Operating a wind turbine with sensors implemented by a trained machine learning model

Patent metadata

Field	Value
Publication number	US-11867154-B2
Application number	US-202017132942-A
Country	US
Kind code	B2
Filing date	Dec 23, 2020
Priority date	Dec 23, 2019
Publication date	Jan 9, 2024
Grant date	Jan 9, 2024

Abstract

The present invention relates to a method of operating a wind turbine with an operational parameter where values of the operational parameter are obtained by different sensors and compared to determine the validity of the value. A first value and a second value of the operational parameter are obtained different sensors and validated by comparing the two values. The wind turbine being operated using a validated value as the operational parameter. The two sensors are selected among a trained machine learning model, a reference sensor and a computerized physical model.

First claim

The invention claimed is:

1. A method of operating a wind turbine in accordance with an operational parameter, the method comprising:
obtaining a first value of the operational parameter by a first source, wherein the first source is a trained machine learning model taking a set of operating parameters as input and outputting the first value;
obtaining a second value of the operational parameter by a second source, wherein the second source is a reference sensor, and the reference sensor provides one of the operating parameters of the set of operating parameters being input into the trained machine learning model;
comparing a difference between the first value and the second value with a comparison criterion to determine a validity of the first value; and
in response determining the first value is invalid, shutting down the wind turbine or changing a control strategy of the wind turbine.

2. The method of claim 1 , wherein the operational parameter is a wind speed, a wind direction, a blade pitch angle, a blade load, a temperature, a rotor azimuth angle, a rotor speed, a tower acceleration, an electrical power, a nacelle position, a light intensity, or a parameter value derived from a combination thereof.

3. The method of claim 1 , wherein the comparing the first value with the second value in accordance with the comparison criterion is based on comparing a statistical value of the first value and a statistical value of the second value which are obtained by statistically analysing a time series of the first and second values.

4. The method of claim 1 wherein the comparing the first value with the second value in accordance with the comparison criterion is based on comparing the first value and the second value to generate an indicator; and comparing the indicator with a validation threshold.

5. The method of claim 1 , wherein the reference sensor is at least one of a temperature sensor, a wind speed sensor, a wind direction sensor, a blade load sensor, a blade pitch angle sensor, a rotor azimuth angle sensor, a rotor speed sensor, a tower top acceleration sensor, a power sensor, a nacelle position sensor, a light intensity sensor, or a time of day sensor.

6. The method of claim 1 , wherein the machine learning model is trained using training data set from one or more of: a temperature sensor, a wind speed sensor, a wind direction sensor, a blade load sensor, a blade pitch angle sensor, a rotor azimuth angle sensor, a rotor speed sensor, a tower top acceleration sensor, a power sensor, a nacelle position sensor, a light intensity sensor and a time of day sensor.

7. A wind turbine system comprising:
a wind turbine, and
a control system configured to perform an operation, comprising:
obtaining a first value of an operational parameter by a first source, wherein the first source is a trained machine learning model taking a first set of operating parameters as input and outputting the first value;
obtaining a second value of the operational parameter by a second source, wherein the second source is a computerized physical model taking a second set of operating parameters as input and outputting the second value;
comparing a difference between the first value and the second value with a comparison criterion to determine a validity of the first value;
operating the wind turbine using the first value as the operational parameter if the first value is determined to be valid; and
shutting down the wind turbine or changing a control strategy of the wind turbine if the first value is determined to be invalid.

8. The wind turbine system of claim 7 , wherein the operational parameter is a wind speed, a wind direction, a blade pitch angle, a blade load, a temperature, a rotor azimuth angle, a rotor speed, a tower acceleration, an electrical power, a nacelle position, a light intensity, or a parameter value derived from a combination thereof.

9. The wind turbine system of claim 7 , wherein comparing the first value with the second value in accordance with the comparison criterion is based on comparing a statistical value of the first value and a statistical value of the second value, which are obtained by statistically analysing a time series of the first and second values.

10. The wind turbine system of claim 7 , wherein comparing the first value with the second value in accordance with the comparison criterion is based on:
comparing the first value and the second value to generate an indicator; and
comparing the indicator with a validation threshold.

11. The wind turbine system of claim 7 , wherein the machine learning model is trained using training data set from one or more of: a temperature sensor, a wind speed sensor, a wind direction sensor, a blade load sensor, a blade pitch angle sensor, a rotor azimuth angle sensor, a rotor speed sensor, a tower top acceleration sensor, a power sensor, a nacelle position sensor, a light intensity sensor and a time of day sensor.

12. A wind turbine system comprising:
a wind turbine, and
a control system configured to perform an operation, comprising:
obtaining a first value of an operational parameter by a first source, wherein the first source is a first trained machine learning model taking a first set of operating parameters as input and outputting the first value;
obtaining a second value of the operational parameter by a second source, wherein the second source is a second trained machine learning model taking a second set of operating parameters as input and outputting the second value;
comparing a difference between the first value and the second value with a comparison criterion to determine a validity of the first value;
operating the wind turbine using the first value as the operational parameter if the first value is determined to be valid; and
shutting down the wind turbine or changing a control strategy of the wind turbine if the first value is determined to be invalid.

13. The wind turbine system of claim 12 , wherein the first trained machine learning model is trained by a first training data set, and the second trained machine learning model is trained by feeding a second training data set, wherein the second training data set is different to the first training data set.

14. The wind turbine system of any of claim 13 wherein the first training data set is a first set of sensor inputs from a first set of sensors, and the second training data set is a second set of sensor inputs from a second set of sensors, wherein the second set of sensors is different to the first set of sensors.

15. The wind turbine system of claim 12 , wherein the first trained machine learning model is trained by a first machine learning method, and the second trained machine learning model is trained by a second machine learning method, wherein the second machine learning method is different to the first machine learning method.

16. The wind turbine system of claim 12 , wherein the operational parameter is a wind speed, a wind direction, a blade pitch angle, a blade load, a temperature, a rotor azimuth angle, a rotor speed, a tower acceleration, an electrical power, a nacelle position, a light intensity, or a parameter value derived from a combination thereof.

17. The wind turbine system of claim 12 , wherein comparing the first value with the second value in accordance with the comparison criterion is based on comparing a statistical value of the first value and a statistical value of the second value, which are obtained by statistically analysing a time series of the first and second values.

18. The wind turbine system of c

Assignees

• Vestas Wind Sys As

Inventors

• Duarte Pereira Goncalo Artur
• Pinto Frutuoso Ines
• Nielsen Johnny
• Van Schelve Jens

Classifications

• F03D7/046 Primary

  with learning or adaptive control, e.g. self-tuning, fuzzy logic or neural network · CPC title

• F03D7/045

  with model-based controls · CPC title

• F03D17/00

  Monitoring or testing of wind motors, e.g. diagnostics (testing during commissioning of wind motors F03D13/30) · CPC title

• F05B2260/84

  Modelling or simulation · CPC title

• F05B2270/404

  active, predictive, or anticipative · CPC title