Forecasting output power of wind turbine in wind farm

1. A method for forecasting an output power of a plurality of wind turbines in a wind farm, the method comprising:

categorizing a first plurality of wind turbines in a wind farm that are densely packed in a first group, each of the first plurality of wind turbines being within a density threshold distance of at least one of the first plurality of wind turbines;

categorizing a second plurality of wind turbines in the wind farm that are sparsely packed in a second group, each of the second plurality of wind turbines being at least outside the density threshold distance of the plurality of wind turbines in the wind farm;

a high resolution wind turbine density dependent weather forecasting model providing a high resolution weather forecast, the high resolution wind turbine density dependent weather forecasting model generating the high resolution weather forecast using real-time environmental data collected from first meteorological sensors at first grid points in a first geographic area of the first plurality of wind turbines, each of the first meteorological sensors being within a dense grid point threshold distance to adjacent meteorological sensors of the first meteorological sensors;

a low resolution wind turbine density dependent weather forecasting model providing a low resolution weather forecast, the low resolution wind turbine density dependent weather forecasting model generating the low resolution weather forecast using real-time environmental data collected from second meteorological sensors at second grid points in a second geographic area of the second plurality of wind turbines, each of the second meteorological sensors being outside of the dense grid point threshold distance to adjacent meteorological sensors of the second meteorological sensors;

collecting real-time environmental data from at least one sensor in the wind farm, wherein the at least one sensor comprises a meteorological sensor at a wind tower in the wind farm; and

for each of the plurality of wind turbines in the wind farm:

generating a corrected data set based on the real-time environmental data, wherein the corrected data set includes a measure of wind direction at a hub-height of each of the wind turbines in the wind farm that is calculated without the use of a wind direction sensor disposed at the respective wind turbine;

correcting, for each of the wind turbines categorized as the first plurality of wind turbines in the wind farm, the high resolution wind turbine density dependent weather forecasting model by using the corrected data set to produce a corrected weather forecasting model;

correcting, for each of the wind turbines categorized as the second plurality of wind turbines in the wind farm, the low resolution wind turbine density dependent weather forecasting model by using the corrected data set to produce the corrected weather forecasting model;

obtaining a forecast value of wind information at each of the plurality of wind turbines in the wind farm based on the corrected weather forecasting model; and

forecasting output power of each of the plurality of the wind turbines based on the forecast value and a power forecasting model.

2. The method according to claim 1 , wherein the at least outside the density threshold is greater than a sparse threshold distance data.

3. The method according to claim 2 , wherein the wind information comprises wind direction and wind velocity, and

the measure of wind direction is obtained based on at least one of:

calculating the wind direction based on yaw angle of the wind turbine in the wind farm;

calculating the wind direction based on wind direction at the wind tower in the wind farm;

obtaining the wind direction based on fluid dynamics analysis; and

obtaining the wind direction based on power curve deviation analysis.

4. The method according to claim 1 , wherein the correcting the high resolution wind turbine density dependent weather forecasting model by using the corrected data set comprises:

correcting the high resolution wind turbine density dependent weather forecasting model by using the corrected data set according to a Hybrid Data Assimilation method.

5. The method according to claim 4 , wherein the correcting the high resolution wind turbine density dependent weather forecasting model by using the corrected data set comprises:

selecting a data member from the corrected data set; and

correcting the high resolution wind turbine density dependent weather forecasting model based on the data member in each round.

6. The method according to claim 5 , wherein the selecting the data member from the corrected data set comprises:

filtering a plurality of data members in the corrected data set according to the first grid points in the high resolution wind turbine density dependent weather forecasting model.

7. The method according to claim 6 , wherein the correcting the high resolution wind turbine density dependent weather forecasting model based on one of the data member in each round further comprises:

sorting the at least one data member of the plurality of data members according to a dependence in the at least one data member in the corrected data set; and

using each data member to correct the high resolution wind turbine density dependent weather forecasting model in a sequence.

8. The method according to claim 1 , wherein the at least one sensor in the wind farm comprises:

a wind velocity sensor at the wind turbine in the wind farm.

9. The method according to claim 1 , wherein the power forecasting model is at least one of:

a power curve of the wind turbine; and

a function related to a plurality of properties of the wind turbine, air density and the forecast value.

10. An apparatus for forecasting output power of a wind turbine in a wind farm, the apparatus comprising:

one or more computer processors;

one or more computer readable storage media;

program instructions stored on the one or more computer readable storage media for executing by at least one of the one or more processors, the program instructions comprising:

categorizing a first plurality of wind turbines in a wind farm that are densely packed in a first group, each of the first plurality of wind turbines being within a density threshold distance of at least one of the first plurality of wind turbines;

categorizing a second plurality of wind turbines in the wind farm that are sparsely packed in a second group, each of the second plurality of wind turbines being at least outside the density threshold distance of the plurality of wind turbines in the wind farm;

a high resolution wind turbine density dependent weather forecasting model providing a high resolution weather forecast, the high resolution wind turbine density dependent weather forecasting model generating the high resolution weather forecast using real-time environmental data collected from first meteorological sensors at first grid points in a first geographic area of the first plurality of wind turbines, each of the first meteorological sensors being within a dense grid point threshold distance to adjacent meteorological sensors of the first meteorological sensors;

a low resolution wind turbine density dependent weather forecasting model providing a low resolution weather forecast, the low resolution wind turbine density dependent weather forecasting model generating the low resolution weather forecast using real-time environmental data collected from second meteorological sensors at second grid points in a second geographic area of the second plurality of wind turbines, each of the second meteorological sensors being outside of the dense grid point threshold distance to adjacent meteorological sensors of the second meteorological sensors;

for each of the plurality of wind turbines in the wind farm:

generating a corrected data set based on the real-time environmental data, wherein the corrected data set includes a measure of wind direction at a hub-height of each of the wind turbines in the wind farm that is calculated without the use of a wind direction sensor disposed at the respective wind turbine;

correcting, for each of the wind turbines categorized as the first plurality of wind turbines in the wind farm, the high resolution wind turbine density dependent weather forecasting model by using the corrected data set to produce a corrected weather forecasting model;

correcting, for each of the wind turbines categorized as the second plurality of wind turbines in the wind farm, the low resolution wind turbine density dependent weather forecasting model by using the corrected data set to produce the corrected weather forecasting model;

obtaining a forecast value of wind information at each of the plurality of wind turbines in the wind farm based on the corrected weather forecasting model; and

forecasting output power of each of the plurality of the wind turbines based on the forecast value and a power forecasting model.

11. The apparatus of claim 10 , wherein the at least outside the density threshold is greater than a sparse threshold distance.

12. The apparatus of claim 11 , wherein:

the wind information comprises wind direction and wind velocity; and

the measure of wind direction is obtained based on at least one of:

calculating the wind direction based on a yaw angle of the wind turbine in the wind farm;

calculating the wind direction based on a wind direction at the wind tower in the wind farm;

obtaining the wind direction based on a fluid dynamics analysis; and

obtaining the wind direction based on a power curve deviation analysis.

13. The apparatus of claim 10 , wherein the program instruction of correcting the high resolution wind turbine dependent weather forecasting model by using the corrected data set comprises:

correcting the high resolution wind turbine density dependent weather forecasting model by using the corrected data set according to a Hybrid Data Assimilation method.

14. The apparatus of claim 13 , wherein the program instruction of correcting the high resolution wind turbine density dependent weather forecasting model by using the corrected data set comprises:

selecting a data member from the corrected data set; and

correcting the high resolution wind turbine density dependent weather forecasting model based on the data member in each round.

15. The apparatus of claim 14 , wherein the program instruction of selecting the data member from the corrected data set comprises:

filtering a plurality of data members in the corrected data set according to the first grid points in the high resolution wind turbine density dependent weather forecasting model.

16. The apparatus of claim 15 , wherein the program instruction of correcting the high resolution wind turbine density dependent weather forecasting model based on one of the data member in each round further comprises:

sorting the at least one data member of the plurality of data members according to a dependence in the at least one data member in the corrected data set; and

using each data member to correct the high resolution wind turbine density dependent weather forecasting model in a sequence.

17. The apparatus of claim 10 , wherein the at least one sensor in the wind farm comprises:

a wind velocity sensor at a wind turbine in the wind farm.

18. The apparatus of claim 10 , wherein the power forecasting model is at least one of:

a power curve of the wind turbine; and

a function related to a plurality of properties of the wind turbine, air density, and the forecast value.