Transient fault detection system and method

• US 20040176901A1
• Filed: 03/03/2003
• Published: 09/09/2004
• Est. Priority Date: 03/03/2003
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Kim, Kyusung, Uluyol, Onder, Nwadiogbu, Emmanuel O., Menon, Sunil K.
• Granted Patent: US 7,043,348 B2
• Time in Patent Office: Days
• US Class Current: 701/100
• CPC Class Codes: F02C 9/00 Controlling gas-turbine plants; Controlling fuel supply in air- breathing jet-propulsion plants controlling air intakes F02C7/057; controlling turbines F01D; controlling compressors F04D27/00; controlling in general G05

Abstract

A transient fault detection system and method is provided that facilitates improved fault detection performance in transient conditions. The transient fault detection system provides the ability to detect symptoms of fault in engine that occur in transient conditions. The transient fault detection system includes a feature extractor that measures sensor data during transient conditions and extracts salient features from the measured sensor data. The extracted salient features are passed to a classifier that analyzes the extracted salient features to determine if a fault has occurred during the transient conditions. Detected faults can then be passed to a diagnostic system where they can be passed as appropriate to maintenance personnel.

Claims

1. A transient fault detection system for detecting transient faults in a turbine engine, the transient fault detection system comprising:

a) a feature extractor, the feature extractor receiving measured turbine sensor data from the turbine engine during a transient condition, the feature extractor extracting salient features from the measured turbine sensor data; and

b) a classifier, the classifier receiving the extracted salient features and analyzing the extracted salient features to determine if a fault occurred during the transient condition.

2. The system of claim 1 wherein the classifier comprises a neural network classifier.

3. The system of claim 2 wherein the neural network classifier comprises a multilayer perceptron neural network.

4. The system of claim 2 wherein the neural network classifier is trained using a leave-one out method.

5. The system of claim 2 wherein the neural network classifier is trained using sensor data including a plurality of known good transient condition data sets and a plurality of known bad transient condition data sets.

6. The system of claim 1 wherein the feature extractor performs a principal component analysis to extract the salient features from the measured turbine sensor data.

7. The system of claim 6 wherein the feature extractor extracts principal components selected to improve performance of the classifier.

8. The system of claim 1 wherein the sensor data comprises engine speed.

9. The system of claim 1 wherein the sensor data comprises exhaust gas temperature.

10. The system of claim 1 wherein the transient condition comprises turbine engine startup.

11. An apparatus comprising:

a) a processor;

b) a memory coupled to the processor;

c) a transient fault detection program residing in the memory and being executed by the processor, the transient fault detection program including;

i) a feature extractor, the feature extractor receiving measured turbine sensor data from the turbine engine during a transient condition, the feature extractor extracting salient features from the measured turbine sensor data; and

ii) a classifier, the classifier receiving the extracted salient features and analyzing the extracted salient features to determine if a fault occurred during the transient condition.

12. The apparatus of claim 11 wherein the classifier comprises a neural network classifier.

13. The apparatus of claim 12 wherein the neural network classifier comprises a multilayer perceptron neural network.

14. The apparatus of claim 12 wherein the neural network classifier is trained using a leave- one out method.

15. The apparatus of claim 12 wherein the neural network classifier is trained using sensor data including a plurality of known good transient condition data sets and a plurality of known bad transient condition data sets.

16. The apparatus of claim 11 wherein the feature extractor performs a principal component analysis to extract the salient features from the measured turbine sensor data.

17. The apparatus of claim 16 wherein the feature extractor extracts principal components selected to improve performance of the classifier.

18. The apparatus of claim 11 wherein the sensor data comprises engine speed.

19. The apparatus of claim 11 wherein the sensor data comprises exhaust gas temperature.

20. The apparatus of claim 11 wherein the transient condition comprises turbine engine startup.

21. A program product comprising:

a) a transient fault detection program, the transient fault detection program including;

i) a feature extractor, the feature extractor receiving measured turbine sensor data from the turbine engine during a transient condition, the feature extractor extracting salient features from the measured turbine sensor data; and

ii) a classifier, the classifier receiving the extracted salient features and analyzing the extracted salient features to determine if a fault occurred during the transient condition; and

b) signal bearing media bearing said program.

22. The program product of claim 21 wherein the signal bearing media comprises recordable media.

23. The program product of claim 21 wherein the signal bearing media comprises transmission media.

24. The program product of claim 21 wherein the classifier comprises a neural network classifier.

25. The program product of claim 24 wherein the neural network classifier comprises a multilayer perceptron neural network.

26. The program product of claim 24 wherein the neural network classifier is trained using a leave-one out method.

27. The program product of claim 24 wherein the neural network classifier is trained using sensor data including a plurality of known good transient condition data sets and a plurality of known bad transient condition data sets.

28. The program product of claim 21 wherein the feature extractor performs a principal component analysis to extract the salient features from the measured turbine sensor data.

29. The program product of claim 28 wherein the feature extractor extracts principal components selected to improve performance of the classifier.

30. The program product of claim 21 wherein the sensor data comprises engine speed.

31. The program product of claim 21 wherein the sensor data comprises exhaust gas temperature.

32. The program product of claim 21 wherein the transient condition comprises turbine engine startup.

33. A method of detecting faults in transient conditions in a turbine engine, the method comprising the steps of:

a) receiving measured turbine sensor data from the turbine engine during a transient condition;

b) extracting salient features from the measured turbine sensor data; and

c) classifying the extracted salient features to determine if a fault occurred during the transient condition.

34. The method of claim 33 wherein the step of classifying comprises classifying with a neural network classifier.

35. The method of claim 34 wherein the neural network classifier comprises a multilayer perceptron neural network.

36. The method of claim 34 further comprising the step of training the neural network classifier using a plurality of known good transient condition data sets and a plurality of known bad transient condition data sets.

37. The method of claim 36 wherein the step of training the neural network classifier comprises training using a leave-one out method.

38. The method of claim 33 wherein the step of extracting salient features comprises performing a principal component analysis.

39. The method of claim 38 where the step of extracting salient features comprises extracting principal components selected to improve performance of the classifying step.

40. The method of claim 33 wherein the sensor data comprises engine speed.

41. The method of claim 33 wherein the sensor data comprises exhaust gas temperature.

42. The method of claim 33 wherein the transient condition comprises turbine engine startup.