Method and apparatus for prediction-based wind turbine control

1. A method of automatically controlling a wind turbine, comprising:

determining a time-series of measurement values of an aerodynamic flow property of a wind turbine rotor blade;

generating a predictive wind field model representing a structure of a wind field acting on the wind turbine rotor blade based on the time-series of measurement values;

generating, based on the wind-field model, a control value to produce a desired operational condition of the wind turbine rotor blade; and

adjusting the wind turbine rotor blade according to the control value.

2. The method according to claim 1 , further comprising:

determining an azimuth measurement value of the wind turbine rotor blade.

3. The method according to claim 2 , wherein the predictive wind field model is further based on the determined azimuth measurement value.

4. The method according to claim 1 , further comprising:

performing a statistical trend analysis based on the time-series of measurement values.

5. The method according to claim 4 , further comprising:

calibrating the generating of the predictive wind field model based on the statistical trend analysis.

6. The method according to claim 4 , wherein the generating of the predictive wind field model comprises machine learning based on the statistical trend analysis.

7. The method according to claim 5 , further comprising:

determining a load measurement value, and

wherein the calibrating the generating of the predictive wind field model is based on the load measurement value.

8. The method according to claim 1 , wherein the desired operational condition of the wind turbine rotor blade comprises a blade thrust.

9. The method according to claim 1 , wherein the desired operational condition of the wind turbine rotor blade comprises the blade load.

10. The method according to claim 1 , wherein the wind turbine rotor blade is adjusted cyclically.

11. A wind turbine control apparatus, comprising

a memory that comprises a time-series of measurement values of an aerodynamic flow property of, or in the vicinity of, a wind turbine rotor blade; and

a processing unit that comprises:

a wind field prediction module generates, based on the time-series of measurement values, a predictive wind field model representing a structure of a wind field acting on the wind turbine rotor blade, and

a control value generating module generates a control value based on the wind-field model,

wherein the wind turbine rotor blade is adjusted according to the control value.

12. The wind turbine control apparatus according to claim 11 , wherein the processing unit further comprises an aerodynamic property prediction module with an input from at least the wind field prediction module and an output to the control value generating module.

13. The wind turbine control apparatus according to claim 12 ,

wherein the input from the wind field prediction module is a vector of predicted wind field properties comprising forecasted values for vertical and/or horizontal linear shear and/or for horizontal-axis and/or cross-axis 2p-shear and/or for the mean wind, and

wherein the output to the control value generating module is a vector of a predicted aerodynamic property value the wind turbine rotor blade.

14. The wind turbine control apparatus according to claim 13 , wherein that the vector of the predicted aerodynamic property value comprises a blade load value and/or a blade thrust value.

15. The wind turbine control apparatus according to claim 14 , wherein the control value generating module performs a blade thrust control.

16. The wind turbine control apparatus according to claim 14 , wherein the control value generating module performs blade load control.

17. The wind turbine control apparatus according to claim 13 ,

wherein the memory further comprises a rotor blade azimuth measurement value, and

wherein the predictive wind field model is further based on the determined azimuth measurement value.

18. The wind turbine control apparatus according to claim 13 , wherein the wind field prediction module self-calibrates on the time-series of measurement values.

19. The wind turbine control apparatus according to claim 13 ,

wherein the memory comprises a load measurement value, and

wherein the wind field prediction module to self-calibrates based on the load measurement value.

20. The wind turbine control apparatus according to claim 11 , wherein the memory further comprises a reference control signal which is a pitch angle reference control signal which is added to an output of the control value generating module.