Controlling activation of individual pitch control of wind turbine rotor blades based on blade load volatility

Abstract

The invention provides a controller for a wind turbine having a plurality of rotor blades. The controller is for controlling activation of individual pitch control of the rotor blades. The controller is configured to receive a flap load signal, from a blade load sensor of each of the plurality of rotor blades, indicative of flap loading on each of the respective rotor blades. The controller is configured to determine, based on the received flap load signals, a standard deviation parameter indicative of volatility of loading on the rotor blades of the wind turbine. The controller is configured to control activation of individual pitch control based on the determined standard deviation parameter.

Description

[0001] CONTROLLING ACTIVATION OF INDIVIDUAL PITCH CONTROL OF WIND TURBINE ROTOR BLADES BASED ON BLADE LOAD VOLATILITY

[0002] TECHNICAL FIELD

[0003] The invention relates to controlling activation of individual pitch control of rotor blades of a wind turbine. In particular, activation of the individual pitch control is based on blade load volatility, e.g. caused by wind volatility in the wind field, of the wind turbine.

[0004] BACKGROUND

[0005] Wind turbines as known in the art include a wind turbine tower supporting a nacelle and a rotor with a number of - typically, three - pitch-adjustable rotor blades mounted thereto. Control strategies for adjusting blade pitch can be used to maximise energy production of a wind turbine while minimising loads experienced by various components of the wind turbine.

[0006] Rotor blades may be adjusted as part of a collective pitch control (CPC) routine, in which each of the (three) blades is adjusted in the same way at the same time, where such CPC may be used to control wind turbine speed, for instance. Rotor blades may also be adjusted as part of an individual pitch control (I PC) routine, in which each blade has its own individual pitch reference, possibly as an adjustment to a collective pitch reference from a collective pitch controller. I PC may be used to alleviate loads caused by rotational sampling of the wind field in the vicinity of wind turbine as the rotor rotates.

[0007] Continuous or excessive activation of an individual (or collective) pitch controller may result in excessive wear of the blade bearings. As such, I PC schemes may be combined with activation strategies that allow for certain key loading issues associated with wind turbine operation to be handled or addressed without putting excessive demands on blade bearings.

[0008] Existing activation strategies for I PC may not perform as needed. For instance, in some cases achieving the desired performance of such activation strategies may only happen as a result of being strongly coupled to the activation schemes of wind turbine control features other than I PC. Also, existing activation strategies for handling cases of excessive tilt and yaw loading of the wind turbine may not activate in certain wind conditions that cause such excessive tilt / yaw loading.

[0009] There is a need to further improve I PC activation strategies to ensure that problematic tilt / yaw loads are alleviated across different wind conditions, while continuing to ensure that excessive demands are not placed on the blade bearing.

[0010] It is against this background to which the present invention is set.

[0011] SUMMARY OF THE INVENTION

[0012] According to an aspect of the invention there is provided a controller for a wind turbine having a plurality of rotor blades. The controller is for controlling activation of individual pitch control of the rotor blades. The controller is configured to receive a flap load signal, from a blade load sensor of each of the plurality of rotor blades; apply an m-blade coordinate transformation to the flap load signal to obtain first and second mutually orthogonal components in a fixed coordinate frame of the wind turbine; determine based on the first and second mutually orthogonal components a standard deviation parameter; and control activation of individual pitch control based on the determined standard deviation parameter. The flap load signal is indicative of flap loading on each of the respective rotor blades. The standard deviation is determined based on the received flap load signals and is indicative of volatility of loading on the rotor blades of the wind turbine.

[0013] The individual pitch control may be activated when a tilt / yaw load magnitude signal, indicative of one or both of a tilt moment and a yaw moment on a rotor of the wind turbine, is greater than a tilt / yaw load threshold value. To control activation of individual pitch control, the controller may be configured to adjust the tilt / yaw load threshold value based on the determined standard deviation parameter.

[0014] The controller is configured to decrease the tilt / yaw load threshold value when the determined standard deviation parameter increases. Optionally, the tilt / yaw load threshold value may be decreased linearly as the determined standard deviation parameter increases. The controller may be configured to adjust the tilt / yaw load threshold value to be no less than a defined minimum tilt / yaw load threshold value and / or no greater than a defined maximum tilt / yaw load threshold value.

[0015] To determine the standard deviation parameter the controller may be configured to determine first and second standard deviation components and determine a magnitude of the first and second standard deviation components to obtain the standard deviation parameter. The first and second standard deviation components may be based on the respective first and second mutually orthogonal components.

[0016] Prior to determining the magnitude, the controller may be configured to apply a low pass filter to the first and second standard deviation components to remove high frequency content.

[0017] Prior to determining the first and second standard deviation components, the controller may be configured to isolate frequency content indicative of volatility of loading on the rotor blades of the wind turbine in the first and second mutually orthogonal components.

[0018] The controller may be for implementing individual pitch control of the rotor blades. The controller may be configured to determine, based on the first and second mutually orthogonal components, a tilt / yaw load magnitude signal, and control implementation of individual pitch control based on the determined tilt / yaw load magnitude signal. The tilt / yaw magnitude signal may be indicative of one or both of a magnitude of a tilt moment and a magnitude of a yaw moment on a rotor of the wind turbine.

[0019] To control implementation of individual pitch control, the controller may be configured to define a function describing a relationship between a maximum pitch amplitude that is permitted to implement individual pitch control and the tilt / yaw load magnitude signal. The relationship may be defined between a minimum available pitch amplitude and a maximum available pitch amplitude. The function may define at least that the maximum permitted pitch amplitude of the rotor blades increases as the determined tilt / yaw load magnitude signal increases.

[0020] The controller may be configured to shift the function such that the maximum permitted pitch amplitude of the rotor blades increases for a given tilt / yaw load magnitude signal as the tilt / yaw threshold value increases. The controller being configured to isolate 3P frequency content components in the respective first and second mutually orthogonal components and determine a 3P magnitude signal of the respective 3P frequency content components. The tilt / yaw load magnitude signal may be determined based on the determined 3P magnitude components.

[0021] Prior to determining the standard deviation parameter, the controller may be configured to determine a corrected flap load signal indicative of the received flap load signal corrected to include an estimation of additional loads reduced by individual pitch control of the wind turbine. The standard deviation parameter may be determined based on the corrected flap load signal.

[0022] According to another aspect of the invention there is provided a wind turbine comprising an activation controller as defined above.

[0023] According to another aspect of the invention there is provided a method for a wind turbine having a plurality of rotor blades. The method is for controlling activation of an individual pitch controller that is for controlling individual pitch of the rotor blades. The method comprises receiving a flap load signal indicative of flap loading on each of the respective rotor blades applying an m-blade coordinate transformation to the flap load signal to obtain first and second mutually orthogonal components in a fixed coordinate frame of the wind turbine, determining based on the first and second mutually orthogonal components a standard deviation parameter indicative of volatility of loading on the rotor blades of the wind turbine; and controlling activation of the individual pitch controller based on the determined standard deviation parameter. The flap load signal is received from a blade load sensor of each of the plurality of rotor blades. The standard deviation parameter is determined based on the flap load signal.

[0024] BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Examples of the invention will now be described with reference to the accompanying drawings, in which:

[0026] Figure 1 schematically illustrates a wind turbine in accordance with an aspect of the invention; Figure 2 schematically illustrates an overall controller of the wind turbine of Figure 1 in accordance with an aspect of the invention;

[0027] Figure 3 schematically illustrates modules or components of a pitch adjustment unit of the controller of Figure 2;

[0028] Figure 4 schematically illustrates an activation module of the pitch adjustment unit of Figure 3;

[0029] Figure 5 schematically illustrates a function describing an activation threshold against a standard deviation parameter as determined by the activation module Figure 4;

[0030] Figure 6 schematically illustrates how a permitted maximum pitch amplitude to be used by the pitch adjustment unit of Figure 3 is determined;

[0031] Figure 7 schematically illustrates how the pitch adjustment unit of Figure 3 determines a load magnitude signal in a 1P controller;

[0032] Figure 8 schematically illustrates how the pitch adjustment unit of Figure 3 determines a load magnitude signal in a 2P controller;

[0033] Figures 9(a) and 9(b) show plots illustrating results obtained by implementing the controller of Figure 2;

[0034] Figure 10 summarises steps of a method performed by the controller of Figure 2;

[0035] Figure 11 schematically illustrates how the pitch adjustment unit of Figure 3 determines a corrected load magnitude signal; and

[0036] Figure 12 schematically illustrates how a permitted maximum pitch amplitude to be used by the pitch adjustment unit of Figure 3 is determined based on the corrected load magnitude signal of Figure 11. DETAILED DESCRIPTION

[0037] Figure 1 illustrates, in a schematic view, an example of a wind turbine 10. The wind turbine 10 includes a tower 102, a nacelle 103 disposed atop the tower 102, and a rotor 104 operatively coupled to a generator housed inside the nacelle 103. In addition to the generator, the nacelle 103 houses other components required for converting wind energy into electrical energy and various components needed to operate, control, and optimise the performance of the wind turbine 10. The rotor 104 of the wind turbine 10 includes a central hub 105 and three rotor blades 106 that project outwardly from the central hub 105. Moreover, the wind turbine 10 comprises a control system or controller (not shown in Figure 1). The controller may be placed inside the nacelle 103, in the tower 102, or distributed at a number of locations inside (or externally to) the turbine 10 and communicatively connected to one another.

[0038] The rotor blades 106 are pitch-adjustable. The blades 106 can be adjusted in accordance with a collective pitch setting, where each of the blades are set to the same pitch value. In addition, the blades 106 are adjustable in accordance with individual pitch settings, where each blade 106 may be provided with an individual pitch setpoint.

[0039] The wind turbine 10 includes blade load sensors 112 placed at, or in the vicinity of, each blade root 109 in a manner such that the sensor detects loading of the blade 106. Blade load signals from such sensors may be used to determine how to adjust the pitch of each of the individual blades 106. Depending on the placement and the type of the blade load sensors 112, loading may be detected in the flap (flapwise) direction 110 (in / out of plane) or in the edge (edgewise) direction 108 (in-plane). Such sensors may be strain gauge sensors or optical Bragg-sensors, for instance.

[0040] As the blade load sensors 112 are placed on the rotating blades 106, such load signals for each of the adjustable rotor blades 106 are measured in a rotating reference frame of the rotor 104. In different examples, more than one blade load sensor may be provided in each blade 106. In a preferred example, the blade load sensors 112 of the described example are arranged to measure flap loading on the blades 106.

[0041] Figure 2 schematically illustrates an example of an overall controller 20 of the wind turbine 10 implemented to determine individual pitch actuation signals capable of reducing blade loads experienced by the rotor blades 106. In the illustrated implementation, a speed controller (control module / block) 202 of the overall controller 20 minimises a speed error (ω - ωref) between the actual rotor speed, ω, and a reference rotor speed, ωref, in order to output a requested power P (in the form of a power setpoint), and a collective pitch reference 0coi. The collective pitch reference as determined by the speed controller 202, in view of the rotor speed, may also take further sensor values into account. This is referred to in Figure 2 as a measurement set, ms, being input into the speed controller 202. The feedback speed controller 202 may be implemented by a PI (proportional-integral), PID (proportional-integral-derivative), or similar control scheme. In one example, the speed controller 202 may alternatively be a model predictive controller which, based on minimising a cost function, is arranged to determine the collective pitch reference and / or the power reference.

[0042] Figure 2 further illustrates a control block / module or controller 204, of the overall controller 20, which may be referred to as a pitch actuation unit (PAU). In the PAU 204, pitch reference offset values A01, A02, A03 are determined based on one or more input signals 210. The input signals include blade load signals from the blade load sensors 112 of the blades 106.

[0043] The PAU 204 determines pitch reference offset values A01, A02, A03 for each blade 106. These offsets are superimposed onto the collective pitch reference 9COito provide resulting, or overall, pitch signals 0A, 0B, 0C that can be applied to the pitch actuators of the blades 106 individually.

[0044] In the example shown in Figure 2, a collective pitch reference 0COifor the pitch-adjustable blades 106 is being determined based on a rotor speed, w, and a resulting pitch signal 0A, 0B, 0c is applied to the pitch-adjustable blades 106. The resulting pitch signal 0A, 0B, 0C is applied to the pitch-adjustable blades 106 individually, and, for each individual blade 106 is based on a signal of the collective pitch reference 0coi and the respective individual pitch reference offset values A01, A02, A03. In one example, the individual pitch reference offset values A01, A02, A03 is being applied in a cyclic manner.

[0045] The described controller 20 may be in the form of any suitable computing device, for instance one or more functional units or modules implemented on one or more computer processors. Such functional units may be provided by suitable software running on any suitable computing substrate using conventional or custom processors and memory. The one or more functional units may use a common computing substrate (for example, they may run on the same server) or separate substrates, or one or both may themselves be distributed between multiple computing devices. A computer memory may store instructions for performing the methods performed by the controller 20, and the processor(s) may execute the stored instructions to perform the methods.

[0046] The present invention is directed towards improved activation strategies for individual or cyclic pitch control of wind turbine rotor blades 106. The improved activation strategies aim to optimise a balance between alleviation of component loading that is achievable via the implementation of individual pitch control (I PC), while not putting excessive demands on a wind turbine blade bearing as a result of IPC being activated. In particular, the invention provides for the detection of wind events in the vicinity of a wind turbine 10 that are associated with an increased likelihood of high component loading, specifically loading in tilt and / or yaw directions of the wind turbine rotor, and providing for activation of component loading alleviation via IPC upon detection of such wind events.

[0047] To achieve these beneficial effects, the present invention takes into account a level of volatility in a wind field in which a wind turbine 10 operates as at least part of an activation strategy for IPC. In particular, an indication of wind field volatility is determined via a consideration of standard deviation of blade flap loading on wind turbine blades 106. Specifically, the standard deviation of blade flap loading is used to adjust an activation threshold loading level for an individual pitch controller of the wind turbine. That is, the invention provides for a dynamic threshold loading level at which IPC is activated, where the activation threshold is adjusted in dependence on the volatility of the blade flap loading, which is indicative of the volatility of the wind field in which the wind turbine 10 operates.

[0048] In a general sense, higher fluctuations on the loading experienced by the wind turbine blades typically require a lower activation threshold for IPC. Basing the activation threshold on load volatility ensures that IPC control is activated for more / all cases in which problematic loading occurs. Typically, higher wind speeds result in higher fluctuations in blade loading. However, the degree of such fluctuations can vary for a given wind speed.

[0049] In the described example, the PAU 204 of Figure 2 is used to determine the pitch adjustment values, or pitch reference offset values A01, A02, A03 for each blade 106 to target certain rotor-speed harmonic frequencies in order to alleviate component loading as part of an IPC routine. The PAU 204 is then also used to determine whether to activate such IPC - i.e. whether or not to apply the pitch reference offset values, A01, A02, A03, to the collective pitch reference, 0coi, to be output by the controller 20 - based on detected load volatility.

[0050] Figure 3 schematically illustrates modules or components of the PAU 204 in one example in accordance with the invention. In particular, the PAU 204 includes an individual pitch controller or module 31 for determining pitch reference offset values A01, A02, A03 for the blades 106, for instance to counteract certain components at harmonic frequencies in the input signal 210.

[0051] The input signal 210 is sensor data from the blade load sensors 112. In the described example, the blade load sensors 112 are flap load sensors that indicate loading on a respective blade 106 in the flapwise direction. In different examples, the input signal may additionally include edge loads on the rotor blades.

[0052] In the described example, the individual pitch controller 31 targets or counteracts so-called 1P and 2P frequency content in the input signal 210. In the described example, the individual pitch controller 31 includes separate 1P and 2P controllers 312, 313 for generating pitch reference offset values A01, A02, A03 for counteracting 1P and 2P frequency content, respectively. A 1P frequency is a rotational frequency of the wind turbine rotor 104, i.e. the frequency with which a full rotation of the rotor 104 is completed; such disturbances tend to appear at OP in the fixed frame. 2P cyclic disturbances in a rotor coordinate frame (rotating reference frame) are twice the rotor rotational frequency, but such disturbances can appear at 3P in a fixed coordinate frame. As such, a 2P controller 313 may be used to counteract loading at 3P in the fixed frame, where 3P is the frequency at which the blades 106 pass the tower 102 in a three-blade wind turbine 10, i.e. three times per complete rotation of the rotor 104. The skilled person will be aware of how such 1 P and 2P controllers operate in the art.

[0053] In the example illustrated in Figure 3, the PAU 204 comprises an activation module 32 configured to determine whether the individual pitch controller is to be activated. This determination is with reference to a (current) blade load level / value (input signal 210) relative to a threshold load level. Specifically, the threshold value is a dynamic value, and is determined / adjusted based on the input signal 210. In an example, the activation module 204 may be configured to determine more than one threshold value. A first threshold value may be determined at the 1 P activation module 322 and a second threshold value may be determined at the 2P activation module 324. The first threshold value may be used to initiate 1P pitching by the 1P controller 312 and the second threshold value may be used to initiate 2P pitching by the 2P controller 313. The first and second threshold values may be the same or different.

[0054] Figure 4 schematically illustrates control scheme components / modules of the activation module 32 of the PAU 204 for determining the dynamic activation threshold Lon(tilt / yaw load activation threshold), in accordance with an example of the invention. The activation threshold for the 1P controller 312 and / or the 2P controller 313 may be determined in this manner. The activation threshold Lonis determined based on a load volatility signal, which may be indicative of a volatility of the wind field in the vicinity of the wind turbine 10. In particular, the activation threshold Lonis determined in dependence on a standard deviation parameter which is indicative of variation in the blade load signal 210, by applying an m-blade coordinate transformation to the flap load signal to obtain first and second mutually orthogonal components in a fixed coordinate frame of the wind turbine, and determine a standard deviation parameter indicative of volatility of loading on the rotor blades of the wind turbine based on the mutually orthogonal components. The standard deviation parameter may be obtained directly from the two orthogonal components or from a signal treated version, i.e. a filtered version, of these components.

[0055] As illustrated in Figure 4, the activation threshold module 32 receives the measured blade loading signal 210 in the form of blade bending moments Ma, Mb, Mc. In one example, the activation module 32 may correct the flap bending moments Ma, Mb, Mcsuch that they are indicative of the flap bending moments that would be experienced by the rotor blades 106 were the individual pitch controllers 31 that target 1 P and 2P frequency content not in operation. The corrected bending moments Mal, Mbl, Mc, may be determined by first measuring the current pitching angle of each of the rotor blades 106 and subtracting an average pitch angle, therefore calculating a resulting pitching angle experienced by each blade 106 due to the I PC, specifically. The average pitch angle is determined by finding the sum of the measured current pitching angles of the rotor blades 106 and dividing it by the total number of rotor blades 106.

[0056] The activation threshold module 32 may then be configured to determine a pitching force signal experienced by each blade 106 due to the resulting pitching angle experienced by each rotor blade 106 as a result of only the I PC, for example through the use of a look-up table. The look-up table may map the sensitivity of tilt / yaw loading experienced at the rotor to small pitch angle changes for an operating point. An additional blade load signal may then be calculated at the activation threshold module 32 for each blade 106. The additional blade load signals are indicative of an additional load which would be experienced by each blade 106 if the pitch reference offset values, A01, A02, A03, were equal to zero. The corrected bending moments Mal, Mbl, Mc, are then calculated by adding together the flap bending moments Ma, Mb, Mcand the additional blade load signals. The corrected bending moments Mal, Mbl, Mc, are indicative of the load which would be experienced by each blade 106 if pitch reference offset values, A01, A02, A03, were equal to zero. The corrected bending moments Ma,, Mbl, Mc, are described in more detail in EP 3 055 557 B1 where they are referred to as corrected operational loads. By using corrected bending moments Ma„ Mbl, Mcl, the activation system is uninfluenced by the final output of the IPC, which avoids unwanted premature deactivation of the IPC.

[0057] In the following description, examples of control schemes and activation strategies will be described with reference to the uncorrected blade bending moments Ma, Mb, Mc. However, this is not intended to restrict the scope of the application and, throughout the description, references to the blade bending moments Ma, Mb, Mcmay equivalently be replaced by references to the corrected bending moments Mal, Mbl, Mcl.

[0058] The m-blade (multi-blade) transformation module 402 is configured to transform the (corrected) blade bending moments Ma, Mb, Mc, which are in a rotational coordinate frame of the wind turbine 10 (e.g. centred at the rotor 104), into two mutually orthogonal components Md, Mqin a fixed coordinate frame of the wind turbine 10. The m-blade transformation takes the three rotating signals into a fixed reference frame along a first reference direction d and a second reference direction q. The m-blade coordinate transformation may be in the form of a Coleman transformation. The Coleman transformation may be defined as follows:

[0059] vM

[0060] Md_ Fcos( cos(i > + 2TT / 3) cos(x — 2TT / 3)1 “ Mqsin(i sin( > + 2TT / 3) sin( > — 2TT / 3)

[0061] LMB

[0062] C.

[0063] where Mabcis the three-dimensional vector, where each value of the vector indicates a flap bending moment Ma, Mb, Mcassociated with a respective one of the three blades 106, (in the rotor coordinate frame), which in the described example is in the flap direction (but in different examples could be in the edge direction), xp is the (1P) phase, and Mdqis a vector in the fixed coordinate frame. When Mabcis the vector containing the out-of-plane components of the flap bending moments Ma, Mb, Mcand the phase is set as the rotor azimuth, the first and second reference directions d, q are tilt and yaw directions; however, in general, the Coleman transformation simply transforms the rotating signals into mutually orthogonal first and second components in the fixed reference frame.

[0064] In the described example, the activation module 32 applies a high pass filter 404 to the two mutually orthogonal components Md, Mq. The high pass filter 404 may be configured to retain frequency content associated with volatile loading on the rotor blades which indicate the volatility in the wind field. For instance, the high pass filter may be configured to retain frequency content above a level that is greater than 3P. Optionally, the activation module 32 may also be configured to apply a notch filter 406 to remove 3P content in the fixed frame signal.

[0065] After the filtering has been performed, the activation module 32 is configured to calculate two standard deviation components in a known manner based on the obtained filtered components Md filt, Mq filt. In the general example where the illustrated filters are not used, the standard deviation parameter indicative of volatility of loading on the rotor blades are obtained from the mutually orthogonal components, Md, Mq.

[0066] The activation module 32 may then be configured to apply respective low pass filters 412, 414 to the standard deviation components. This may restrict the rate at which the activation threshold Lonmay change. This may be beneficial because it prevents the activation threshold Lonfrom varying at a higher rate than the tilt / yaw load magnitude signal. If the activation threshold Lonwere to change at a greater rate than the tilt / yaw load magnitude signal, then the evolution of the activation threshold Lonmay introduce strong dynamics into the system. In one example, the content of the standard deviation components may be restricted to frequencies less than 0.2P such that the standard deviation components evolve slowly.

[0067] The activation module 32 is then configured to determine the magnitude of the (filtered) standard deviation components at module 416 in a known manner. The standard deviation parameter obtained from the magnitude module 416 is used to determine the dynamic activation threshold Lonat module 418. In particular, the activation threshold Lonmay be determined based on a defined relationship between the standard deviation parameter (i.e. the magnitude value in this example) and the activation threshold. The relationship / function may be defined in the form of pairs of values of the standard deviation parameter and activation threshold and may be accessible via a look-up table of the pairs of values stored in a memory accessible by the controller 20.

[0068] Figure 5 schematically illustrates an example of a defined function 501 describing a relationship between a standard deviation parameter (obtained as outlined above) and the activation threshold Lon. In the illustrated example, the relationship is a linear relationship and as the standard deviation parameter increases, the activation threshold Londecreases. In the illustrated example, the activation threshold Lonis restricted to values between a high activation threshold

[0069]

[0070] and a low activation threshold Llon, where Llon> Lon- Advantageously, by restricting the values of the activation threshold Lon, as described above, the PAU 204 is configured to initiate I PC during only specific wind conditions, thus reducing the loading experienced at the blade bearing over time. The low activation threshold Llonmay be defined such that the I PC is inactive for the majority of operation of the wind turbine 10.

[0071] A determination of whether to activate the individual pitch controller 31 (e.g. one or both of the 1P and 2P controllers 312, 313) is performed based on the value of a load magnitude signal Ltrelative to the determined activation threshold (or load magnitude threshold) Lon. In particular, when the load magnitude signal Lt is less than the activation threshold Lon, the individual pitch controller 31 is / remains deactivated. When the load magnitude signal Ltis greater than the activation threshold Lon, the individual pitch controller 31 is / remains activated.

[0072] Figure 6 schematically illustrates a relationship describing a permitted maximum pitch amplitude

[0073]

[0074] as a function of the load magnitude signal Ltfor use in performing individual pitch control. The determined permitted maximum pitch amplitude |eperm| may be used by the individual pitch controller 31 to determine the pitch signals A01, A02, A03 output by the PAU 204. The determination of the permitted maximum pitch amplitude may be performed by the individual pitch controller module 31 or by the activation threshold module 32. However, note that both the tilt / yaw load magnitude signal Ltand the activation threshold Lonare needed to perform this determination.

[0075] The tilt / yaw load magnitude signal Lt is indicative of current loading on wind turbine components and is determined based on the measured load signal 210 from the blade load sensors. The tilt / yaw load magnitude signal may be determined differently by / for the 1P and 2P controllers 312, 313. An example of how Lt may be determined for each of the 1P and 2P controllers 312, 313 is described in detail below.

[0076] Referring back to Figure 6, the (pitching) function describing permitted maximum pitch amplitude

[0077]

[0078] against load level is defined as a function of the activation threshold Lon(which varies as described above, e.g. as in Figure 5). Figure 6 illustrates a first function 62 that describes the relationship between

[0079]

[0080] and load level when the activation threshold Lonis equal to the high activation threshold L^. In this case, |0perm| is equal to a minimum value 19min| of the maximum pitch amplitude that can be used for I PC when the (current) load level Lt is less than the high activation threshold L^. As Lt increases above the permitted maximum pitch amplitude

[0081]

[0082] increases linearly until |0perm| is equal to a maximum value |0max| of the maximum pitch amplitude that can be used for I PC, and remains at |0max| as Lt continues to increase.

[0083] Figure 6 also illustrates a second function 64 that describes the relationship between and load level when the activation threshold Lonis equal to the low activation threshold Llon. In a corresponding manner to above,

[0084]

[0085] is equal

[0086]

[0087] when the (current) load level Ltis less than the low activation threshold Lln. As Ltincreases above Lln, the permitted maximum pitch amplitude

[0088]

[0089] increases linearly until |0perm| is equal to l^maxL and remains at |0max| as Ltcontinues to increase.

[0090] The second function 64 may be regarded as the first function 62 shifted along the load level axis. More generally, the defined function may shift along the load level axis as the activation threshold Lonvaries. In this case the defined function retains a constant - in this example, linear - shape as the activation threshold Lonvaries. It will be understood that in different examples this need not be the case. It will be understood that the defined function may be any suitable shape, e.g. quadratic.

[0091] Once the pitching function has been defined, e.g. as one of the functions 62, 64 or a different function, the (current) load magnitude signal Lt is used to determine the permitted maximum pitch amplitude based on its position on the defined pitching function.

[0092] Figure 7 schematically illustrates a 1P activation module 322 of the PAU 204 configured to determine the tilt / yaw load magnitude signal Ltfor the 1P controller module 312. Note that the determination of

[0093]

[0094] may alternatively be performed by the 1P controller module 312. In the example control strategy shown in Figure 7, the PAU 204 receives the blade bending moments Ma, Mb, Mc, and preferably determines corrected blade bending moments as described above. The m-blade transformation module 702 is configured to transform the (corrected) blade bending moments Ma, Mb, Mcinto two mutually orthogonal components Md, Mq. The two mutually orthogonal vectors Md, Mqmay be in the tilt and yaw directions, respectively. The 1 P activation module 322 is then configured to apply a notch filter 704 to each of the two mutually orthogonal components Md, Mq. The notch filter 704 may be configured to remove frequency content at / around 3P. After applying the 3P notch filter 704, the 1 P activation module 322 applies a low pass filter 706 to each of the two mutually orthogonal components Md, Mq. The low pass filter 706 is configured to attenuate high frequency content in the two mutually orthogonal components Md, Mq.

[0095] After applying the low pass filter 706, the 1 P activation module 322 may be configured to apply a difference module 712 to calculate the difference between each of the two mutually orthogonal components Md, Mqand respective reference mutually orthogonal components Md ref, A / q re. The mutually orthogonal components Md ref, Mq refmay be introduced through applying a OP reference module 708. The differences calculated may constitute respective tilt and yaw magnitude values. The respective reference tilt and yaw magnitude values may respective reference OP tilt and yaw magnitude values \Mdpref!■ I»C and the signal processing performed by the PAU 204 may result in the frequency content of the two mutually orthogonal components Md, Mqbeing concentrated at OP.

[0096] The 1P activation module 322 may be configured to apply respective addition modules 716, 718 to add the two mutually orthogonal OP magnitude components to two mutually orthogonal 3P magnitude components to determine two respective mutually orthogonal total magnitude components. The 1P activation controller 322 is then configured to determine a magnitude, at magnitude block 720, of the two respective mutually orthogonal total magnitude components. The magnitude value calculated at the PAU 204 constitutes the tilt / yaw load magnitude signal

[0097]

[0098] which is indicative of blade bending moments Ma, Mb, Mcwith a frequency of 1P (or preferably corrected blade bending moments).

[0099] The Mabcis a vector representing the blade loading as measured in the blade root of each blade. This vector may be used in an unmodified form and passed directly through the Coleman transformation, or it may be modified to contain the out-of-plane part of the signal and this modified out-of-plane version may be passed through the Coleman transform.

[0100] In the illustrated example the difference is based on a comparison between the components Md, Mqand respective reference components Md ref, Mq ref. In an embodiment where the unmodified flap loads are passed directly through the Coleman transformation, the d- and q- components may be transferred to estimated tilt and yaw moments and compared to reference tilt and yaw moments. Such transformation may be placed in the signal path between the transformation module 702 and the difference module(s) 710, 712.

[0101] In an embodiment, a scheme may be implemented to ensure a balance is obtained between reducing tilt / yaw moments in the fixed frame and flap loads on the blades. There may be situations where a reduction in the error of the tilt / yaw moments (difference to reference values 710, 712) comes at the expense of an increased flap load, like there may be situations where a reduction in in the error of the tilt / yaw moments does not increases, or even decreases, the flap loads. A modified scheme can be introduced which activates the controller more readily in conditions that are damaging to the flap loads and constrain activation in conditions where reducing tilt and / or yaw comes at a cost to the flap loading. This may be implemented by introducing signal paths which further determines a measure of the flap load, either individually and / or collectively, and uses the maximum of the fixed frame tilt / yaw moments and the flap loads as input into the magnitude block 720 to determine the load magnitude signal Lt.

[0102] Figure 8 schematically illustrates a 2P activation module 324 configured to determine the tilt / yaw load magnitude signal Lt. The 2P activation module 324 receives the blade bending moments Ma, Mb, Mcand optionally determines corrected blade bending moments as described above. An m-blade transformation module 802 (e.g. a Coleman transformation) transforms the (corrected) blade bending moments Ma, Mb, Mcinto two mutually orthogonal components Md, Mqin a fixed coordinate frame, representing the tilt and yaw directions, respectively. The 2P activation module 324 then applies a high pass filter 804 to the two mutually orthogonal components Md, Mqconfigured to attenuate low frequency content in the two mutually orthogonal components Md, Mq.

[0103] After applying the high pass filter 804, the 2P activation module 324 applies a respective 3P transformation modules 810a, 810b to each of the two mutually orthogonal components Md, Mq, individually. The 3P transformation modules 810a, 810b are configured to transform each of the two mutually orthogonal components Md, Mqinto two 3P transformed mutually orthogonal components Md3, Mq3. Each 3P transformation module 810a, 810b requires two input components. In the example control scheme illustrated in Figure 8, the second input component for each of the two 3P transformation modules 810a, 810b is a zero component introduced through a OP block 806. In one example, the 3P transformation 810 is configured apply the following transformation to the first mutually orthogonal component Md.

[0104]

[0105] and the following to the second mutually orthogonal component Mq.

[0106]

[0107] where3Pis the 3P phase, and Md3and M3qare first and second 3P components.

[0108] The 2P activation module 324 then applies notch filters 808, 810 to the obtained 3P transformed mutually orthogonal components Md31, Md32, Mq31, Mq32, which may be configured to remove frequency content around 6P. The 2P activation module 324 is then configured to determine a magnitude, at magnitude block 812, of the obtained signal components. In the example control scheme shown in Figure 8, this magnitude value constitutes the tilt / yaw load magnitude signal Lt.

[0109] Figures 9(a) and 9(b) show simulation plots of the main bearing moment (kNm) in the tilt direction, and the main bearing moment (kNm) in the yaw direction, against the mean wind speed (m / s) far upstream of the rotor measured over a 10-minute period, respectively, in cases where the described activation scheme is and is not used. When the described activation scheme is used - indicated by the open circles in the figures -lower tilt and yaw loads are experienced by the main bearing, particularly at higher wind speeds, compared to when the described activation scheme is not used - indicated by the crosses in the figures.

[0110] Figure 10 summarises the steps of a method 100 that is performed by the PAU 204, including the activation module 32 (as part of the overall controller 20), to control activation of I PC of the rotor blades 106 of the wind turbine 10. At step 1001, the activation module 32 receives blade flap bending moments Ma, Mb, Mc, from a blade load sensor 112 of each of the three blades 106, indicative of flap loading on each of the respective blades 106. At step 1002 of the method 100, the module 32 determines, a standard deviation parameter based on the flap bending moments Ma, Mb, Mc, the standard deviation parameter being indicative of a volatility in a wind field in which the wind turbine 10 operates. In some examples, prior to determining the standard deviation parameter, the received flap bending moments Ma, Mb, Mcmay be processed to include an estimation of additional loads reduced by individual pitch control of the wind turbine 10, with the standard deviation parameter being determined based on the corrected bending moments. That is, the measured loading signals may be modified to include cyclic contributions, e.g. 1P and / or 2P content, that would be present if IPC were absent. In the above-described examples, the magnitude of the standard deviation of two mutually orthogonal components Md, Mqin the fixed coordinate plane - obtained via an m-blade transformation of the measured loading signals in the rotor rotational plane - is determined as the standard deviation parameter. It will be understood that a standard deviation parameter, based on the measured loading signals, may be determined differently in different examples.

[0111] At step 1003 of the method 100, the module 32 controls activation of IPC of the wind turbine 10 based on the determined standard deviation parameter. The standard deviation parameter is indicative of volatility of loading on the rotor blades. The volatility of blade loading is indicative of the volatility of wind in the vicinity of the wind turbine. In one example, the activation of the IPC comprises using the standard deviation parameter to calculate an activation threshold Lonfor comparison with a tilt / yaw load magnitude signal Ltmeasured at the wind turbine blades 106 and operating the PAU 204 to initiate the IPC when the tilt / yaw load magnitude signal Lt is greater than the activation threshold Lon.

[0112] Optionally, hysteresis may be introduced for activation and deactivation thresholds to guard against repeated activation and deactivation of IPC. For instance, the threshold to deactivate IPC may be less than the threshold to activate it in the first place.

[0113] In an optional embodiment of the activation scheme and method described above, the tilt / yaw load magnitude signal Lt may instead be a tilt load magnitude signal calculated based only on the blade loading in the tilt direction. In this embodiment, the tilt load magnitude is determined as described previously regarding the tilt / yaw load magnitude signal Lt, in Figure 7, except the magnitude block only receives the output signal of the tilt addition block 716. The tilt load magnitude signal is then calculated as described above and used to determine a permitted maximum pitch amplitude

[0114]

[0115] by implementing the relationship illustrated in Figure 6. This optional embodiment will provide more precise I PC activation for volatile tilt loads. Advantageously, determining activation based purely on tilt loads may improve blade-tower clearance during transient wind shear events.

[0116] Figures 11 and 12 illustrate an optional feature of the described activation scheme. As has been described above, the motivation of the present invention is to provide an activation scheme which is able to detect volatile load conditions indicative of turbulent wind conditions and initiate an I PC designed to counteract the load experienced at the blades 106 due to the volatile conditions whilst limiting the wear and tear of the blade bearings. In Figures 11 and 12, the activation module 32 is provided with an activation strategy which can detect transient wind events during a period of otherwise low load volatility. As will be described in more detail below, the activation scheme comprises providing the activation module 32 with load magnitude signals Lt and the magnitude of the standard deviation components. It is noted that the load magnitude signals Lt may have been determined using the received flap bending moments Ma, Mb, Mcor, preferably, using the corrected bending moments Mal, Mbl, Mc, as has been described previously and the magnitude of standard deviation components are taken directly from the output of the magnitude module 416 described in relation to Figure 4.

[0117] Figures 11 and 12 illustrate an activation strategy for initiating IPC in a wind turbine 10 which may be carried out by the activation module 32. Referring to Figure 11, the activation module 32 is provided with a plurality of load signals, preferably tilt / yaw load magnitude signals Lt (e.g. as described above with reference to Figures 7 or 8) comprising content at OP and / or 3P in the fixed frame, measured from a first time t to a current time t2and a current tilt / yaw load magnitude signal Lt. The activation module 32 is then configured to apply an analysis module 1106, which is configured to determine a derivative of the load signals with respect to time at the current time t2.

[0118] The magnitude of the standard deviation components, e.g. determined by operation of the magnitude module 416 in Figure 4, and the derivative may then be provided to a ratio module 1112. The ratio module is configured to divide the derivative by the magnitude of the standard deviation components. The result of this operation of the ratio module 1112 constitutes a (corrected) load level Et.

[0119] As is illustrated in Figure 12, the load level Etis then input into a pitching block 1202. The pitching block 1202 defines a pitching function in a similar manner to the function illustrated in Figure 6, i.e. defining a relationship between maximum pitch amplitude and load level. A current (corrected) load level Et- determined as described above - is used to determine a corresponding permitted maximum pitch amplitude |0perm|. The pitching function may be defined as described above with reference to Figure 6. The pitching function is defined between minimum and maximum pitching amplitudes \0min\, |0max|, which may be determined in the same manner described above. The pitching function is defined such that it returns the minimum pitching amplitude

[0120]

[0121] when the load level Et is equal to a minimum pitching load Eminand returns the maximum pitching amplitude |0max| when the load level Et is equal to a maximum pitching load Emax.

[0122] In an optional embodiment of the activation scheme shown in Figures 11 and 12, the load level Etmay be calculated based only on the blade loading in the tilt direction. In this embodiment, the tilt / yaw load magnitude Ltis determined as described previously in Figure 7 except the magnitude block only receives the output signal of the tilt addition block 716. The load level Et is then calculated as described above and provided to the pitching block 1202 in order to determine a permitted maximum pitch amplitude |0perm|. This optional embodiment will provide more precise I PC activation for volatile tilt loads. Advantageously, determining activation based purely on tilt loads may improve blade-tower clearance during transient wind shear events.

[0123] Many modifications may be made to the described examples without departing from the scope of the appended claims. For example, the example activation schemes have been described as using the standard deviation of blade loads to indicate transient volatile wind events. However, it will be understood that the activation schemes disclosed herein may use other statistical measurements (statistical dispersion parameter signals) of blade load variance to infer transient wind events such as the variance, inter-quartile range, and / or any other suitable measure of statistical variation.

Claims

CLAIMS1. A controller for a wind turbine having a plurality of rotor blades, the controller being for controlling activation of individual pitch control of the rotor blades, the controller being configured to:receive a flap load signal, from a blade load sensor of each of the plurality of rotor blades, indicative of flap loading on each of the respective rotor blades;apply an m-blade coordinate transformation to the flap load signal to obtain first and second mutually orthogonal components in a fixed coordinate frame of the wind turbine;determine, based on the first and second mutually orthogonal components, a standard deviation parameter indicative of volatility of loading on the rotor blades of the wind turbine; andcontrol activation of individual pitch control based on the determined standard deviation parameter.

2. A controller according to Claim 1, wherein individual pitch control is activated when a tilt / yaw load magnitude signal, indicative of one or both of a tilt moment and a yaw moment on a rotor of the wind turbine, is greater than a tilt / yaw load threshold value, wherein to control activation of individual pitch control the controller is configured to: adjust the tilt / yaw load threshold value based on the determined standard deviation parameter.

3. A controller according to any preceding claim, wherein the controller is configured to decrease the tilt / yaw load threshold value when the determined standard deviation parameter increases; optionally, wherein the tilt / yaw load threshold value is decreased linearly as the determined standard deviation parameter increases.

4. A controller according to Claim 2 or Claim 3, wherein the controller is configured to adjust the tilt / yaw load threshold value to be no less than a defined minimum tilt / yaw load threshold value and / or no greater than a defined maximum tilt / yaw load threshold value.

5. A controller according to any of Claims 1 to 4, wherein to determine the standard deviation parameter the controller is configured to:determine first and second standard deviation components based on the respective first and second mutually orthogonal components; anddetermine a magnitude of the first and second standard deviation components to obtain the standard deviation parameter.

6. A controller according to Claim 5, wherein, prior to determining the magnitude, the controller is configured to:apply a low pass filter to the first and second standard deviation components to remove high frequency content.

7. A controller according to Claim 5 or Claim 6, wherein, prior to determining the first and second standard deviation components, the controller is configured to:isolate frequency content indicative of volatility of loading on the rotor blades of the wind turbine in the first and second mutually orthogonal components.

8. A controller according to any previous claim, the controller being for implementing individual pitch control of the rotor blades, the controller being configured to:determine, based on the first and second mutually orthogonal components, a tilt / yaw load magnitude signal, indicative of one or both of a magnitude of a tilt moment and a magnitude of a yaw moment on a rotor of the wind turbine; andcontrol implementation of individual pitch control based on the determined tilt / yaw load magnitude signal.

9. A controller according to Claim 8, wherein, to control implementation of individual pitch control, the controller is configured to:define a function describing a relationship between a maximum pitch amplitude that is permitted to implement individual pitch control and the tilt / yaw load magnitude signal, wherein the relationship is defined between a minimum available pitch amplitude and a maximum available pitch amplitude, the function defining at least that the maximum permitted pitch amplitude of the rotor blades increases as the determined tilt / yaw load magnitude signal increases.

10. A controller according to Claim 9, wherein the controller is configured to shift the function such that the maximum permitted pitch amplitude of the rotor blades increases for a given tilt / yaw load magnitude signal as the tilt / yaw threshold value increases.

11. A controller according to any of claims 8 to 10, the controller being configured to:isolate 3P frequency content components in the respective first and second mutually orthogonal components; anddetermine a 3P magnitude signal of the respective 3P frequency content components, wherein the tilt / yaw load magnitude signal is determined based on the determined 3P magnitude components.

12. A controller according to any previous claim, wherein, prior to determining the standard deviation parameter, the controller is configured to determine a corrected flap load signal indicative of the received flap load signal corrected to include an estimation of additional loads reduced by individual pitch control of the wind turbine, and wherein the standard deviation parameter is determined based on the corrected flap load signal.

13. A wind turbine comprising an activation controller according to any previous claim.

14. A method for a wind turbine having a plurality of rotor blades, the method being for controlling activation of an individual pitch controller that is for controlling individual pitch of the rotor blades, the method comprising:receiving a flap load signal, from a blade load sensor of each of the plurality of rotor blades, indicative of flap loading on each of the respective rotor blades;applying an m-blade coordinate transformation to the flap load signal to obtain first and second mutually orthogonal components in a fixed coordinate frame of the wind turbine;determining, based on the first and second mutually orthogonal components, a standard deviation parameter indicative of volatility of loading on the rotor blades of the wind turbine; andcontrolling activation of the individual pitch controller based on the determined standard deviation parameter.

Images