Temperature estimation in wind turbines

Abstract

A method of estimating a temperature of a component of a wind turbine is provided, the method comprising, during a calibration period, receiving a measurement of a temperature of the component and a measurement of one or more corresponding operating parameters; and using the measurements to calculate coefficients of a model of the temperature of the component. The model associates a temperature at a current time with a temperature T n—1 at a previous time. The model is divided into different bins based on a wind speed or power produced by the wind turbine. The method further comprises using the model to estimate a temperature of the wind turbine.

Description

Technical Field

[0001] This invention relates to wind turbines, and more particularly to the estimation of temperature of components in wind turbines. Background Technology

[0002] Knowing the temperature of the various components of a wind turbine is crucial to ensuring they operate as intended and to prevent damage to any part of the turbine. Typically, multiple temperature sensors are used to track the temperature of a given component. Using multiple sensors allows for comparison and verification of measurements and provides redundancy in case of failure.

[0003] However, one or more of these temperature sensors may fail during the turbine's lifespan and may be difficult to replace. Without knowing the operating temperatures of critical components, it may be necessary to take the turbine offline.

[0004] Therefore, alternative measurements for the temperature of components are needed. Summary of the Invention

[0005] A first aspect of the present invention provides a method for estimating the temperature of a component of a wind turbine, the method comprising:

[0006] During the calibration cycle:

[0007] Receives a measurement of the temperature of the component as measured by the temperature sensor of the wind turbine;

[0008] Receive measurements of one or more operating parameters of the wind turbine corresponding to the temperature measurement, the one or more operating parameters including at least a measurement of wind speed or power generated by the wind turbine;

[0009] The coefficients of the model for the temperature of the component are calculated using the measurement of the temperature and the measurement of one or more operating parameters, wherein:

[0010] The model will determine the temperature T of the component at the current time. n The values ​​of one or more operating parameters at the current time and the temperature T of the component at the previous time. n-1 Related;

[0011] Based on the wind speed or power generated by the wind turbine, the model is divided into different boxes, such that for each box, the model includes the temperature T of the wind turbine at the current time. n The values ​​of the one or more operating parameters at the current time and the temperature T of the component at the previous time. n-1 The associated corresponding coefficients; and

[0012] computing coefficients of the model comprises assigning each measurement of the temperature and the one or more operating parameters to one or more bins of the model, and fitting the coefficients of the respective bins to the measurements assigned to the bins; and

[0013] using the model to estimate the temperature of the component of the wind turbine.

[0014] In some embodiments, using the model to estimate the temperature of the component of the wind turbine further comprises the steps of:

[0015] during an operating period:

[0016] receiving a measurement of the one or more operating parameters taken at a first time;

[0017] assigning the measurement of the one or more operating parameters to a bin of the model based on the wind speed or power produced at the first time; and

[0018] inputting the measurement of the one or more operating parameters into the model to estimate the temperature of the component at the first time T x-1 using coefficients of the bin to which the measurement is assigned and an estimate of the temperature T x at a time prior to the first time.

[0019] The method can further comprise operating the wind turbine in dependence on the estimate of the temperature of the component at the first time.

[0020] In some embodiments, the method can further comprise estimating an uncertainty of the temperature of the component at the first time, wherein the uncertainty is based on a statistical uncertainty of the estimate of the temperature at the first time and on an uncertainty at a time prior to the first time.

[0021] In some embodiments, the method can further comprise the steps of, during the operating period:

[0022] receiving a measurement of the temperature of the component measured by a temperature sensor of the wind turbine at the first time; and

[0023] comparing the measurement to the estimate of the temperature at the first time to validate the measurement.

[0024] In some such embodiments, the method can further comprise the steps of:

[0025] determining that a difference between the measurement of the temperature and the estimate of the temperature exceeds a predetermined threshold; and

[0026] determining that the measurement from the temperature sensor is invalid.

[0027] In some embodiments, using the model to estimate a temperature of the component of the wind turbine can comprise or further comprise: calculating a steady state temperature of the component from the model. Such embodiments can comprise comparing the steady state temperature of the component of the wind turbine to a corresponding steady state temperature from one or more other wind turbines. Some embodiments can further comprise identifying an anomaly in the wind turbine based on the comparison. Alternatively or additionally, the method can comprise comparing the steady state temperature to an expected steady state temperature of the wind turbine.

[0028] In some embodiments, the component can be at least one of: a generator; a generator winding; a transformer; a transformer winding; a gearbox; gearbox oil; hydraulic oil; a converter; one or more bearings; and a cooling water system.

[0029] In some embodiments, the component can be a generator or a generator winding, the one or more operating parameters can further comprise at least one of: a generator rotor speed; a generator voltage; and a reactive power.

[0030] In some embodiments, the model can relate a temperature of the generator or generator winding at a current time to:

[0031] a square of a current power generated by the turbine;

[0032] a current generator rotor speed; and

[0033] a temperature of the generator or generator winding at a previous time.

[0034] A second aspect of the application provides a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method of any of the embodiments of the first aspect.

[0035] A third aspect of the application provides a controller for a wind turbine, the controller comprising a processor and a memory; wherein the controller is configured to receive measurements of operating parameters from one or more sensors of the wind turbine; and wherein the memory stores instructions which, when executed by the processor, cause the processor to carry out the method of any of the embodiments of the first aspect.

[0036] A fourth aspect of the present application provides a wind turbine comprising one or more sensors for measuring an operating parameter during operation of the wind turbine; and a controller according to any one of the embodiments of the third aspect of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0037] Embodiments of the present application will now be described with reference to the accompanying drawings, in which

[0038] Figure 1 is a schematic illustration of a wind turbine;

[0039] Figure 2 is a schematic illustration of a control system of a wind turbine; and

[0040] Figure 3 illustrates a method of estimating a temperature of a component of a wind turbine. DETAILED DESCRIPTION

[0041] Figure 1 An example of a wind turbine 100 is illustrated in a schematic perspective view. The wind turbine 100 comprises a tower 102, a nacelle 103 at the top of the tower, and a rotor 104 operatively coupled with a generator housed within the nacelle 103. In addition to the generator, the nacelle houses a variety of components required for converting wind energy into electrical energy and various components required for operating, controlling and optimizing the performance of the wind turbine 100. The rotor 104 of the wind turbine comprises a central hub 105 and a plurality of blades 106 protruding outwardly from the central hub 105. In the illustrated embodiment, the rotor 104 comprises three blades 106, but the number can vary. Furthermore, the wind turbine comprises a control system. The control system can be placed within the nacelle, or distributed at a plurality of locations within the turbine and communicatively connected.

[0042] The wind turbine 100 can be included in a collection of other wind turbines belonging to a wind power plant (also referred to as a wind farm or wind park) which functions as a power plant connected to an electrical grid via transmission lines. The electrical grid is typically made up of a network of power stations, transmission circuits, and substations coupled by a network of transmission lines that transmit power to loads in the form of electric utilities and other customers.

[0043] Figure 2Embodiments of the control system 200 and elements of the wind turbine are schematically illustrated. The wind turbine includes rotor blades 106 that are mechanically connected to a generator 202 via a gearbox 203. In direct drive systems and other systems, the gearbox 203 can be absent. Electrical power generated by the generator 202 is injected into an electrical grid 204 via an electrical converter 205. The generator 202 and converter 205 can be based on a full scale converter (FSC) architecture or a doubly fed induction generator (DFIG) architecture, although other types can be used.

[0044] The control system 200 includes a number of elements, including at least one master controller 220 having a processor and memory, such that the processor is capable of performing computational tasks based on instructions stored in the memory (such as the methods discussed below). In general, the wind turbine controller ensures that the wind turbine produces the required power output level in operation. This is achieved by adjusting the pitch angle of the blades 106 and / or the power extraction of the converter 205. To this end, the control system includes a pitch system including a pitch controller 207 using a pitch reference 208, and a power system including a power controller 209 using a power reference 206. The wind turbine rotor includes rotor blades that are capable of being pitched by a pitch mechanism. The rotor includes individual pitch systems that are capable of individually pitching the rotor blades, and can include a common pitch system that adjusts all of the pitch angles on all of the rotor blades simultaneously. The control system or elements of the control system can be placed in a power plant controller (not shown) such that the turbine can operate based on externally provided instructions.

[0045] The control system 200 also includes a number of temperature sensors 230 (only one is shown for clarity, Figure 2 The temperature sensors 230 are positioned at various locations in the wind turbine 100 to measure the temperature of particular components of the turbine, such as the generator windings. The temperature sensors 230 provide temperature measurements to the controller 220, which monitors the temperature of various components to ensure proper operation of the turbine, and can adjust the operation of the turbine based on the temperature readings. For example, if a component becomes too hot, the controller 220 can control the operation of the turbine 100 to reduce the temperature of the component (e.g., reduce the physical or electrical load on the component) to prevent permanent damage to the wind turbine 100.

[0046] Ideally, the temperature sensors 230 will function properly for the entire lifetime of the turbine 100. In practice, however, it is likely that the sensors 230 will fail. Repairing or replacing a failed temperature sensor 230 can be difficult and expensive, for example requiring the turbine 100 to be shut down. However, continued operation of the wind turbine 100 requires knowledge of the component temperatures, leaving the operator with no choice but to repair the sensor 230 or risk permanent damage to the turbine 100.

[0047] Figure 3 A method 300 of estimating the temperature of components of a wind turbine 100 is exemplified. Using the method 300, the temperature can be estimated from current operating parameters of the wind turbine, for example the turbine power output. This estimated temperature can be used to virtually replace one or more failed temperature sensors, allowing the turbine 100 to continue safe operation.

[0048] The method 300, as exemplified below, comprises two different phases. In a first calibration phase, data measured on a particular wind turbine 100 is used to calculate parameters of a model relating component temperature to operating parameters. Importantly, the model also relates current component temperature to previous temperatures, to account for thermal inertia. In a second operational phase, the model is used to estimate component temperature based on current operating parameters of the same turbine 100.

[0049] As used herein, temperature can refer to the absolute or relative temperature of a component. For example, the temperature can be relative to a reference temperature such as the current cooling water temperature.

[0050] The method 300 begins at step 301, in which a measurement of the temperature of a component is measured by one or more temperature sensors 230.

[0051] The method then proceeds to step 302, in which a measurement of one or more operating parameters of the wind turbine corresponding to the measurement of temperature is measured. The one or more operating parameters include at least a measurement of the wind speed or power produced by the wind turbine, and can also include operating parameters such as generator speed. As will be appreciated by those skilled in the art, each of these measurements can be performed by appropriate sensors on the wind turbine, for example sensors reporting to a data acquisition and monitoring system.

[0052] The measurements of the operating parameters correspond to the measurements of the temperature in that each measurement represents the value of the operating parameter at the time the corresponding temperature measurement was made. For example, the corresponding measurements can be made at the same or similar times, e.g. within a predetermined time period of each other, e.g. within 1 minute or within 5 minutes. For data acquisition and monitoring data, where measurements are typically made and averaged over a 10 minute period, the temperature and operating parameters can correspond as they are from the same data acquisition and monitoring data period.

[0053] The measurements performed in steps 301 and 302 can be passed to the turbine controller 220, which can perform the further steps detailed below. Alternatively, the measurements can be passed to an external system, which can perform the steps below on behalf of the wind turbine 100.

[0054] At step 303, the coefficients of a model of the temperature of the component are calculated using the measurements of the temperature and the operating parameters. The model is of the form:

[0055] T n ~ F(x, y, z, T n-1 ).

[0056] Here, T n represents the temperature at the current time. T n-1 represents the temperature at the previous time, in particular the most recent temperature measurement / estimate. x, y, z represent the values of the operating parameters at the current time, e.g. power generation. Although three operating parameters are represented, any number of parameters can be used. For example, the model can relate only the current temperature to power / wind speed and the previous temperature. An example model for the case of generator winding temperature is discussed in more detail below.

[0057] The coefficients of the model can be found using regression techniques or any method capable of deriving from training data, i.e. from the measurements made in steps 301 and 302. For example, machine learning or probabilistic modelling techniques can be applied to calculate the coefficients of the model.

[0058] The model is divided into a number of bins based on the wind speed or power generated by the wind turbine 100. The model includes a respective coefficient for each bin. For example, consider a simple model, e.g.:

[0059] T n ~ a i (PWR) 2 + b i T n-1 , (1)

[0060] where PWR represents the output power of the turbine 100. Here, a i , bi are the square of the power and the coefficients associated with the previous and current temperatures, respectively.

[0061] Based on the generated power, the model is divided into a number of bins. A first bin can be used for output power in the range 100-200 kW; a second bin can be used for output power in the range 200-300 kW, and so on. Each bin includes respective coefficients a i , b i for that bin. Thus, for the first bin, the model is:

[0062] T n ~ a1(PWR) 2 + b1T n-1 , (2)

[0063] where a1, b1are the respective coefficients for the first bin. Likewise, for the second bin, the model is:

[0064] T n ~ a2(PWR) 2 + b2T n-1 , (3)

[0065] where a2, b2are the respective coefficients for the second bin.

[0066] When the coefficients are calculated in step 303, each temperature and operating parameter is assigned to a bin based on the power / wind speed at the time the measurement was made. Using the regression techniques described above, the measurements assigned to a bin are only used to calculate the coefficients for that bin.

[0067] Thus, the model effectively comprises a number of different temperature models, each adapted to a particular power / wind speed range. It has been found that this division of the model provides a more accurate temperature estimate than a global model for all power / wind speeds.

[0068] In some embodiments, the bins can overlap. For example, a first bin can be used for output power in the range 80-220 kW, a second bin for power in the range 180-320 kW, a third bin for power in the range 280-420 kW, and so on. In such embodiments, each measurement of a temperature and operating parameter can be assigned to a plurality of bins in step 303, where the wind speed / power output at the time of those measurements is covered by those plurality of bins. The measurements are then used to find the coefficients for each bin to which they are assigned.

[0069] The number of bins used and / or the amount of overlap between bins can be chosen based on the amount of data available for the generation of the model. For example, if data for a year or more is available, then a large amount of data is available to allow for bins with a lot of overlap. For example, the bins can be spaced between 5kW and 50kW, for example 10kW (i.e. the spacing of the bin centres is 10kW). In this case, the width of the bins can be ±500kW of the bin centre (or selected from a range of ±100kW to ±1000kW). In this case, the large amount of data available in each bin provides robustness to the generated model. When the parameters of the model for each bin have been determined, weighting factors such as Gaussian weighting factors can be applied to the data within the bins so that the data at the centre of the bin has a greater influence on the resulting parameters than the data at the edges of the bin.

[0070] It should be noted that the models provided in equations (1) to (3) are merely illustrative. Any number of operating parameters can be used in the model, selected to suit the particular component being modelled, as will be understood by the skilled person. Furthermore, any model can include a constant offset coefficient calculated in step 303 with a coefficient of the operating parameter.

[0071] Once the coefficients of the bins of the model have been determined, the calibration period is complete and the model is used to estimate the temperature of the component. In particular, the method 300 can move to an operational period, which begins at step 304. The method 300 can proceed immediately to step 304, or there can be a delay. For example, the method 300 can proceed to step 304 only when a temperature estimate is required, for example when the temperature sensor 230 fails.

[0072] At step 304, measurements of the operating parameters (i.e. the parameters used in the model) are taken at a first time t x (i.e. the current time). These measurements are taken in the same way as step 302 discussed above.

[0073] At step 305, the measurements taken in step 304 are assigned to bins of the model based on the output power / wind speed at the first time t x In the case where overlapping bins are used, as discussed above, each measurement taken in step 304 can be assigned to the bin in which the measurement is closest to the bin centre point.

[0074] At step 306, the measurements are input into the model to estimate the temperature T x at the first time t x-1 using the values of the operating parameters at the first time t x-1 and the estimate T x of the temperature at the previous time t xFor example, for the simple exemplary model shown in equations (1) to (3) above, if the output power at t x is within the range of the first bin, then the operating parameter at t x is measured and the estimate of the temperature at the previous time t x-1 is input into equation (2) to estimate the temperature T x of the component at the first current time t x-1 The previous time t x-1 may in particular be the time of the last available operating parameter measurement. For example, SCADA data from a wind turbine is typically measured and averaged into 10 minute data periods. The first time can correspond to the most recent 10 minute data period, while the previous time can correspond to the immediately preceding 10 minute data period.

[0075] The estimate T x-1 of the temperature at the previous time t x-1 may itself be generated using the method 300, which uses a measurement of the operating parameter at the previous time t x-2 and an estimate T x-2 of the temperature at a time t x-1 preceding that. The temperature T x may for example be stored in a memory, for example a memory associated with the turbine controller 220, and can be retrieved from the memory as part of step 306 to estimate the temperature T x at the first time t x-1 .

[0076] Alternatively, in particular where no estimate from a previous run of the method 300 is available, a value can be chosen for the temperature T x-1 at the previous time t x-1 . For example, a preset "start-up" value such as 50°C can be chosen, or an estimate can be made based on past observations of typical temperatures in the wind turbine 100 or similar wind turbines. A temperature measurement made by the temperature sensor 230 can be used as the previous temperature T x , for example the last temperature measurement made before the sensor 230 failed.

[0077] Once the temperature T x at the first current time has been estimated, the method 300 can end. Alternatively, steps 304 to 306 can be run continuously, updating the estimate of the temperature for each new measurement period based on the temperature estimate in the previous measurement period. In either case, the estimate of the temperature can be used to inform operation of the wind turbine 100. For example, if the temperature becomes too hot, the controller 220 can change the operation of the wind turbine 100 to reduce the temperature of the component.

[0078] In some implementations, an uncertainty value can also be determined for each estimated temperature value. Uncertainty value σ x It can be based on the statistical uncertainty ∑ from the model used for that single temperature, and the uncertainty σ based on previous temperature estimates (i.e., the previous temperature estimates used in generating the current temperature estimate). x-1 Of course. Specifically, the temperature estimate T at the first time point. x Uncertainty σ x It can be calculated as:

[0079]

[0080] Here, b i It is T x With the generation of T x The specific box of the model T x-1 The associated coefficients. Therefore, for example, in the simple model above, when the temperature is calculated from equation (2) using the first box of the model, the coefficient b1 will be used in the uncertainty calculation.

[0081] In this way, the inherent thermal inertia of the components is reflected in the uncertainty, which provides a more valuable indication of the uncertainty of a single temperature estimate compared to statistical uncertainty alone. This improved uncertainty can then be used to inform decisions regarding turbine operation.

[0082] The temperature (and uncertainty) provided by method 300 can be used to verify temperature measurements from sensor 230, for example, in cases where other temperature sensors 230 have failed and no redundancy has been provided. In this case, one or more measurements from the (remaining) sensor 230 can be received and compared with the corresponding temperature estimate generated from method 300. If the difference between the actual measurement and the estimate exceeds a threshold, it can be determined that there is a fault in sensor 230 or a physical problem (e.g., overheating) in the component. The threshold can be a simple predetermined threshold based on the absolute temperature difference between the actual measurement and the estimated measurement. Alternatively, the uncertainty value discussed above can be used. The threshold can be a value where the deviation of the actual temperature measurement from the estimated measurement exceeds a predetermined multiple of the uncertainty. For example, if the actual measurement is greater than 1.5σ from the estimated temperature. x or 2σ x Then it can be determined that there is a problem.

[0083] The model generated as part of method 300 can also be used to determine the steady-state temperature for a given set of fixed operating parameters. Specifically, after the coefficients of each bin of the model have been determined in step 303 of method 300, for the fixed operating parameters, the steady-state temperature can be determined for case T. n =Tn-1 The equations of the model are solved to determine the steady state temperature. The calculation of the steady state temperature can be performed without estimating any operating temperatures, in other words, steps 304 to 306 of the method 300 can be omitted when the model is used to determine the steady state temperature.

[0084] In such an embodiment, a single steady state temperature can be calculated using the coefficients of a single selected bin of the model and the selected operating parameter. Alternatively, a plurality of steady state temperatures can be calculated for a range of different output powers / wind speeds (and thus different bins of the model) and a range of other operating parameters included in the model, effectively generating a steady state temperature map.

[0085] The steady state temperature calculated in this way can be used to compare the performance of the turbine 100 to other turbines, or to the expected performance from the design of the turbine 100, to determine whether the turbine 100 is operating correctly. For example, the steady state temperature for a single turbine 100 can be compared to the average behaviour of corresponding steady state temperatures from a collection of similar turbines (e.g. same model, similar location, etc.). As altitude will affect temperature, the collection can only include turbines at the same or similar altitude as the single turbine 100, or an altitude correction factor can be determined from a regression of the steady state temperatures of a population of turbines against altitude, which allows turbines at different altitudes to be compared. As an alternative to building a machine population model, all the data measured on a collection of wind turbines can be pooled and used to calculate the coefficients of one (binned) collective model, which will then provide the average behaviour of the collection and be able to be compared to the results from the single turbine 100 to identify anomalies in the performance of that single turbine 100.

[0086] In any of the embodiments discussed above, the component can be at least one of: a generator; a generator winding; a transformer; a transformer winding; a gearbox; gearbox oil; hydraulic oil; a converter; one or more bearings; and a cooling water system. In general, the component can be any component of the wind turbine whose temperature behaviour exhibits thermal inertia.

[0087] In a particular example, the method 300 can be applied to estimate the temperature of a generator winding of the wind turbine 100. In this case, the operating parameters used as part of the model can include the output power and at least one of the generator rotor speed, the generator voltage and the reactive power. For example, the model can take the form:

[0088] T n ~ c i (PWR) 2 + d i (RPM) + e iV + f i + g i T n-1 , (5)

[0089] where PWR is the turbine output power, RPM is the generator speed, V is the generator voltage, and f i is a constant offset factor. In some variations of this model, V can be omitted.

[0090] Any of the above methods can be implemented as a computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out any of the above methods. In particular, the computer program can be stored on a turbine controller such as the controller 220 and can cause the processor of the controller to carry out any of the above methods.

[0091] While the application has been described above by reference to one or more preferred embodiments, it should be understood that various changes or modifications can be made without departing from the scope of the application, which is defined by the following claims.

Claims

1. A method for estimating the temperature of a component of a wind turbine, the method comprising: During the calibration cycle: Receives a measurement of the temperature of the component as measured by the temperature sensor of the wind turbine; Receive measurements of one or more operating parameters of the wind turbine corresponding to the measurement of the temperature, the one or more operating parameters including at least a measurement of wind speed or power generated by the wind turbine; The coefficients of the model for the temperature of the component are calculated using the measurement of the temperature and the measurement of one or more operating parameters, wherein: The model will determine the temperature T of the component at the current time. n The values ​​of one or more operating parameters at the current time and the temperature T of the component at the previous time. n-1 Related; Based on the wind speed or power generated by the wind turbine, the model is divided into different boxes, such that for each box, the model includes the temperature T of the wind turbine at the current time. n The values ​​of the one or more operating parameters at the current time and the temperature T of the component at the previous time. n-1 The associated corresponding coefficients; and Calculating the coefficients of the model includes: assigning each measurement of the temperature and one or more operating parameters to one or more bins of the model, and fitting the coefficients of each bin to the measurements assigned to that bin; and The model is used to estimate the temperature of the components of the wind turbine.

2. The method according to claim 1, wherein, Using the model to estimate the temperature of the components of the wind turbine includes: During the operating cycle: Receive measurements of one or more operating parameters acquired in a first-time event; Based on the wind speed or power generated at the first time, the measurements of the one or more operating parameters are assigned to the box of the model; and The measurements of one or more operating parameters are input into the model to use the coefficients of the chamber to which the measurements are assigned and the estimated temperature T at the time prior to the first time. x-1 To estimate the temperature T of the component at the first time point. x .

3. The method according to claim 2, further comprising: The uncertainty of the temperature of the component at the first time point is estimated, wherein the uncertainty is based on the statistical uncertainty of the estimated temperature at the first time point and the uncertainty based on the time prior to the first time point.

4. The method according to claim 2 or 3, further comprising, during the operation cycle: Receives a measurement of the temperature of the component as measured by the temperature sensor of the wind turbine at the first time; and The measurement is compared with the estimated temperature at the first time point to verify the measurement.

5. The method according to claim 4, further comprising: The difference between the temperature measurement and the temperature estimate exceeds a predetermined threshold. as well as The measurement from the temperature sensor was determined to be invalid.

6. The method according to claim 1, wherein, Using the model to estimate the temperature of the components of the wind turbine includes calculating the steady-state temperature of the components from the model.

7. The method according to claim 6, further comprising: The steady-state temperature of the component of the wind turbine is compared with the corresponding steady-state temperature from one or more other wind turbines.

8. The method according to claim 7, further comprising: The anomalies in the wind turbine are identified based on the comparison.

9. The method according to any one of claims 6 to 8, further comprising: The steady-state temperature is compared with the expected steady-state temperature of the wind turbine.

10. The method according to any one of claims 1 to 3, wherein, The component is at least one of the following: a generator; a generator winding; a transformer; a transformer winding; a gearbox; gearbox oil; hydraulic oil; a converter; one or more bearings; and a cooling water system.

11. The method according to any one of claims 1 to 3, wherein, The component is a generator or generator winding, and the one or more operating parameters further include at least one of generator rotor speed, generator voltage, and reactive power.

12. The method according to claim 11, wherein, The model correlates the temperature of the generator or generator winding at the current time with the following: The square of the current power generated by the turbine; Current generator rotor speed; and The temperature of the generator or generator winding at the previous time.

13. A computer program product, the computer program of the computer program product comprising instructions that, when the computer program is executed by a computer, cause the computer to perform the method of any one of claims 1 to 12.

14. A controller for a wind turbine, the controller including a processor and a memory; in, The controller is configured to receive measurements of operating parameters from one or more sensors of the wind turbine; and The memory stores instructions that, when executed by the processor, cause the processor to perform the method of any one of claims 1 to 11.

15. A wind turbine comprising: One or more sensors, the one or more sensors being used to measure operating parameters during operation of the wind turbine; and The controller according to claim 14.

Images