AUTOMATED ENGINE DATA DIAGNOSTIC ANALYSIS

• US 20080228338A1
• Filed: 03/15/2007
• Published: 09/18/2008
• Est. Priority Date: 03/15/2007
• Status: Abandoned Application

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Stramiello, Andrew D., Howard, Joseph S.
• Application Number:
  US11/686,777
• Publication Number:
  US 20080228338A1
• Time in Patent Office: Days
• US Class Current: 701/29
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G05B 23/0278 Qualitative, e.g. if-then rules; Fuzzy logic; Lookup tables; Symptomatic search; FMEA

Abstract

A method for diagnosing potential faults reflected in operational data for a turbine engine includes the steps of generating a diagnostic pattern for the operational data and comparing the diagnostic pattern with a plurality of historical patterns, to thereby identify one or more likely potential faults reflected in the operational data. The diagnostic pattern comprises a plurality of scalars. Each scalar represents an arithmetic relationship between values of the operational data and values predicted by a baseline thermodynamic model. Each historical pattern is linked to one or more specific faults.

Claims

3-1. The method of claim 1, wherein at least one of the scalars represents an additive relationship between values of the operational data and values predicted by the baseline thermodynamic model.

11. A program product for diagnosing potential faults reflected in operational data for a turbine engine, the program product comprising:

(a) a program configured to;

generate a diagnostic pattern for the operational data, the diagnostic pattern comprising a plurality of scalars, each scalar representing an arithmetic relationship between values of the operational data and values predicted by a baseline thermodynamic model; and

compare the diagnostic pattern with a plurality of historical patterns, each historical pattern linked to one or more specific faults, to thereby identify one or more likely potential faults reflected in the operational data; and

(b) a computer-readable signal bearing media bearing the program.

12. The program product of claim 11, wherein the program is further configured to:

generate a matrix of operating parameter perturbations to simulate a plurality of engine faults; and

run the matrix through the baseline thermodynamic model, to thereby generate a historical pattern for each fault, each historical pattern representing a deviation from the baseline thermodynamic model resulting from the fault.

13. The program product of claim 11, wherein at least one of the scalars represents a multiplicative relationship between values of the operational data and values predicted by the baseline thermodynamic model.

14. The program product of claim 11, wherein at least one of the scalars represents an additive relationship between values of the operational data and values predicted by the baseline thermodynamic model.

15. The program product of claim 11, wherein at least one of the scalars represents a relationship between:

a first operational value from the operational data; and

an expected value of the first operational value, determined at least in part based on a second operational value from the operational data and a known relationship, based at least in part on one or more laws of physics, between the first and second operational values.

16. The program product of claim 11, wherein:

each historical pattern includes a plurality of historical scalars, each historical scalar representing a deviation from the baseline thermodynamic model; and

the program is further configured to normalize the scalars and the historical scalars.

17. The program product of claim 11, wherein the program is further configured to:

quantify an expected severity of the one or more potential faults, based at least in part on the comparison between the diagnostic pattern and the plurality of historical patterns.

18. The program product of claim 11, wherein the program is further configured to:

identify multiple likely potential faults based at least in part on the comparison of the diagnostic pattern with the plurality of historical patterns, each likely potential fault having a different historical pattern; and

assign probability values to each of the identified likely potential faults based at least in part on the comparison between the diagnostic pattern and the plurality of historical patterns, each probability value representing a probability that the engine has a particular fault.

19. The program product of claim 18, wherein the program is further configured to generate user instructions for further diagnosis of the multiple likely potential faults, based at least in part on the assigned probability values.

20. A program product for diagnosing potential faults reflected in operational data for a turbine engine, the program product comprising:

(a) a program configured to;

generate a matrix of operating parameter perturbations to simulate a plurality of engine faults;

run the matrix through the baseline thermodynamic model, to thereby generate a historical pattern for each fault, each historical pattern representing a deviation from the baseline thermodynamic model resulting from the fault;

generate a diagnostic pattern for the operational data, the diagnostic pattern comprising a plurality of scalars, each scalar representing an arithmetic relationship between values of the operational data and values predicted by the baseline thermodynamic model;

compare the diagnostic pattern with a plurality of historical diagnostic patterns, each historical pattern linked to one or more specific faults, to thereby identify multiple likely potential faults based at least in part on the comparison of the diagnostic pattern with the plurality of historical patterns, each likely potential fault having a different historical pattern;

assign probability values to each of the identified likely potential faults based at least in part on the comparison between the diagnostic pattern and the plurality of historical patterns, each probability value representing a probability that the engine has a particular fault; and

generate user instructions for further diagnosis of the multiple likely potential faults, based at least in part on the assigned probability values; and

(b) a computer-readable signal bearing media bearing the program.