a first stage, when the flying animal approaches the wind turbine, the target speed can be set to 80% of the nominal speed. In a second stage, when the flying animal approaches the wind turbine further, the target speed can be set to 50% of the nominal speed, to give an example. If the flying animal then approaches the wind turbine even further, the target speed can be set to a further stage, for example, 30% of the nominal speed, or to the safety speed.

[0138] According to the invention, a wind turbine is also proposed. This wind turbine is connected to an electrical grid, has a rotor with adjustable blade angles, is operable at a variable speed, and is configured to generate power from wind. It has a blade angle control for adjusting the blade angles, a speed control for regulating or limiting the speed, and a power control for limiting the turbine power.

[0139] The wind turbine is operable at a predeterminable operating point, wherein the operating point is characterized at least by the rotational speed and the turbine power, preferably additionally by the blade angle. The wind turbine is configured to execute a method according to at least one aspect described above. In particular, the wind turbine has a turbine controller in which such a method is implemented. The turbine controller can thus have a process computer in which the method is stored by means of a program code.

[0140] The invention will now be explained in more detail below with reference to the accompanying figures. Figure 1 shows a wind turbine in a perspective view. Figure 2 shows a schematic of a control structure with an indicated wind turbine. Figure 3shows a flow chart when a reduced speed is specified as a reduction request. Figure 4 shows a flow chart when a reduced power is specified as a curtailment request. Figure 5 shows a structure for calculating the pilot blade angle. Figure 6 shows a diagram with a filtered jump size. Figure 7 shows simulation results of an applied procedure.

[0141] Figure 1 shows a wind turbine 100 with a tower 102 and a nacelle 104. A rotor 106 with three rotor blades 108 and a spinner 110 is arranged on the nacelle 104. During operation, the rotor 106 is set into rotation by the wind and thereby drives a generator in the nacelle 104.

[0142] The pitch angles, i.e. blade angles of the rotor blades 108, can be changed by pitch motors at the rotor blade roots 109 of the respective rotor blades 108.

[0143] The wind turbine 100 has an electrical generator 101, which is indicated in the nacelle 104. Electrical power can be generated by means of the generator 101. For feeding in electrical power, a feed-in unit 105 is provided, which can be designed particularly as an inverter. This can generate a three-phase feed-in current and / or a three-phase feed-in voltage according to amplitude, frequency, and phase for feeding into a grid connection point PCC. This can be done directly or jointly with other wind turbines in a wind farm. A system controller 103 is provided to control the wind turbine 100 and the feed-in unit 105. The system controller 103 can also receive default values from external sources, in particular from a central farm computer.

[0144] Figure 2shows a control structure 200, which in particular has a speed control 201 with controller block 202 and a blade angle feedforward control with modeling block 204. During regular operation, when no change from a first operating point to a target operating point is planned, the speed control 201 is active. It operates in such a way that a difference between the setpoint speed n S and the actual speed ni is formed in the first summing point 206. This speed difference thus forms the control error e for the speed, which is passed to the controller block 202. Here and throughout the description, the speed is referred to simply and thus synonymously as the rotor speed.

[0145] The controller block 202 converts the control error e into a manipulated variable, namely a blade angle setpoint α S . This blade angle setpoint α S is then passed on directly unchanged via the selection block 208. The meaning of the selection block 208 will be described later. In normal operation, if in the structure of the Figure 2 If only the speed control 201 is active, the selection block 208 has no significance and can be regarded as simply forwarding the blade angle setpoint α S from the controller block 202 to the blade angle control 210. The blade angle control 210 is only indicated schematically here and it performs the adjustment of the blade angle, i.e. the blade angle of all rotor blades. For this purpose, it can in particular control corresponding pitch motors, which is why it is included in the control structure 200 of the Figure 2also referred to as "pitch". The fact that the blade angle control 210 thus acts on the rotor blades is indicated by the fact that its output leads to the schematically illustrated wind turbine 100. The wind turbine 100 can, like the wind turbine 100 of the Figure 1 be trained.

[0146] By controlling the wind turbine 100 and due to the wind, a speed is set which is recorded as the actual speed ni and fed back to the speed control 201, namely to the first summing point 206.

[0147] In order to change the operation of the wind turbine from a first operating point to a target operating point after receiving a curtailment request, a blade angle feedforward control 212 is provided. An important element of the blade angle feedforward control is the modeling block 204. The modeling block 204 receives, in particular, from the target specification block 214 the parameters that characterize the target operating point to be controlled, namely the target turbine power PZ , the target speed n Z and the target blade angle α Z . The target specification block 214 receives the curtailment request for this purpose. This can thus be a reduced turbine power PR or a reduced rotor speed n R. Both are in Figure 2 as input variables for the target specification block 214, but in reality usually or exclusively only one of the two variables will be specified.

[0148] The modeling block 204 then determines the blade angle α V to be pre-controlled from the input variables it has received. For this purpose, it can receive further information which is contained in the control structure of the Figure 2 are not shown. This includes the current wind speed, which it can receive or determine, in particular estimate, from other information. This other information can be information about the first operating point from which the start is made, namely turbine power, rotor speed, and blade angle of the first operating point, i.e., the starting operating point.

[0149] In addition, the modeling block 204 can receive a rotor acceleration as an input variable, or calculate it. It can also receive an aerodynamic power to be adjusted as an input variable, or calculate it.

[0150] In any case, the modeling block 204 can contain a model of the wind turbine, i.e. a model of the behavior of the wind turbine, and based thereon predetermine a suitable turbine behavior and derive the pilot blade angle α V.

[0151] One possible implementation in modeling block 204 is to store one or more tables. Such a table can, for example, contain suitable pilot blade angles or suitable profiles for the pilot blade angle for different wind speeds and different target operating points. The first operating point, as the starting operating point, can be assumed, in particular, to be one that results at the respective wind speed when the wind turbine is not yet subject to a curtailment requirement.

[0152] In any case, the modeling block 204 outputs the pilot blade angle α V , which is to be passed to the blade angle control 210 for implementation. The modeling block 204 can, in particular, continuously output a blade angle that is only controlled, i.e., not changed by any feedback, and certainly not by speed deviations. In one case, the blade angle pilot control 212 can operate in such a way that only values from the first operating point, i.e., the starting operating point, are taken into account as recorded values, and further values are predetermined and / or a profile of the pilot blade angle is predetermined. The blade angle pilot control 212 can also operate in such a way that it constantly adjusts the pilot blade angle based on the current operating point of the system, i.e., constantly recalculates the pilot angle for new operating points.

[0153] In addition, however, the selection block 208 is provided, which decides whether the target blade angle α S as the output of the controller block 202 or the pilot blade angle α V as the output of the modeling block 204 is passed to the blade angle control 210 as the blade angle α to be set. The selection block 208 is thus symbolized as a switching block that can switch between the speed control 201 and the blade angle pilot control 212. The switching criterion is considered to be which of the two manipulated variables, the target blade angle α S of the speed control 201 and the pilot blade angle α V of the blade angle pilot control 212, is greater in magnitude. For the sake of simplicity, the structure of the Figure 2 this amount formation is not shown.

[0154] To form this magnitude comparison, the second summing point 216 is provided. The difference between the values between the target blade angle α S and the pilot blade angle α V is thus formed at this second summing point 216. Depending on this difference, namely whether it is positive or negative, the selection block 208 then switches to the target blade angle α S or the pilot blade angle α V .

[0155] What is achieved here is that the speed continues to run normally, but is inactive, as long as the blade angle precontrol 212 is dominant, i.e., as long as its precontrol blade angle α V is greater than the target blade angle α S . This is the case when the difference output by the second summing point 216 is negative. However, if this is no longer the case, the speed control can take over again, and this will happen especially when the target operating point is almost reached.

[0156] It should be noted that the specification of a target blade angle α S or a pilot blade angle α V is illustrated in Figure 2 is shown. Alternatively, an adjustment rate for the blade angle can also be output instead of a blade angle. In this case, the controller block 202 outputs a target adjustment rate for the blade angle instead of the target blade angle α S . Likewise, the modeling block 204 then outputs a pilot control adjustment rate for the blade angle instead of the pilot control blade angle α V . This variant can be preferred because it results in the speed controller 201 having an integrating behavior that achieves steady-state accuracy for the speed to be regulated without the need for an integrator in the controller block 202. In particular, with this variant, the speed controller 201 can continue to run continuously, even if the manipulated variable it generates is temporarily not used.

[0157] The flow chart 300 of the Figure 3 Explains a process for the case where a reduced speed, i.e., a speed request, is specified as the curtailment request. Then, in start step 302, the reduced speed n R is specified for the speed n. This is passed on to the target determination step 304. In the target determination step 304, the target operating point is determined, and for this purpose, the values for the target blade angle α Z , the target rotor speed n Z , and the target turbine power P Z are specified. The target rotor speed n Z then corresponds to the specified reduced speed n R .

[0158] With these specified values, the process continues in the pilot control determination step 306. In the pilot control determination step 306, a rotor acceleration, symbolized there as dn / dt, is specified or calculated. Alternatively, an aerodynamic power P ae is specified or calculated. Based on this, the pilot control blade angle α V is determined, which can be specified as an absolute value or as a curve, or instead, an adjustment rate can be specified as the pilot control adjustment rate. The pilot control blade angle, its curve, or the pilot control adjustment rate are specified such that the aerodynamic power P ae is achieved. In all of this, a system power PG is also taken into account or determined in the pilot control determination step 306.

[0159] Preferably, the turbine power or, alternatively, the generator torque is set as high as possible until the target operating point has been reached to support aerodynamic braking. The turbine power or the generator torque affects the rotor speed, thus affecting the change in the operating point and can therefore influence the determination of the pilot blade angle if it is constantly being recalculated depending on the current operating point. However, such behavior can also be pre-calculated entirely, e.g., through a simulation. Then the pilot blade angle can be adjusted or tracked without constantly recording the current operating point.

[0160] The pre-control determination step 306 can in particular be carried out in a model-based manner, in that the mentioned variables are determined based on a model or a simulation, or have been previously determined and are stored in a memory for use in the pre-control determination step 306.

[0161] The sequence then continues to pre-control step 308. In pre-control step 308, the pre-control is executed. The pre-control blade angle, or its profile or the pre-control adjustment rate, are passed on to the blade angle control without any control feedback, in particular without taking any speed deviation into account, thus only as a control. This is indicated by the symbol in the block of pre-control step 308. At the same time, the system power PG, which was also taken into account or determined in the pre-control determination step 306, is converted. For this purpose, it can be passed as a setpoint to a corresponding frequency converter and / or active rectifier or other control of the generator. The system power PG is designated here with the index G to indicate that it acts on the generator and differs from the aerodynamic power P ae.

[0162] The pilot control determination step 306 and the pilot control step 308 can be repeated continuously in order to redetermine or adjust the pilot control angle continuously or in short time steps and then also to apply it, which is indicated by a corresponding repeat loop 309.

[0163] While the pilot blade angle α V , or its curve or the pilot adjustment rate, is being piloted in pilot control step 308, speed control is inactive. However, near the target operating point, the pilot control can transition to speed control, and for this purpose the flow chart 300 transitions from the pilot control step 308 to the speed control step 310. In the speed control step 310, the speed control then takes over the adjustment of the blade angle again and then a corresponding feedback takes place again, namely the feedback of the actual speed. Here, only small deviations from the target operating point exist and these can best be corrected by the speed control. The target operating point is then reached and the process ends.

[0164] Figure 4shows a flowchart 400 in which a power limitation is specified as the curtailment request. This is symbolized in the start step 402 by specifying a reduced power PR for the power P.

[0165] The method then proceeds to target determination step 404, in which the target operating point is defined and the values for the target blade angle α Z , the target speed n Z , and the target power PZ are determined. If a speed reduction is specified as a curtailment request, the target turbine power PZ thus corresponds to the reduced power PR .

[0166] In this process, when a reduced power is specified as a curtailment request, the system power is reduced, in particular immediately or as quickly as possible, which is carried out in the power reduction step 405. The system power is therefore reduced to the reduced power, in particular as far as or as quickly as this is technically feasible. However, such technical implementation can be carried out in a period of less than one second or at least a few seconds, such as a maximum of 5 seconds. In particular, this can be achieved by the inverters feeding into the electrical supply grid immediately reducing their power to the specified power. However, if the curtailment request allows for a slower power reduction, in particular in a range of a maximum of 5 seconds or a maximum of 3 seconds, the power can be reduced accordingly more slowly, e.g.via a corresponding linear regulation, i.e. via a temporal ramp function.

[0167] Such a power reduction request can often be a requirement of a grid operator and requires correspondingly rapid action. In this case, the generator may not be able to reduce its output as quickly as the power fed into the grid is reduced. The resulting short-term power surplus can be consumed in so-called chopper resistors. It may also be the case, or alternatively, that the power generated by the generator is reduced faster than the aerodynamic power, which can lead to rotor acceleration.

[0168] In any case, an attempt should be made to adjust the blade angle to the target blade angle as quickly as possible, and for this purpose, at least the pilot control blade angle α V is determined in the pilot control determination step 406. The determination can be carried out in the same way as in the pilot control determination step 306 of the Figure 3 described, whereby, however, the system output has the value of the reduced output from the outset or according to the reduction carried out, which was carried out, for example, using the explained time ramp function.

[0169] According to schedule 400 of the Figure 4 The process continues in pilot control step 408. In pilot control step 408, the pilot control blade angle or its profile or a pilot control adjustment rate is thus sent directly to the blade angle control for implementation there.

[0170] How to Figure 3As explained, the pilot control determination step 406 and the pilot control step 408 can also be repeated continuously in order to redetermine or adjust the pilot control angle constantly, i.e. continuously or in short time steps, and then also apply it, which is indicated by a corresponding repeat loop 409.

[0171] Here too, it is proposed that when the target operating point is reached or shortly before, the process transitions from the pre-control step 408 to the speed control step 410. Thus, here too, the speed control can replace the pre-control when the target operating point is reached or almost reached. The process for changing from the first operating point to the target operating point is thus completed and the speed control can, which also applies to the process of Figure 3 applies, continue the operation of the wind turbine at this new target operating point.

[0172] Figure 5shows a structure 500 for determining the pilot blade angle α V according to one embodiment. The pilot blade angle α V can be determined in the characteristic map block 502 from the current tip speed ratio λ and a predetermined target value for the power coefficient CP,S. For this purpose, a characteristic map can be stored that specifies a relationship between the tip speed ratio λ, the power coefficient CP, and the blade angle α for the rotor used.

[0173] The tip speed ratio λ is derived from the current rotor speed n, which is divided by the current wind speed V w in the quotient block 504. The speed is known from the turbine control system, and the wind speed could be measured. However, here it is proposed to estimate the wind speed V w, which is indicated by the estimation block 506. This allows the wind speed to be measured with high accuracy and low noise.

[0174] The wind estimation is based on the recorded rotor speed n, the current aerodynamic power P ae , i.e. the power that is currently extracted from the wind, and the currently set blade angle α, which is known from the turbine control system that controls the wind turbine as a whole.

[0175] The aerodynamic power P ae can be calculated from the currently generated electrical power P el and the moment of inertia J of the rotor, taking into account the rotor speed n and its change, according to the formula: P ae =P el +J*n*dn / dt.

[0176] The wind estimator 506 can determine the wind speed from these three variables using a state observer or from empirical values that can be stored in a table.

[0177] The setpoint of the power coefficient CP,S , i.e., the desired and thus specified power coefficient, is determined from the current wind speed V w and a specified aerodynamic power P ae,s . This is symbolized by the CP block 508. In the CP block, a quotient of the aerodynamic power P ae and a wind power P w assigned to the wind speed V w can be calculated. The existing wind power P w in the rotor area A of the rotor can be calculated from the wind speed and the air density ρ. This can be converted into a calculation of the specified power coefficient CP and leads to the formula: C P = P ae * 2 / A * ρ * V w 3

[0178] It is proposed to select the aerodynamic power P ae such that only realistic CP values can be achieved. However, the wind turbine is usually located at operating points where unrealistic CP values are not expected. Ultimately, the aerodynamic power should be reduced compared to the current operating point so that an excessively high CP value is not expected.

[0179] If a power target, i.e., a power reduction, is specified as a curtailment request, the setpoint of the aerodynamic power P ae,s can be set to the value of the power target, i.e., the system power to be curtailed, which can correspond to the target system power. It is also possible to initially specify a lower aerodynamic power, e.g., with the value zero.

[0180] However, if the wind turbine is to be braked very quickly, i.e., if the curtailment requirement is a speed specification, i.e., a speed reduction, particularly to ensure bird or bat protection, a very high aerodynamic power can be specified.

[0181] In order not to overload the wind turbine, it is proposed to set the aerodynamic power P ae,s to the negative nominal power PN of the wind turbine.

[0182] This is illustrated by the activation block 510. The activation block 510 thus switches from a previous value, which could have been a maximum of +PN, to -PN and outputs this as the ideal braking variable, in the embodiment shown as the ideal aerodynamic power P ae,i . This ideal value is therefore a step value, but one that can lead to the excitation of the first tower vibration, so it is suggested that it be first passed through the filter function 512. The filter function 512 is designed as a band-stop filter and filters out a frequency component that could excite a first tower natural frequency. The result is the filtered setpoint for the aerodynamic power P ae,s .

[0183] In such a bird protection application, which can also be a bird protection stop of the rotor, a blade angle is requested that aerodynamically brakes the wind turbine with negative nominal power. The first tower natural frequency is filtered out of this erratic signal to prevent tower excitation.

[0184] The effect of specifying the ideal aerodynamic performance by the activation block 510, or generally an ideal braking quantity, and filtering with the filter function 512 is shown in the Figure 6 illustrated.

[0185] Figure 6 shows a time diagram in which the electrical power P el , the desired aerodynamic power P ae,s , and the ideal aerodynamic power P ae;i are represented. The curtailment request occurs at time t 0 . Until then, the electrical power P el and the aerodynamic power P ae are balanced. The slight differences between the curves are only due to the graphical representation.

[0186] At time t0, the ideal aerodynamic power P ae,i is then entered abruptly, changing from PN to -PN. However, this negative jump is modified by filtering, resulting in a modified curve for the specified aerodynamic power P ae, whose curve is to be understood more symbolically. However, after some time, it should asymptotically approach the final value of the ideal function curve. The aerodynamic power P ae is thus not attenuated; its curve is only initially altered to avoid exciting a tower vibration.

[0187] Figure 7 shows the simulation of a curtailment process near nominal wind speed. The index "0" represents the response of a turbine that uses only conventional feedback control, and the index "inv" represents the response of a turbine that responds to the curtailment request using the described feedforward control. The top sub-diagram shows the power feed-in curve P normalized to nominal power. It is the same for both cases. The second sub-diagram shows the blade angle curve, which can also be referred to as the pitch angle curve. The curve α inv shows a significantly earlier response than the curve α 0 , which underlines the advantages of feedforward control. The third sub-diagram shows the rotor speed curve resulting from the power reduction and pitch angle curve, normalized to a nominal speed. It is clear that the speed increase is reduced by the earlier pitch-out at n inv compared to the curve n 0 . The fourth sub-diagram shows the tower base load normalized to a maximum torque as torque M 0 or M inv . It is clear that no significant overshoot occurs due to the early pitch-out.

[0188] Further aspects, considerations and effects of the invention are described below.

[0189] It was recognized that there are many applications for rapid speed or power reductions, especially aerodynamic power reduction with minimal mechanical stress.

[0190] For example, bird detection systems can be used to promptly stop or reduce turbine speed in order to contribute to species conservation. However, this approach reduces the yield when bird contact is detected, or can even lead to unacceptable mechanical stress on the wind turbine if the stops are too frequent, too severe, or too rapid.

[0191] Another example is the rapid or even sudden reduction of the maximum electrical power, for example, by a grid operator. While the electrical system can reduce the feed-in power quickly, usually in less than a second, the reduction of the aerodynamic power takes place more slowly, due to the movement of the blades. If this process is too slow , an overspeed may occur, which can result in high mechanical stress on the wind turbine. On the other hand, too fast Moving the rotor blades has its dangers. It is widely known that this can reverse the rotor thrust, causing the nacelle to swing forward an unacceptable amount. This also leads to unacceptable mechanical stress and is further exacerbated by the rapid power reduction.

[0192] What all applications have in common is that the aerodynamic power absorbed by the rotor blades is reduced as quickly as possible, while keeping the load on the wind turbine as low as possible. This applies to both operational and extreme loads.

[0193] The reason for reducing aerodynamic performance depends on the application. Bird protection, which can also be more generally referred to as species protection, involves regulating the speed below a threshold that poses no danger to birds or bats. In this case, the generator can be used additionally by simultaneously increasing the generator torque or generator power as much as possible to reduce aerodynamic performance.

[0194] The primary purpose of maximum power limitations is to adapt the aerodynamic performance to the remaining feed-in power of the generator as quickly as possible.

[0195] The main aim is therefore to reduce the rotor speed of a wind turbine in a way that protects the load, without implementing emergency stop situations.

[0196] The invention improves, among other things, the speed reduction time for bird protection events.

[0197] The invention avoids increased loads, which can be caused particularly by overspeed or delayed or uncoordinated control procedures, during maximum power reductions.

[0198] Previously, the turbine was shut down slowly during bird protection events. While this reduces load, it takes more time and requires bird detection within a larger radius around the turbine. It also leads to premature shutdown of the wind turbine.

[0199] Previously, overspeed and tower loads were accepted during curtailment, based on the expectation that these events were neither relevant to extreme loads nor significantly relevant to operating loads. An improvement is now proposed.

[0200] In particular, a solution is proposed that achieves the fastest possible rotor speed reduction without, or without significantly higher, tower base loads than with a slower rotor speed reduction.

[0201] Regardless of the application, the implementation of a rapid, yet load-friendly speed or power reduction is carried out in a very similar way according to the invention. The goal is to bring the blade angle to the new desired equilibrium state as quickly as possible, particularly to the target blade angle, but ideally not beyond it. A closed control loop always runs the risk of overshooting if the setting is too strong. If the setting is too weak, the system is too slow.

[0202] Therefore, according to the invention, a feedforward control takes place in which a blade angle is specified and controlled, which is precalculated. This angle, which forms the target blade angle, is calculated using a mathematical model of the wind turbine such that the wind turbine, which can also be referred to as a turbine for simplicity and synonymously, achieves the aerodynamic target power after reaching this angle. Ideally, the target blade angle can be approached as quickly as possible; it can be precalculated as the ideal blade angle. Once this angle is reached, the process is stopped, or only the closed control loop remains active, or the turbine control then uses the closed control loop again, namely speed control.

[0203] According to one aspect, it is proposed to additionally adjust parameters of the closed control loop, i.e., the speed control. In particular, it is proposed to increase the controller gain of the speed control while the blade angle is being changed toward the target blade angle. This allows the target blade angle to be reached more quickly if the speed control is activated in the meantime.

[0204] Since feedforward controls are always based on assumptions about the controlled system, inaccuracies may exist.

[0205] It is therefore proposed, according to one aspect, to terminate the pre-control already when the desired state in approximatelyis reached, particularly based on the difference between the target blade angle or ideal blade angle and the current blade angle. This can, but does not have to, be done this way. The closed control loop can therefore always remain active without being activated. This is particularly useful when the feedforward control is actually deactivated. To continue to achieve fast control behavior in these situations, it is suggested to adapt the closed control loop, which is otherwise designed for normal operation.

[0206] Since the feedforward control is not continuously active, but only in the above-mentioned application cases, a suggested procedure consists of the following steps: 1. Identification of the application, namely the reduction requirement, in particular limited maximum power or bird protection. 2. Identification of the need for feedforward control, namely checking whether the new operating point, i.e. the target operating point, is so far removed from the current state, i.e. the current or first operating point, that feedforward control intervention appears worthwhile, or whether speed control is sufficient. 3. Activation of feedforward control, i.e. targeted approach to a blade angle, namely the target blade angle, at which the desired aerodynamic power is absorbed or released. 4. If necessary, adjustment of the parameters of the closed control loop, in particular the speed control. 5. Deactivation of feedforward control as soon as the new operating point, in particular the associated new speed, is approximately reached. 6. If necessary, undo adjustment of the control parameters of the speed controller.

[0207] Depending on the application, further steps or aspects may be added.

[0208] For bird protection, which here is representative of species protection, and particularly includes bat protection, the following should be noted: In bird protection, in the target state in which the rotor is supposed to rotate slowly, almost at a standstill, no aerodynamic power can be absorbed at all. However, the present invention focuses on the transient speed reduction process for bird protection, during which aerodynamic power is released or "negative power" is absorbed. Here, the aim is to quickly reduce the aerodynamic power so that a low-risk rotor speed is achieved. In practice, the aerodynamic power is selected such that the rotor is sufficiently decelerated within a desired time without mechanically overloading the wind turbine. For example, the aerodynamic power can correspond to the negative value of the maximum positive aerodynamic power during normal operation. Specifically, for an exemplary turbine with +4.For a nominal power of 2 MW, an aerodynamic deceleration power of, for example, -4.2 MW can be parameterized, provided that the rotor blade can absorb the same loads in the positive and negative thrust directions.

[0209] Due to the relationship dω / dt = J*ΔM, the approximate deceleration time is also known, as long as ΔM also includes the generator torque. Alternatively, it is suggested that the desired deceleration time be used as a reference for calculating the aerodynamic performance.

[0210] In order to provide aerodynamic power and, together with the generator power, deceleration power at any time, the actual speed or actual tip speed ratio is used to calculate the ideal blade angle.

[0211] Since changes in aerodynamic performance lead to a rate of change in speed, it is proposed to define a corresponding deactivation criterion. For example, that the target speed has been approximately reached (see point 5 above, "Deactivation of the feedforward control"). However, stopping the feedforward control beforehand is also possible. Feedforward control could overshoot the target due to model uncertainties. This might be harmless in terms of speed or bird protection. However, the tower could be overloaded, both at its extreme load, especially if the thrust becomes too negative, and at its operating load, especially if aerodynamic damping has become too small due to large blade angles.Therefore, in such cases, we suggest stopping the feedforward control beforehand and controlling the remaining control error with the closed control loop, whereby the closed control loop has been adapted to the situation in its dynamics (see point 4 above, "Adjusting the parameters if necessary").

[0212] The following should be noted regarding the adjustment of a target rotor speed in non-load-critical conditions (see point 4 above, "Adjusting the parameters if necessary"): Particularly for bird protection applications, rapid adjustment of a target speed is necessary even in situations in which the thrust has already been reduced in a load-protecting manner. This is typically the case from a blade angle of 20°. Further adjustment of the parameters of the closed control loop is recommended for the period between leaving the range of load-critical operating points until the desired target speed is reached. For this section of the rotor speed reduction, significantly higher feedback gains are possible than for load-relevant operating points, e.g., a quadrupling.

[0213] The following should be noted regarding maximum power reduction as a curtailment request: When pre-controlling a maximum power, the aerodynamic power can be set directly so that it corresponds to the new desired operating point, i.e., the new maximum feed-in power specified by the curtailment request plus the power losses to be compensated by the generator, which were neglected in the above considerations and descriptions. Therefore, when calculating the ideal blade angle, there is a deviation from the bird protection application. While the actual speed was used there, here, when specifying a power reduction, the target speed, i.e., the new target speed of the new operating point, should be used. This way, the system steers directly to the new operating point.

[0214] A pre-control of a transient for speed reduction is proposed according to one embodiment.

Claims

1. A method for controlling a wind power installation (100) which is connected to an electrical supply network and which has a rotor (106) with rotor blades (108) which are adjustable in terms of their blade angle, is able to be operated at a variable speed and is prepared for generating an installation power (PG) from wind, wherein - a blade angle control (210) is provided for adjusting the blade angles, - a closed-loop speed control (201) is provided for closed-loop control of the speed, - a closed-loop power control is provided for limiting the installation power (PG), and - the wind power installation (100) is able to be operated at an operating point which can be specified, wherein the operating point is characterized at least by the speed and the installation power (PG), comprising the steps of - operating the wind power installation (100) at a first operating point with a first blade angle, - checking for a curtailment request, where a reduction in the speed and / or installation power (PG) is requested, and - if there is a curtailment request, - determining a new operating point as the target operating point depending on the curtailment request, wherein the target operating point is characterized by a target speed (nZ) and a target installation power (PZ), and - determining a setpoint blade angle as the target blade angle (αZ) for the target operating point, wherein - in order to change the blade angle to the target blade angle (αZ), a feedforward control blade angle, or a feedforward adjustment rate describing an adjustment rate of the blade angle, is specified via a feedforward control blade angle control (212), wherein - the feedforward control blade angle control (212) gives the feedforward control blade angle or the feedforward control adjustment rate directly to the blade angle control (210) for implementation.

2. The method as claimed in claim 1, characterized in that - the feedforward control process specifies the feedforward control blade angle or the feedforward control adjustment rate - independently of a speed deviation as a deviation of a recorded speed from a specified speed, and / or - independently of the recorded speed and independently of the recorded installation power (PG) and / or - independently of the closed-loop speed control (201) and independently of the closed-loop power control process, and in particular - the change in the blade angle from the first blade angle to the target blade angle (αZ) is specified completely or predominantly by the feedforward control process.

3. The method as claimed in claim 1 or 2, characterized in that - the target blade angle (αZ) is defined as an ideal blade angle which, under steady-state conditions at the target operating point, leads to the target speed (nZ) and target power (PZ), and / or - the feedforward control process specifies a fixed blade angle or a fixedly specified temporal blade angle profile as the feedforward control blade angle or a fixed adjustment rate as the feedforward adjustment rate, and / or - the feedforward control blade angle or the feedforward control adjustment rate is determined using a model which reproduces the behavior of the wind power installation (100), and / or - a target tip-speed ratio is provided for the target operating point as the tip-speed ratio (λ)) characterizing the target operating point, and - the target blade angle (αZ) is determined in such a way that the target tip-speed ratio is set at the target operating point under steady-state conditions.

4. The method as claimed in any of the preceding claims, characterized in that - a curtailment request is selected from the list comprising: - a speed request as a request for a speed reduction, in particular for the protection of an approaching animal, in particular a bird or a bat, wherein the speed request specifies, in particular, the level of a specific target speed (nZ), and - a power request as a request for a reduction of the installation power (PG), in particular for supporting the electrical supply network, wherein the power request specifies, in particular, the level of a specific target power (PZ), and / or that - in the case of a power request as a curtailment request, - a maximum power is specified for the installation power (PG), - the current installation power (PG), if it is greater than the maximum power, is reduced to the maximum power, and, - in order to change the speed to the target speed (nZ), the blade angle is adjusted to the target blade angle (αZ) by means of the blade angle feedforward control process (212), and / or, - in the case of a speed request as a curtailment request, - the target speed (nZ) is specified, - the installation power (PG) is set depending on the specified target speed (nZ), and, - in order to change the speed to the target speed (nZ), the blade angle is adjusted to the target blade angle (αZ) by means of the blade angle feedforward control process (212), wherein, in particular, - in order to decelerate the rotor (106), the installation power (PG) or a generator torque are specified in order to achieve the greatest possible deceleration of the rotor (106), achieved in combination with an aerodynamic power (Pae) which has been subjected to a feedforward control.

5. The method as claimed in any of the preceding claims, characterized in that - the blade angle feedforward control process (212) specifies the blade angle or the adjustment rate as a manipulated variable, and in particular - the closed-loop speed control (201) specifies a blade angle or an adjustment rate of the blade angle as a manipulated variable depending on a speed deviation, and - the closed-loop speed control (201), at least temporarily, is subordinate to the blade angle feedforward control process (212) such that the manipulated variable specified by the closed-loop speed control (201) is at least temporarily not transmitted to the blade angle control (210) or only with a weighting of less than 50%, wherein it is proposed, in particular, that - of the closed-loop speed control (201) and the blade angle feedforward control process (212), the one which outputs the larger blade angle to be set or the larger adjustment rate to be set as the manipulated variable is active, and / or that - the closed-loop speed control (201) is given a speed profile which is specified, in particular, based on a model.

6. The method as claimed in any of the preceding claims, characterized in that - an aerodynamic braking variable is specified in order to change the speed, wherein the aerodynamic braking variable is provided as a rotor acceleration or an aerodynamic power (Pae), wherein the aerodynamic power (Pae) is a power generated by the rotor (106) from the wind, and - the feedforward control blade angle (αV) is determined depending on the specified braking variable in order to achieve the specified braking variable, wherein, in particular, - a blade angle which results in the specified aerodynamic braking variable is determined as the feedforward control blade angle (αV) depending on the speed, wind speed (VW) and specified aerodynamic braking variable.

7. The method as claimed in any of the preceding claims, characterized in that - the specification of the feedforward control blade angle (αV) is repeated continuously, or at least in short-term steps of less than one second, and / or, - in order to specify the feedforward control blade angle (αV), in particular from a or the aerodynamic braking variable, - a power coefficient (CP) to be set is determined and the feedforward control blade angle (αV) is determined depending on the power coefficient (CP) in order to achieve the power coefficient (CP).

8. The method as claimed in any of the preceding claims, characterized in that - an ideal braking variable is specified in order to specify a or the aerodynamic braking variable, and - the ideal braking variable is filtered by means of a filter in order to obtain the aerodynamic braking variable, wherein, - in order to obtain the aerodynamic braking variable, the filter filters from the specified ideal braking variable at least one natural oscillation frequency of the wind power installation (100), which is in particular a natural tower frequency, and - the filter is, in particular, a band-stop filter and / or a notch filter.

9. The method as claimed in any of the preceding claims, characterized in that - the blade angle is determined depending on speed, wind speed (VW) and specified aerodynamic power (Pae) based on a model and / or based on predetermined relationships and / or - the blade angle feedforward control process (212) specifies the blade angle or the adjustment rate as a manipulated variable in order to perform feedforward control of the aerodynamic power (Pae) and the closed-loop speed control (201), at least temporarily, is subordinate to the blade angle feedforward control process (212), wherein it is proposed, in particular, that - of the closed-loop speed control (201) and the blade angle feedforward control process (212), the one which outputs the larger blade angle to be set or the larger adjustment rate to be set is active.

10. The method as claimed in any of the preceding claims, characterized in that - the closed-loop speed control (201) outputs a setpoint blade angle (αS) or a setpoint adjustment rate as the difference between an actual speed (ni) and a setpoint speed (nS) depending on a closed-loop speed control deviation, in particular to the blade angle control (210), and - the closed-loop speed control (201) is changed or suspended, while the blade angle feedforward control process (212) directly specifies the blade angle to be set or the adjustment rate to be set.

11. The method as claimed in any of the preceding claims, characterized in that - a time profile of a setpoint speed (nS) is specified in order to control the target operating point of the closed-loop speed control (201), and / or - a time profile of a setpoint power is specified in order to control the target operating point of the closed-loop power control, and / or - while the blade angle feedforward control process (212) directly specifies the blade angle to be set or the adjustment rate to be set, - the closed-loop speed control (201) is modified in terms of its parameterization, in particular in terms of at least one gain factor.

12. The method as claimed in any of the preceding claims, characterized in that - a change in blade angle from the first blade angle to the target blade angle (αZ) can be described as a relative blade angle change of from 0% to 100%, wherein 0% corresponds to the first blade angle and 100% corresponds to the target blade angle (αZ), and - the adjustment of the blade angle by means of the blade angle feedforward control process (212) from the first blade angle to the target blade angle (αZ) is interrupted before the target blade angle (αZ) is reached and / or is taken over by the closed-loop speed control (201), wherein, in particular, - the blade angle is adjusted by means of the blade angle feedforward control process (212) for at least 5% to 90%, in particular for 0% to 95%, of the blade angle change and / or - the closed-loop speed control (201) is changed or deactivated during this time.

13. The method as claimed in any of the preceding claims, characterized in that - in order to control the target operating point, an aerodynamic braking power (Pae) with which the rotor (106) is to be braked by the wind is determined, and - the blade angle is determined by the blade angle feedforward control process (212) in such a way that the rotor (106) outputs the aerodynamic braking power (Pae) to the wind, wherein, in particular, - the aerodynamic braking power (Pae) is selected in a range of from 10% to 120% of a rated power (PN) of the wind power installation (100), in particular in a range of from 50% to 100% of the rated power.

14. The method as claimed in any of the preceding claims, characterized in that - in order to determine the target operating point, a wind speed (VW) is estimated based on the speed, the power and the first blade angle, and / or - the speed (n) and / or the power at the target operating point is greater than zero, and / or - a desired reduction time within which the target operating point is to be reached is specified, and - the specification of the temporal blade angle profile is determined depending on the desired reduction time, in particular by using a or the model, and / or - after the blade angle adjustment has been completed using the blade angle feedforward control process (212), - the closed-loop speed control (201) controls or limits the speed (n) to a setpoint speed (nS).

15. The method as claimed in any of the preceding claims, characterized in that - if there is a curtailment request, - there is a check to determine whether the target operating point is at a sufficient distance from the first operating point, and - the blade angle is adjusted by means of the blade angle feedforward control process (212) only if the distance of the target operating point is sufficient, wherein - a sufficient distance is present if the distance is above a predeterminable comparison distance, wherein, in particular, - a difference between the current speed and the target speed (nZ) is considered as the distance and is considered sufficient if it is greater than a minimum speed difference that can be specified, and / or - a difference between the current installation power (PG) and the target power (PZ) is considered as the distance and is considered sufficient if it is greater than a minimum power difference that can be specified, and / or - a difference between the first blade angle and the target blade angle (αZ) is considered as the distance and is considered sufficient if it is greater than a minimum angle difference that can be specified.

16. The method as claimed in any of the preceding claims, characterized in that - an approach of a flying animal, in particular a bird or a bat, leads to a speed request as a curtailment request and - the target speed (nZ) is determined depending on one criterion, a plurality or all of the criteria from the list comprising: - a position of the flying animal, - a species of the flying animal, - a speed of movement of the flying animal, and - a direction of movement of the flying animal and / or - a behavior of the flying animal is observed further and a new target speed (nZ) is specified depending on the observed behavior of the flying animal.

17. A wind power installation (100) which is connected to an electrical supply network and which has a rotor (106) with rotor blades (108) which are adjustable in terms of their blade angle, is able to be operated at a variable speed and is prepared for generating an installation power (PG) from wind, wherein - a blade angle control (210) is provided for adjusting the blade angles, - a closed-loop speed control (201) is provided for controlling or limiting the speed, - a closed-loop power control is provided for limiting the installation power (PG), - the wind power installation (100) is able to be operated at a predeterminable operating point, wherein the operating point is characterized at least by the speed and the installation power (PG), and - the wind power installation (100) is prepared to carry out a method as claimed in any one of the preceding claims, wherein - the wind power installation (100) has an installation control system (103) in which such a method is implemented.