System and method for predicting wind turbine component failures

• US 8,326,577 B2
• Filed: 09/20/2011
• Issued: 12/04/2012
• Est. Priority Date: 09/20/2011
• Status: Active Grant

• Current Assignee: GE Infrastructure Technology, LLC (General Electric Company)
• Original Assignee: General Electric Company
• Inventors: Graham, Fred Gorum III, Nanda, Subrat, Vittal, Sameer, Talukdar, Atanu
• Primary Examiner(s): Bui, Bryan
• Application Number:
  US13/237,073
• Publication Number:
  US 20120143565A1
• Time in Patent Office: 441 Days
• Field of Search: 702/181, 702/34, 416/41, 416/44, 416/61
• US Class Current: 702/181
• CPC Class Codes: F03D 17/00 Monitoring or testing of wind motors, e.g. diagnostics testing during commissioning of wind motors F03D13/30
• CPC Class Codes: F05B 2260/821 Parameter estimation or prediction
• CPC Class Codes: F05B 2270/332 Maximum loads or fatigue criteria
• CPC Class Codes: G05B 23/0237 based on parallel systems, e.g. comparing signals produced at the same time by same type systems and detect faulty ones by noticing differences among their responses

Abstract

Methods and systems for use in predicting wind turbine failures are provided. One example method includes determining a parametric profile for a component of a wind turbine from operating data of a plurality of wind turbines, determining an anomaly profile for the component of the wind turbine from anomaly alerts from a plurality of wind turbines, and determining a probability of failure for the component of the wind turbine based on the parametric profile and the anomaly profile. The parametric profile defines at least one parametric event associated with the component prior to failure of the component, and the anomaly profile defines at least one anomaly associated with the component prior to failure of the component.

Claims

1. A computer-based method for use in predicting a wind turbine component failure, said computer-based method comprising:

determining, at a computing device, a parametric profile for a component of a wind turbine from operating data of a plurality of wind turbines, the parametric profile defining at least one parametric event associated with the component prior to failure of the component;

determining, at the computing device, an anomaly profile for the component of the wind turbine from anomaly alerts from a plurality of wind turbines, the anomaly profile defining at least one anomaly associated with the component prior to failure of the component; and

,

determining a probability of failure for the component of the wind turbine based on the parametric profile and the anomaly profile

wherein the anomaly alerts are generated when a component parameter falls outside of an expected range.

2. The computer-based method of claim 1, wherein determining the parametric profile includes analyzing at least one parameter of the component relative to one of an expected value and a set point.

3. The computer-based method of claim 2, wherein analyzing at least one parameter includes analyzing the at least one parameter based on a turbine status.

4. The computer-based method of claim 2, wherein the at least one parameter includes at least one of voltage, current, power, wind speed, torque, and temperature associated with the wind turbine.

5. The computer-based method of claim 1, wherein determining the parametric profile includes analyzing at least one parameter of the component relative to a second parameter associated with the wind turbine.

6. The computer-based method of claim 1, wherein determining the probability of failure includes determining the probability of failure based on domain knowledge.

7. The computer-based method of claim 6, wherein the domain knowledge includes at least one of a maintenance record for at least one wind turbine and theoretical operation of a wind turbine.

8. The computer-based method of claim 1, further comprising assigning the probability of failure to a wind turbine based on an occurrence of the at least one parametric event and the at least one anomaly.

9. The computer-based method of claim 1, wherein determining the anomaly profile includes identifying the anomaly alerts associated with the component.

10. The computer-based method of claim 9, wherein determining the anomaly profile includes validating the accuracy of the anomaly profile based on the anomaly alerts and selecting the anomaly profile that exceeds a threshold accuracy.

11. The computer-based method of claim 1, wherein the component comprises an electronic component of the wind turbine.

12. The computer-based method of claim 11, wherein determining the anomaly profile includes validating the anomaly profile against the analogy alerts to determine a validation score, and wherein determining the probability of failure includes weighting the anomaly profile based on at least the validation score.

13. A data analyzer for use in predicting a wind turbine component failure, said data analyzer comprising:

a database including operating data for a plurality of wind turbines and anomaly alerts for a plurality of wind turbines; and

,

a processing device coupled to said database;

the processing device configured to;

determine a parametric profile for a component of a wind turbine from the operating data, the parametric profile defining at least one parametric event associated with the component prior to failure of the component;

determine an anomaly profile for the component of the wind turbine from the anomaly alerts, the anomaly profile defining at least one anomaly associated with the component prior to failure of the component; and

,

determine a probability of failure for the component of the wind turbine based on the parametric profile and the anomaly profile

wherein the anomaly alerts are generated when a component parameter falls outside of an expected range.

14. The data analyzer of claim 13, wherein the processing device is configured to identify anomaly alerts associated with the component, validate the accuracy of the anomaly profile based on the anomaly alerts, and select the anomaly profile that exceed a threshold accuracy in order to determine the anomaly profile.

15. The data analyzer of claim 13, further comprising a communication interface for coupling to at least one of a network and a system monitor to receive said operating data and said anomaly alerts.

16. The data analyzer of claim 15, wherein the processing device is configured to assign the probability of failure to a wind turbine based on an occurrence of the at least one parametric event and the at least one anomaly.

17. The data analyzer of claim 15, wherein the at least one parametric event comprises a correlation between a wind speed and a torque output of a wind turbine.

18. One or more non-transitory computer-readable storage media having computer executable instructions embodied thereon, wherein when executed by at least one processing device, said computer executable instructions cause the at least one processing device to:

determine a parametric profile for a component of a wind turbine from operating data of a plurality of wind turbines, the parametric profile defining at least one parametric event associated with the component prior to failure of the component;

determine an anomaly profile for the component of the wind turbine from anomaly alerts generated from a plurality of wind turbines, the anomaly profile defining at least one anomaly associated with the component prior to failure of the component; and

,

determine a probability of failure for the component of the wind turbine based on the parametric profile and the anomaly profile

wherein the anomaly alerts are generated when a component parameter falls outside of an expected range.

19. The one or more non-transitory computer-readable storage media of claim 18, wherein said computer executable instructions cause the at least one processing device to analyze at least one parameter of the component relative to a set point to determine the parametric profile.

20. The one or more non-transitory computer-readable storage media of claim 18, wherein said computer executable instructions cause the at least one processing device to assign the probability of failure to a wind turbine based on an occurrence of the at least one parametric event and the at least one anomaly.