Method and system for predicting the occurrence of a wind gust at a wind turbine

1. A method for predicting an impending wind gust at a wind turbine comprising:

obtaining as input data, with a predetermined frequency, measured values of two or more parameters selected from wind speed, rotor speed and generator power of the wind turbine;

selecting a group of input data of predetermined size consisting of the most recently obtained two or more selected parameters;

calculating, via a computing device having a processor, the probability of an occurrence of a wind gust on the basis of said selected group of input data by determining the resemblance of the selected group of input data with data of the two or more selected parameters, wherein the data of the two or more selected parameters being used to determine the resemblance comprises predetermined patterns of the two or more selected parameters during one or more predetermined wind gusts conditions;

verifying if the calculated probability of the occurrence of a wind gust is above a predetermined probability level; and

predicting the occurrence of a wind gust if the calculated probability is above said predetermined probability level.

2. The method of claim 1 , wherein the predetermined patterns to determine the resemblance are predetermined wind gusts as defined in the standard IEC 61400.

3. The method of claim 1 , wherein determining the resemblance of the input data with data of the selected parameters relating to the one or more predetermined wind gust comprises using an algorithm obtained with an artificial neural network.

4. The method of claim wherein the two or more selected parameters are wind speed measured at the nacelle of the wind turbine and generator power.

5. The method of claim 1 , wherein said predetermined frequency is one of between 10 and 1000 times per second, between 10 and 100 times per second, or approximately 20 times per second.

6. The method of claim 1 , wherein said group of input data of predetermined size covers a period of between 2 and 8 seconds.

7. The method of claim 1 , wherein said predetermined probability level is between approximately 40% and approximately 85%.

8. The method of claim 1 , further comprising:

when the occurrence of a wind gust is predicted, adapting one of a pitch control and generator torque of the wind turbine to such a wind gust.

9. A non-transitory computer-readable medium having computer-executable instructions embodied thereon that, when executed by a computing device having a processor, perform a method of predicting the occurrence of a wind gust according to claim 1 .

10. A wind turbine, comprising:

a wind turbine control system adapted to

obtain as input data, with a predetermined frequency, measured values of two or more parameters selected from wind speed, rotor speed and generator power of the wind turbine;

select a group of input data of predetermined size consisting of the most recently obtained two or more selected parameters;

calculate the probability of an occurrence of a wind gust on the basis of said selected group of input data by determining the resemblance of the selected group of input data with data of the two or more selected parameters, wherein the data of the two or more selected parameters being used to determine the resemblance comprises predetermined patterns of the two or more selected parameters during one or more predetermined wind gusts conditions,

verify if the calculated probability of the occurrence of a wind gust is above a predetermined probability level; and

predict the occurrence of a wind gust if the calculated probability is above said predetermined probability level.