Method for performing gas turbine performance diagnostics

• US 20040088100A1
• Filed: 10/31/2002
• Published: 05/06/2004
• Est. Priority Date: 10/31/2002
• Status: Active Grant

• Current Assignee: Raytheon Technologies Corporation d/b/a RTX (Rtx Corporation)
• Original Assignee: United Technologies Corporation (Rtx Corporation)
• Inventors: Volponi, Allan J.
• Granted Patent: US 6,909,960 B2
• Time in Patent Office: Days
• US Class Current: 701/100
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

A method of performing diagnostics on a system comprises receiving a plurality of measurement parameters, each corresponding to one of a plurality of parameters at a time k, forming a deviation vector from the plurality of measurement parameters, calculating an initial deviation vector from an initial fault vector, calculating a multiple fault isolation deviation vector using the initial deviation vector and the deviation vector, determining if an event is in progress using the multiple fault isolation deviation vector, performing statistical data validity to set a present inhibit flag and a past inhibit flag, and performing module performance analysis according to the present inhibit flag and the past inhibit flag.

Claims

1. A method of performing diagnostics on a system comprising:

receiving a plurality of measurement parameters, each corresponding to one of a plurality of parameters at a time k;

forming a deviation vector from said plurality of measurement parameters;

calculating an initial deviation vector from an initial fault vector;

calculating a multiple fault isolation deviation vector using said initial deviation vector and said deviation vector;

determining if an event is in progress using said multiple fault isolation deviation vector;

performing statistical data validity to set a present inhibit flag and a past inhibit flag; and

performing module performance analysis according to said present inhibit flag and said past inhibit flag.

2. The method of claim 1, wherein said performing module performance comprises the additional steps of:

receiving said multiple fault isolation deviation vector;

performing trend detection using said multiple fault isolation deviation vector;

performing isolation estimation;

constructing a total fault vector.

3. The method of claim 2, wherein said isolation estimation comprises the step of performing event isolation estimation.

4. The method of claim 2, wherein said isolation estimation comprises the step of performing multiple fault isolation estimation.

5. The method of claim 1, wherein said receiving said plurality of measurement parameter comprises receiving a percent deviation.

6. The method of claim 5, comprising the additional step of processing said plurality of measurement parameters to correct for a plurality of conditions.

7. The method of claim 6, wherein said processing said plurality of measurement parameters to correct for a plurality of conditions comprises processing said plurality of measurement parameters to correct for a plurality of conditions selected from the group consisting of mach number, altitude, and power level.

8. The method of claim 1, wherein said receiving said plurality of measurement parameters comprises receiving a plurality of engine parameters.

9. The method of claim 8, wherein said receiving said plurality of engine parameters comprises receiving a plurality of flow parameters.

10. A method of performing diagnostics on a system comprising:

receiving a plurality of measurement parameters each corresponding to one of a plurality of parameters at a time k;

forming a deviation vector from said plurality of measurement parameters;

calculating an initial deviation vector from an initial fault vector;

calculating a multiple fault isolation deviation vector using said initial deviation vector and said deviation vector;

determining if an event is in progress using said multiple fault isolation deviation vector; and

performing module performance analysis.

11. The method of claim 10, wherein said performing module performance comprises the additional steps of:

receiving said multiple fault isolation deviation vector;

performing trend detection using said multiple fault isolation deviation vector;

performing isolation estimation;

constructing a total fault vector.

12. The method of claim 11, wherein said isolation estimation comprises the steps of performing event isolation estimation.

13. The method of claim 11, wherein said isolation estimation comprises the step of performing multiple fault isolation estimation.

14. The method of claim 10, wherein said receiving said plurality of measurement parameter comprises receiving a percent deviation.

15. The method of claim 14, comprising the additional step of processing said plurality of measurement parameters to correct for a plurality of conditions.

16. The method of claim 15, wherein said processing said plurality of measurement parameters to correct for a plurality of conditions comprises processing said plurality of measurement parameters to correct for a plurality of conditions selected from the group consisting of mach number, altitude, and power level.

17. The method of claim 10, wherein said receiving said plurality of measurement parameters comprises receiving a plurality of engine parameters.

18. The method of claim 17, wherein said receiving said plurality of engine parameters comprises receiving a plurality of flow parameters.