WO2026008115A1

WO2026008115A1 - Calibrating a sensor of a wind turbine

Info

Publication number: WO2026008115A1
Application number: PCT/DK2025/050112
Authority: WO
Inventors: Johnny Nielsen; Goncalo Artur Duarte Pereira; Poul Brandt Christensen; Mahmood MIRZAEI
Original assignee: Vestas Wind Systems AS
Current assignee: Vestas Wind Systems AS
Priority date: 2024-07-04
Filing date: 2025-07-02
Publication date: 2026-01-08
Anticipated expiration: 2027-01-04

Abstract

The invention relates to calibrating a wind turbine sensor. The invention involves obtaining over time a sensor signal, from the sensor, indicative of measured parameter values associated with the wind turbine, obtaining over time a monitored signal indicative of monitored operational parameter values associated with the wind turbine, and obtaining over time a wake detection signal indicating a wake flow experienced by the wind turbine. Based on the sensor, monitored and wake detection signals, the invention involves identifying the measured monitored operational parameter values obtained when the wind turbine is not experiencing wake flow. Based on the identified values of the measured and monitored operational parameters, the invention involves determining over time a calculated operational parameter as a function of the measured parameter, determining a correction to the measured parameter based on the calculated operational parameter, and adjusting the measured parameter based on the correction to calibrate the sensor.

Description

CALIBRATING A SENSOR OF A WIND TURBINE

TECHNICAL FIELD

The invention relates to calibrating a sensor of a wind turbine. In particular, the invention relates to calibrating the wind turbine sensor using parameter values obtained when the wind turbine is not experiencing wake flow.

BACKGROUND

Wind turbines are used to capture energy in the wind as it flows past them, and to generate electrical power from the captured energy, e.g. to be supplied to an electrical grid. Wind turbines operate in a variety of conditions, e.g. environmental conditions, and monitoring and controlling operational parameters of a wind turbine in view of such conditions is important for maximising power production and/or minimising wind turbine component loading.

An increasing number of the functions implemented by wind turbines used to control wind turbine operation can automatically adapt the parameters associated therewith to be optimised for the respective wind turbine being controlled. One example of such a wind turbine function is a function that has the aim of determining a relative wind direction for which a wind turbine is maximising power production, and then adjusting the relative wind direction to where it is measured as zero degrees. Another example is a wind turbine function that calibrates measured wind speed in different wind speed bins based on a difference between measured wind speed and estimated wind speed in each respective wind speed bin.

Many other control functions of wind turbines also perform control actions based on measured values of parameters associated with a wind turbine, such as a wind parameter, e.g. wind direction or wind speed. For instance, an ice detection function may compare estimated wind speed against measured wind speed to detect when ice has formed on rotor blades of a wind turbine. Wind turbine efficiency generally decreases when ice is present, and so control actions to remove ice may be implemented when ice is detected.

The adjustments I calibrations I detections performed by these various wind turbine functions can often be inaccurate. It is against this background to which the present invention is set. SUMMARY OF THE INVENTION

Aspects of the present invention reside in an appreciation that measured parameters of a wind turbine - and, in particular, a relationship between measured parameters and other (monitored) parameters of the wind turbine - are influenced by wake flow experienced by the wind turbine. Examples of the invention therefore take this into account for wind turbine functions that perform calibrations I detections based on measured and monitored parameters.

According to an aspect of the invention there is provided a method of calibrating a sensor of a wind turbine. The method comprises obtaining over time a sensor signal, from the sensor, indicative of values of a measured parameter associated with the wind turbine. The method comprises obtaining over time a monitored signal indicative of values of a monitored operational parameter associated with the wind turbine. The method comprises obtaining over time a wake detection signal indicating a wake flow experienced by the wind turbine. The method comprises, based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and the monitored operational parameter values obtained when the wind turbine is not experiencing wake flow. The method comprises, based on the identified values of the measured parameter and the identified values of the monitored operational parameter, determining over time a calculated operational parameter associated with the wind turbine as a function of the measured parameter. The method comprises determining a correction to the measured parameter based on the calculated operational parameter. The method comprises adjusting the measured parameter based on the correction to calibrate the sensor.

The method may comprise, in an implementation phase after a calibration phase in which the sensor is calibrated: determining a value of a control parameter of the wind turbine based on the adjusted measured parameter; and controlling the wind turbine in accordance with the control parameter value.

The method may comprise, in the implementation phase: obtaining the wake detection signal indicating the wake flow experienced by the wind turbine; and, based on the wake detection signal, controlling the wind turbine in accordance with the control parameter value only when the wind turbine is not experiencing wake flow. The method may comprise, in the implementation phase: determining a further value of the control parameter of the wind turbine; and based on the wake detection signal, controlling the wind turbine in accordance with the further value of the control parameter when the wind turbine is experiencing wake flow.

The method may comprise, in the calibration phase, and based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and monitored operational parameter values obtained when the wind turbine is experiencing wake flow. The method may comprise, in the calibration phase, and based on the identified values of the measured parameter and identified values of the monitored operational parameter when the wind turbine is experiencing wake flow, determining over time a further calculated operational parameter associated with the wind turbine as a function of the measured parameter. The method may comprise, in the calibration phase, determining a further correction to the measured parameter based on the further calculated operational parameter. In the implementation phase, the further value of the control parameter of the wind turbine may be determined based on the measured parameter adjusted according to the determined further correction.

The wake flow signal may include an indication of a type of wake flow being experienced by the wind turbine from a plurality of defined wake flow types. The method may comprise one or more of the following steps for each defined wake flow type. In the calibration phase, the method may comprise, based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and monitored operational parameter values obtained when the wind turbine is experiencing the respective defined wake flow type. In the calibration phase, the method may comprise, based on the identified values of the measured parameter and identified values of the monitored operational parameter when the wind turbine is experiencing the respective defined wake flow type, determining over time a respective wake type-specific calculated operational parameter associated with the wind turbine as a function of the wake parameter. In the calibration phase, the method may comprise determining a respective wake type-specific correction to the measured parameter based on the respective wake type-specific calculated operational parameter. In the implementation phase, the method may comprise determining a respective wake flow-specific value of the control parameter of the wind turbine based on the measured parameter adjusted according to the determined respective wake type-specific correction. In the calibration phase, the method may comprise, based on the wake detection signal, controlling the wind turbine in accordance with the respective wake flow-specific value of the control parameter when the wind turbine is experiencing the respective defined wake flow type. The plurality of defined wake flow types may include one or more of: full wake flow; half wake flow; left-hand wake flow; and right-hand wake flow.

The sensor may be a wind sensor.

The sensor may be a wind direction sensor. The measured parameter may be a wind direction relative to the wind turbine.

The monitored operational parameter may be pitch angle of at least one rotor blade of the wind turbine. The calculated operational parameter may be a statistical representation of the obtained pitch angle values.

The monitored operational parameter may be a wind power parameter of the wind turbine. The calculated operational parameter may be a statistical representation of the obtained wind power values. The wind power parameter may be one of a rotor power, a grid power, a torque and a rotor blade loading, of the wind turbine.

The control parameter may be a yaw angle for the wind turbine.

The sensor may be a wind speed sensor. The measured parameter may be measured wind speed.

The monitored operational parameter may be estimated wind speed in the vicinity of the wind turbine. The calculated operational parameter may be an error parameter indicative of a difference between measured and estimated wind speed.

The control parameter may be one or more of: a rotor blade pitch control parameter; a torque control parameter; and an ice detection control parameter.

According to another aspect of the invention there is provided a non-transitory, computer readable storage medium storing instructions thereon that, when executed by one or more computer processors, causes the one or more computer processors to perform the method defined above.

According to another aspect of the invention there is provided a controller for a wind turbine. The controller is configured to obtain over time a sensor signal, from a sensor of the wind turbine, indicative of values of a measured parameter associated with wind flow in the vicinity of the wind turbine. The controller is configured to obtain over time a monitored signal indicative of values of a monitored operational parameter associated with the wind turbine. The controller is configured to obtain over time a wake detection signal indicating a wake flow experienced by the wind turbine. The controller is configured to, based on the sensor signal, the monitored signal and the wake detection signal, identify the measured parameter values and monitored operational parameter values obtained when the wind turbine is not experiencing wake flow. The controller is configured to, based on the identified values of the measured parameter and identified values of the monitored operational parameter, determine over time a calculated operational parameter associated with the wind turbine as a function of the measured parameter. The controller is configured to determine a correction to the measured parameter based on the calculated operational parameter. The controller is configured to adjust the measured parameter based on the correction.

According to another aspect of the invention there is provided a wind turbine comprising a controller as defined above.

BRIEF DESCRIPTION OF THE DRAWINGS

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

Figure 1 schematically illustrates a wind turbine in accordance with an aspect of the invention;

Figure 2 schematically illustrates a wind farm comprising a plurality of wind turbines, including the wind turbine of Figure 1 ;

Figure 3(a) shows a plot of output power against relative wind direction of the wind turbine of Figure 1 for data collected over a period of time, and Figure 3(b) shows a plot of output power against relative wind direction of the wind turbine of Figure 1 for data collected over the period of time where the data collected when the wind turbine is experiencing wake flow is removed; and

Figure 4 shows the steps of a method, performed by a controller of the wind turbine of Figure 1 , in accordance with an aspect of the invention. DETAILED DESCRIPTION

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 at the apex of, or 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, e.g. a gearbox, 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.

The wind turbine 10 includes a number of sensors for measuring various parameters, e.g. wind parameters, associated with wind turbine operation. The wind turbine 10 may include one or more wind direction sensors for measuring incoming wind direction in the vicinity of the wind turbine 10. The wind direction sensor(s) may be a wind vane or any other suitable type of sensor, and may be located on the rotor 104 or nacelle 103 of the wind turbine, for instance. The wind turbine 10 may include one or more wind speed sensors for measuring wind speed in the vicinity of the wind turbine 10. The wind speed sensor(s) may be an anemometer, for instance. The wind speed sensor may similarly be located on the rotor 104 or nacelle 103, and may be part of a single device that also comprises the wind direction sensor.

Another type of sensorthat may be provided is an accelerometer, e.g. for measuring vibrations (fore-aft and/or side-side) of the tower 102. One or more accelerometers may be placed at the top of the tower 102 or in/on the nacelle 103. A further type of sensor that may be provided is a blade load sensor. For instance, blade load sensors may be placed at, or in the vicinity of, a root end of each blade 106 in a manner such that the sensor detects loading in the blade 106. Depending on the placement and the type of sensor, loading may be detected in the flap (flapwise) direction (in/out of plane) or in the edge (edgewise) direction (in-plane). Such sensors may be displacement sensors, strain gauge sensors or optical Bragg-sensors, for instance.

The wind turbine 10 includes a control system or controller (not shown in Figure 1). The controller may be placed inside the nacelle, in the tower or distributed at a number of locations inside (or externally to) the turbine 10 and communicatively connected to one another. The wind turbine controller performs various functions for controlling operation of the wind turbine 10. The control signals output by the controller - to control wind turbine components, such as pitch of the rotor blades 106 and speed of the wind turbine generator - may be determined on input signals received by the controller, e.g. measurement signals obtained from sensors such as the sensors outlined above.

One example of a function performed by the wind turbine controller is a wind direction adjustment function. Typically, wind turbine power production is maximised when the wind turbine 10 faces directly into the incoming wind, i.e. when the wind turbine rotor 104 is controlled to be aligned with the incoming wind (by adjusting the yaw angle of the wind turbine using a yaw drive mechanism). However, wind direction sensors of the wind turbine 10 may suffer from inaccuracies, e.g. because of the rotor 104 disrupting the wind flow prior to it being measured by the sensor. The aim of the wind direction adjustment function is to determine the relative wind direction that maximises power production of the wind turbine 10 and then to adjust I calibrate the wind direction sensor measuring the relative direction such that the relative wind direction that maximises power production corresponds to zero degrees. This is an example of a wind turbine function that calibrates a sensor, namely, a wind direction sensor, based on measured and monitored parameters associated with the wind turbine.

The wind direction adjustment function may be performed differently in dependence on whether the wind turbine 10 is in full load operation or partial load operation. EP3394434A1 describes an approach for wind direction adjustment during full load operation, in which the maximum power for a given wind direction is set based on the determined maximum pitch angle observed for the given wind direction. On the other hand, EP3394435A1 describes an approach for wind direction adjustment during partial load operation, in which correlation with maximum power for a given wind direction may be performed directly.

Another function that may be performed by the wind turbine controller is a wind speed adjustment function in form of an adaptive correction function which, over time, calibrates measured wind speed - as measured by a wind speed sensor of the wind turbine 10 - in different defined wind speed bins I ranges based on a difference between measured and estimated wind speed in the respective wind speed bin. Such wind speed adjustment is therefore another example of a wind turbine function that calibrates a sensor, namely, a wind speed sensor, based on measured and monitored parameters associated with the wind turbine. An approach for performing such The wind speed adjustment is described in EP3478962A1. The estimated wind speed is determined based on a wind speed estimator which used measured power and wind speed.

A further example of a function performed by the wind turbine controller is an ice detection function which detects when ice forms on the rotor blades 106 or other components of the wind turbine 10. Wind turbine efficiency, e.g. power production, is generally lower when ice is present. The wind turbine controller may therefore instruct control actions to remove the ice, e.g. blade heating. The presence of ice on the blades can also cause a safety issue, e.g. a risk of ice being thrown from the blades as they rotate about the rotor 104, and so wind turbine operation may need to be halted when ice is detected. An ice detection function may determine a Rotor Efficiency Loss (REL) caused by ice. In particular, the ice detection function may perform a comparison of estimated and measured wind speeds and, in particular, a decrease in estimated wind speed against measured wind speed. If the decrease is greater than a threshold value then a determination that ice is present may be performed. Such ice detection is therefore another example of a function in which an accurate measurement of wind speed is needed.

The wind turbine 10 may be part of a wind farm or wind park. Figure 2 shows a wind park 20 that includes a plurality of wind turbines 21 , including the wind turbine 10 of Figure 1 , in a defined geographical area. The wind park 20 may be an onshore or offshore wind farm. Each wind turbine 21 may be controlled to balance maximising the captured energy I power production of the turbine against (minimising) the loading experienced by one or more components of the turbine 21 .

Figure 2 schematically illustrates a direction 22 of wind flow in the wind park 20. As the wind flows past a first one of the turbines 21a in the wind park 20, wake is generated downstream of the wind turbine 21a. This means that wind flow downstream of the wind turbine 21a is perturbed or disturbed relative to upstream of the wind turbine 21a, resulting in a reduction in the speed of the wind flow and/or an increase in the turbulence of the wind flow.

Depending on the positioning of the other wind turbines in the wind park 20 relative to the (first) wind turbine 21a, the wind flow past one or more of the other wind turbines may include wake effects caused by the wind flow past the first wind turbine 21a. For instance, the wind turbine 10 may experience wake flow caused I generated by the first wind turbine 21a as it is downstream of the first wind turbine 21a.

Wake flow experienced by the wind turbine 10 may additionally or alternatively be caused I generated by factors other than neighbouring wind turbines. For instance, wake flow may be generated as wind flows over/round a building or other object, or by geographical features such as hills and valleys. For wind turbine functions that that perform calibrations I detections based on sensor measurements of wind turbine parameters accurate calibrations I detections may be achieved when the wind turbine 10 is operating when the wind turbine 10 is not experiencing wake flow generated by neighbouring wind turbines or other objects I features. That is, wake flow can contribute to inaccurate calibrations of sensors being performed, inaccurate detections being determined, etc. This means that, in practice, wind turbine functions may perform with reduced accuracy as it is common for wind turbines to operate in wake flow at least some of the time. For instance, a wind turbine that operates in the middle of a wind farm may experience wake flow up to 50% of its operating time.

This can lead to drawbacks such as power production of the wind turbine 10 being lower than expected. For instance, if calibration of a wind direction sensor is performed inaccurately, then this will result in misalignment of the wind turbine 10 relative to the incoming wind direction, leading to reduced levels of energy capture. As another example, wake flow may lead to incorrect detection of ice on the blades, as wake flow influences measured wind speed used for ice detection. Even a relatively small number of incorrect detections of ice, e.g. 1%, can lead to a relatively significant loss of power production if the wind turbine 10 is shut down when it does not need to be.

The present invention is advantageous in that it provides for more accurate calibrations of wind turbine sensors to be used with one or more wind turbine functions (such as those outlined above). In turn, this means that more accurate detections I determinations based on sensor measurements may be obtained, which further means that improved control of the wind turbine may be implemented, e.g. increased power production. This is achieved by taking into account wake flow experienced by the wind turbine when calibrating a sensor for use with one or more wind turbine functions. In examples of the invention, only parameter values, e.g. measured parameter values obtained using wind turbine sensors, obtained when the wind turbine is not experiencing wake flow, i.e. when the wind turbine is experiencing free stream flow, are used to calibrate the sensor(s). Further beneficial effects of the invention will become apparent from the following description.

An example of the invention is illustrated with reference to Figure 3. Figure 3(a) is a plot 31 of output power (grid power) of the wind turbine 10 as a function of relative wind direction, where values of the output power and relative wind direction have been obtained over a period of time. Specifically, the period of time during which the data is collected is a period of time in which the wind turbine 10 experiences wake flow for a relatively significant amount of the time, e.g. up to 50% of the time. The wind turbine 10 should produce the most output power when the relative wind direction is 0 degrees, i.e. when the wind turbine rotor 104 is aligned with the wind. However, it is seen in Figure 3(a) that the collected data 31 indicates that output power appears to be maximised at a relative wind direction of about 7.5 degrees (indicated by the line 32). Even fitting a curve 33 to the collected data indicates that the maximum power output is at about 3.5 degrees (indicated by the line 34), i.e. significantly different from 0 degrees. For the purposes of the present test, it was verified that the sensor used to obtain relative wind direction values was accurate to within ±1 degrees (i.e. the sensor is already correctly calibrated), and so the maximum power would be expected to be within one degree of 0. It is identified that data collected when the wind turbine 10 is experiencing wake flow introduces errors in the relative wind direction measurements. If the wind direction adjustment function was to calibrate a wind direction sensor based on the collected data shown in Figure 3(a), then the calibrated sensor would have an error of either about 3.5 or about 7.5 degrees, depending on whether the raw data 31 or the fitted data curve 33 was used for the calibration.

Like Figure 3(a), Figure 3(b) is a plot 35 of output power (grid power) of the wind turbine 10 as a function of relative wind direction, where values of the output power and relative wind direction have been obtained over a period of time. However, unlike in Figure 3(a), in Figure 3(b) the collected data included in the plot 35 only includes data collected when the wind turbine 10 was not experiencing wake flow. This is performed via the use of a wake flow signal that indicates whether the wind turbine 10 is experiencing wake flow or not. When the wake flow signal indicates that the wind turbine 10 is in wake then the obtained power output and relative wind direction data is discarded (i.e. not included in the plot of Figure 3(b)). On the other hand, when the wake flow signal indicates that the wind turbine 10 is not in wake then the obtained power output and relative wind direction data is included in the plot of Figure 3(b). It may be seen in Figure 3(b) that the output power in the collected data 35 is maximised at about 0 degrees (indicated by the line 36). It is further seen that a fitted curve 37 of the collected data is maximised at about 0 degrees. By removing the data that was collected in wake conditions from the data that is to be used to calibrate a sensor, for instance, more accurate calibration is performed.

Figure 4 shows the steps of a method 40 performed by a controller of the wind turbine 10 in accordance with examples of the invention. The method is for calibrating a sensor of the wind turbine 10. At step 401 , the method 40 involves obtaining over time a sensor signal, from the sensor, indicative of values of a measured parameter associated with the wind turbine 10. The measured parameter may be any suitable parameter associated with wind turbine operation. The measured parameter may be a wind parameter. For instance, the measured wind parameter may be wind direction as measured by a wind direction sensor of the wind turbine 10. The measured wind parameter may be wind speed as measured by a wind speed sensor of the wind turbine 10. The measured parameter may be wind turbine tower acceleration, rotor blade loading, or any other suitable parameter. The measured parameter values may be obtained over any suitable period of time during which the wind turbine is operational, e.g. a number of days, weeks or months. The measured parameter values may be obtained at any suitable intervals of time, e.g. at each sampling point of the wind turbine controller or at less frequent intervals such as every few minutes (e.g. every ten minutes).

At step 402, the method 40 involves obtaining over time a monitored signal indicative of values of a monitored operational parameter associated with the wind turbine 10. The monitored operational parameter may be any suitable parameter associated with wind turbine operation. The monitored operational parameter may be a wind power parameter of the wind turbine 10, such as an output power, a grid power, a rotor power, a torque of the wind turbine 10, a rotor loading of the wind turbine 10, etc. The output power may be a measured or estimated value, e.g. determined using a defined equation for power as a function of a defined coefficient of power and wind speed. The monitored operational parameter may be an estimated wind speed in the vicinity of the wind turbine 10. Estimated wind speed may be obtained in any suitable manner, such as based on a power of the wind turbine or on wind turbine thrust, as is known in the art. The monitored operational parameter may be a rotor pitch angle of one or more rotor blades 106 of the wind turbine 10. The monitored pitch angle may for instance be a pitch angle reference output by a controller, e.g. full load controller, of the wind turbine 10. The monitored operational parameter values may be obtained I sampled at the same time points I intervals as the measured parameter values (such that pairs of corresponding values of the measured parameter and monitored operational parameter are obtained).

At step 403, the method 40 involves obtaining over time a wake detection signal indicating a wake flow experienced by the wind turbine 10. In a broadest sense of the invention, the wake detection signal may be a signal that indicates whether the wind turbine 10 is currently experiencing wake flow (of any type) or currently experiencing no wake flow, i.e. a binary signal. In some examples, the wake detection signal may indicate a type of wake flow being experienced by the wind turbine 10 (if the wind turbine is experiencing wake). For instance, the wake detection signal may indicate whether the wind turbine 10 is in full wake or half wake conditions. In full wake conditions, the wind flow past the turbine 10 across a full span of the swept area of the rotor 104 is wake flow. In half wake conditions, the wind flow past the turbine 10 across only half of the swept area of the rotor 104 may be wake flow. For half wake conditions, the wake detection signal may further indicate the type of half wake flow, e.g. lefthand wake or right-hand wake, indicating whether the wake flow is at a left or right hand side of the turbine rotor 104. The type of wake flow indicated in the wake detection signal may further indicate a severity of the wake flow being experienced by the wind turbine 10. The wake detection signal may be received from a wake detection module of the wind turbine controller, or may be received from off-board the wind turbine 10. The wake detection signal may be received at each time step I interval that the measured parameter values and monitored operational parameter values are obtained.

In some examples, the method 40 may involve a step of determining wake flow experienced by the wind turbine 10, and then generating the wake detection signal based on the determination. The determination of wake flow may be performed in any suitable manner. For instance, this could be based on a temporal variation of measured wind speed in the vicinity of the wind turbine 10 or based on power production I output for certain measured wind speeds. One example for determining wake flow in the vicinity of the wind turbine 10 may be found in WO 2023/025365; however, it will be understood that this is an illustrative, non-limiting example only. Determinations as to whether a wind turbine is experiencing wake may be relative, and so may be performed with respect to suitably defined thresholds in order to generate the (possibly binary) wake detection signal.

At step 404, the method 40 involves identifying the measured parameter values (step 401) and the monitored operational parameter values (step 402) obtained when the wind turbine 10 is not experiencing wake flow. This identification is performed with respect to the received wake detection signal (step 403) that indicates the wake flow being experienced by the wind turbine 10. When the wake detection signal indicates that the wind turbine 10 is not experiencing wake flow, then values of the measured parameter and monitored operational parameter obtained during that time are retained (i.e. the ‘identified’ values). On the other hand, when the wake detection signal indicates that the wind turbine 10 is experiencing wake flow, then values of the measured parameter and monitored operational parameter obtained during that time are discarded or ignored. In this way, only parameter values obtained when the wind turbine 10 is experiencing no wake flow are retained for the following calibration steps.

At step 405, the method 40 involves determining over time a calculated operational parameter associated with the wind turbine 10 as a function of the measured parameter. In particular, this determination is performed based on the identified values of the monitored operational parameter, i.e. the values of the monitored operational parameter obtained when the wind turbine 10 was in a no wake condition. The calculated operational parameter may be any suitable parameter to be used to calibrate the wind turbine sensor. For instance, when the monitored operational parameter is a wind power parameter, the calculated operational parameter may be a statistical representation of the obtained wind power values. In the wind direction adjustment function example above, output power values (‘monitored parameter’ values) are collected over time and then a statistical representation of the collected values in the form of average values for each relative wind direction (or range of wind directions) is determined, as shown in the plot 35 in Figure 3(b). In this case, the average output power values may be the calculated operational parameter values. Other types of statistical representations may be used as calculated operational parameters in different contexts, e.g. variation, standard deviation, etc. In another example, when the monitored operational parameter is rotor blade pitch angle of the rotor blades 106, the calculated operational parameter may be a statistical representation of the obtained pitch angle values. In a further example, when the monitored operational parameter is estimated wind speed, the calculated operational parameter may be an error parameter indicative of a difference between measured and estimated wind speed. This may be the case in the wind speed adjustment function mentioned above.

At step 406, the method 40 involves determining a correction to the measured parameter based on the calculated operational parameter. This correction may in some examples be in the form of an offset. For instance, in the wind direction adjustment example described above, it is determined for what relative wind direction the maximum output power is achieved using the plotted data (i.e. using the calculated operational parameter), and then an offset is determined such that the maximum output power is shifted to a relative wind direction of 0 degrees. Indeed, the measured parameter - in this case, the relative wind direction - is adjusted based on the correction I offset to calibrate the sensor. In different examples, the correction may be in a form other than an offset, e.g. a gain I multiple.

By calibrating the sensor based only on parameter data collected when the wind turbine is not in wake, wind turbine functions using the calibrated sensor are optimised for use in no wake conditions. Steps 401 to 406 may be regarded as a calibration phase or training phase of the method 40. Further steps of the method 40 may be regarded as an implementation phase or execution phase, which make use of the calibration performed during the calibration phase to control the wind turbine 10. For instance, once the wind turbine sensor has been calibrated, in an implementation phase the method 40 may involve determining a value of a control parameter of the wind turbine 10 based on the adjusted measured parameter, controlling the wind turbine 10 in accordance with the control parameter value. The control parameter may be any suitable parameter for controlling operation of the wind turbine 10. For instance, the control parameter may be a yaw angle for the wind turbine 10. In particular, the wind turbine 10 may be controlled to be aligned with the incoming wind direction, and the adjusted relative wind direction, obtained using the calibrated wind direction sensor, may be used to control the yaw angle so that the relative wind direction is 0 degrees. In a different example, the control parameter may be a rotor blade pitch control parameter, e.g. for controlling blade pitch angle to maximise power production (in full load). In a further example, the control parameter may be a torque control parameter, e.g. to limit loading on the wind turbine rotor 104. The control parameter may alternatively be an ice detection control parameter. For instance, if a determination of ice being present on the rotor blades 106 is made using the calibrated sensor (as part of a calculation of the error between measured and estimated wind speed), then the wind turbine controller may output a control signal to initiate heating of the blades or to shut down the wind turbine 10.

In some examples, in the implementation phase control of the wind turbine based on measurements from the calibrated sensor is performed irrespective of whether the wind turbine 10 is operating in wake flow conditions or not (despite the sensor having been calibrated based on data obtained in no wake flow conditions). Even having the sensor calibrated correctly for no wake conditions provides a significant increase in wind turbine performance.

In other examples, in the implementation phase control of the wind turbine is additionally based on the wake detection signal, i.e. additionally based on the flow conditions being experienced by the turbine 10 during operation. In such examples, in the implementation phase the method 40 may involve obtaining the wake detection signal indicating the wake flow experienced by the wind turbine 10 and, based on the wake detection signal, controlling the wind turbine in accordance with the control parameter value only when the wind turbine is not experiencing wake flow.

In such examples, control of the wind turbine 10 when the wake detection signal indicates that the wind turbine 10 is in wake flow conditions may be performed in different ways. In one example, a further I different value of the control parameter may be determined and used to control the wind turbine when the wind turbine 10 is experiencing wake flow. The further value of the control parameter may for instance be a default value.

In the above-described examples, in the calibration phase the sensor is calibrated based on parameter data collected during no wake flow conditions and the calibrated sensor is then used in an implementation is no wake flow and/or wake flow conditions. This could be regarded as disabling certain wind turbine functions, when the wind turbine is experiencing wake flow, or not training the wind turbine sensor I function when the wind turbine is in wake flow. In different examples, a further I different I separate calibration of the sensor may be performed based on parameter data collected when the wind turbine 10 is experiencing wake flow. This separate calibration may be used for controlling the wind turbine 10 in the implementation phase when the wind turbine 10 is experiencing wake flow. In this regard, a single further I separate calibration or correction may be determined that is to be applied for wake flow of any type, or separate calibrations I corrections may be performed for each of a plurality of defined types of wake flow, such as full wake, left-hand only wake, right-hand only wake, wake severity above a certain level, etc.

As well as the wind turbine functions described above, the described method may be applied to various other functions of a wind turbine. For instance, the method may be applied to one or more of a tilt-yaw controller, an upwind yaw controller and a converter controller.

A controller of the wind turbine 10 for performing the described method 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, and the processor(s) may execute the stored instructions to perform the method. The controller may be located in one or more locations of the wind turbine 10, e.g. in the wind turbine tower 102.

Many modifications may be made to the described examples without departing from the scope of the appended claims.

Claims

1. A method of calibrating a sensor of a wind turbine, the method comprising: obtaining over time a sensor signal, from the sensor, indicative of values of a measured parameter associated with the wind turbine; obtaining over time a monitored signal indicative of values of a monitored operational parameter associated with the wind turbine; obtaining over time a wake detection signal indicating a wake flow experienced by the wind turbine; based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and the monitored operational parameter values obtained when the wind turbine is not experiencing wake flow; based on the identified values of the measured parameter and the identified values of the monitored operational parameter, determining over time a calculated operational parameter associated with the wind turbine as a function of the measured parameter; determining a correction to the measured parameter based on the calculated operational parameter, and adjusting the measured parameter based on the correction to calibrate the sensor.

2. A method according to Claim 1 , the method comprising, in an implementation phase after a calibration phase in which the sensor is calibrated: determining a value of a control parameter of the wind turbine based on the adjusted measured parameter; and controlling the wind turbine in accordance with the control parameter value.

3. A method according to Claim 2, the method comprising, in the implementation phase: obtaining the wake detection signal indicating the wake flow experienced by the wind turbine; and based on the wake detection signal, controlling the wind turbine in accordance with the control parameter value only when the wind turbine is not experiencing wake flow.

4. A method according to Claim 2 or Claim 3, the method comprising, in the implementation phase: determining a further value of the control parameter of the wind turbine; and based on the wake detection signal, controlling the wind turbine in accordance with the further value of the control parameter when the wind turbine is experiencing wake flow.

5. A method according to Claim 4, the method comprising, in the calibration phase: based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and monitored operational parameter values obtained when the wind turbine is experiencing wake flow; based on the identified values of the measured parameter and identified values of the monitored operational parameter when the wind turbine is experiencing wake flow, determining over time a further calculated operational parameter associated with the wind turbine as a function of the measured parameter; and determining a further correction to the measured parameter based on the further calculated operational parameter, wherein, in the implementation phase, the further value of the control parameter of the wind turbine is determined based on the measured parameter adjusted according to the determined further correction.

6. A method according to Claim 2 or Claim 3, wherein the wake flow signal includes an indication of a type of wake flow being experienced by the wind turbine from a plurality of defined wake flow types, the method comprising, for each defined wake flow type, in the calibration phase: based on the sensor signal, the monitored signal and the wake detection signal, identifying the measured parameter values and monitored operational parameter values obtained when the wind turbine is experiencing the respective defined wake flow type; based on the identified values of the measured parameter and identified values of the monitored operational parameter when the wind turbine is experiencing the respective defined wake flow type, determining over time a respective wake type-specific calculated operational parameter associated with the wind turbine as a function of the wake parameter; determining a respective wake type-specific correction to the measured parameter based on the respective wake type-specific calculated operational parameter, in the implementation phase: determining a respective wake flow-specific value of the control parameter of the wind turbine based on the measured parameter adjusted according to the determined respective wake type-specific correction; and based on the wake detection signal, controlling the wind turbine in accordance with the respective wake flow-specific value of the control parameter when the wind turbine is experiencing the respective defined wake flow type.

7. A method according to Claim 6, wherein the plurality of defined wake flow types includes one or more of: full wake flow; half wake flow; left-hand wake flow; and right-hand wake flow.

8. A method according to any previous claim, wherein the sensor is a wind direction sensor, and wherein the measured parameter is wind direction relative to the wind turbine.

9. A method according to Claim 8, wherein: the monitored operational parameter is pitch angle of at least one rotor blade of the wind turbine, and the calculated operational parameter is a statistical representation of the obtained pitch angle values; or the monitored operational parameter is a wind power parameter of the wind turbine, and the calculated operational parameter is a statistical representation of the obtained wind power values, wherein the wind power parameter is one of a rotor power, a grid power, a torque and a rotor blade loading, of the wind turbine.

10. A method according to Claim 9 when dependent on Claim 2, wherein the control parameter is a yaw angle for the wind turbine.

11. A method according to any previous claim, wherein the sensor is a wind speed sensor, and wherein the measured parameter is measured wind speed.

12. A method according to Claim 11 , wherein the monitored operational parameter is estimated wind speed in the vicinity of the wind turbine, and wherein the calculated operational parameter is an error parameter indicative of a difference between measured and estimated wind speed.

13. A method according to Claim 12 when dependent on Claim 2, wherein the control parameter is one or more of: a rotor blade pitch control parameter; a torque control parameter; and an ice detection control parameter.

14. A controller for a wind turbine, the controller being configured to: obtain over time a sensor signal, from a sensor of the wind turbine, indicative of values of a measured parameter associated with wind flow in the vicinity of the wind turbine; obtain over time a monitored signal indicative of values of a monitored operational parameter associated with the wind turbine; obtain over time a wake detection signal indicating a wake flow experienced by the wind turbine; based on the sensor signal, the monitored signal and the wake detection signal, identify the measured parameter values and monitored operational parameter values obtained when the wind turbine is not experiencing wake flow; based on the identified values of the measured parameter and identified values of the monitored operational parameter, determine over time a calculated operational parameter associated with the wind turbine as a function of the measured parameter; determine a correction to the measured parameter based on the calculated operational parameter; and adjust the measured parameter based on the correction.

15. A wind turbine comprising a controller according to Claim 14.