System and method for turbine engine fault detection using discrete event system modeling

• US 7,369,932 B2
• Filed: 05/04/2006
• Issued: 05/06/2008
• Est. Priority Date: 05/04/2006
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Sturm, Doug, Kim, Kyusung, Mylaraswamy, Dinkar
• Primary Examiner(s): Tran; Khoi H.
• Assistant Examiner(s): Peche; Jorge O
• Application Number:
  US11/418,992
• Publication Number:
  US 20070260390A1
• Time in Patent Office: 733 Days
• Field of Search: 701/100, 701/29, 701/34, 701/99, 700/287, 700/290, 702/FOR.113, 702/182, 702/185, 731/16, 731/172, 105/36, 105/38, 105/61.5, 184/3.11, 415/902, 415/13, 477/30, 477/53
• US Class Current: 701/100
• CPC Class Codes: F01D 21/00 Shutting-down of machines or engines, e.g. in emergency; Regulating, controlling, or safety means not otherwise provided for
• CPC Class Codes: F01D 21/003 Arrangements for testing or measuring for measuring vibrations G01H
• CPC Class Codes: F01D 21/14 responsive to other specific conditions
• CPC Class Codes: F05D 2270/09 to cope with emergencies
• CPC Class Codes: F05D 2270/44 active, predictive, or anticipative
• CPC Class Codes: F05D 2270/708 with comparison tables
• CPC Class Codes: F05D 2270/709 with neural networks
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold

Abstract

A system and method is provided for fault detection in a turbine engine. The fault detection system and method uses discrete event system modeling to provide improved fault diagnosis and prognosis. The fault detection system and method receives sensor taken at multiple dynamic events occurring in different time windows. An event determination mechanism evaluates the received sensor data to determine if specified events have occurred. Indications of the occurrence of specified events are then passed to a discrete event system model. The discrete event system model analyzes the timing and sequencing of the event occurrences to determine if a fault has occurred. This method does not require a detailed modeling of the system, and thus can be applied to complex systems such as turbine engines.

Claims

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

an event determination mechanism, the event determination mechanism adapted to;

(i) receive sensor data from the turbine engine, wherein the sensor data comprises engine speed at peak exhaust gas temperature, exhaust gas temperature at peak exhaust gas temperature, and time to engine start;

(ii) analyze the sensor data to determine if specified events have occurred by comparing the engine speed at peak exhaust gas temperature to a first value to determine engine speed deviation from the first value, comparing the exhaust gas temperature at peak exhaust gas temperature to a second value to determine engine temperature deviation from the second value, and comparing the time to engine start to a third value to determine engine start time deviation from the third value; and

(iii) generate current event indicators responsive to the occurrence of the specified events by generating an engine speed event indicator responsive to the engine speed deviation exceeding a first threshold, generating an engine temperature event indicator responsive to the engine temperature deviation exceeding a second threshold, and generating a start time event indicator responsive to the engine start time deviation exceeding a third threshold; and

a discrete event system model, the discrete event system model adapted to (i) receive the current event indicators, (ii) receive past event indicators, and (iii) analyze timing and sequencing of the current event indicators and past event indicators to determine a likelihood of fault in the turbine engine.

2. The system of claim 1 wherein the discrete event system model further generates new event indicators in response to the timing and sequencing of the current event indicators and past event indicators, and wherein the new event indicators are fed back for use as the past event indicators in a next analysis cycle.

3. The system of claim 1 wherein the discrete event system model comprises a Petri net, and wherein the event indictors comprise markings.

4. The system of claim 3 wherein the Petri net includes a plurality of place nodes and transitions, wherein the Petri net updates tokens at the plurality of place nodes in response to the event indicators.

5. The system of claim 1 wherein the discrete event system model is adapted to analyze timing and sequencing of the event indicators to determine a likelihood of fault in the turbine engine by:

determining if the engine speed event indicator is followed by the engine temperature event indicator within a first specified time period; and

determining if the start time event indicator is within a second specified time period of the engine temperature event indicator.

6. The system of claim 1 wherein the sensor data comprises indications of an event in a component in the turbine engine.

7. A method of detecting faults in a turbine engine, the method comprising the steps of:

a) receiving sensor data from the turbine engine, wherein the sensor data comprises engine speed at peak exhaust gas temperature, exhaust gas temperature at peak exhaust gas temperature, and time to engine start;

b) analyzing the sensor data to determine if specified events have occurred by comparing the engine speed at peak exhaust gas temperature to a first value to determine engine speed deviation from the first value, comparing the exhaust gas temperature at peak exhaust gas temperature to a second value to determine engine temperature deviation from the second value, and comparing the time to engine start to a third value to determine engine start time deviation from the third value;

c) generating current event indicators responsive to the occurrence of the specified events by generating an engine speed event indicator responsive to the engine speed deviation exceeding a first threshold, generating an engine temperature event indicator responsive to the engine temperature deviation exceeding a second threshold, and generating a start time event indicator responsive to the engine start time deviation exceeding a third threshold;

d) receiving past event indicators from a previous analysis cycle of the turbine engine; and

e) analyzing timing and sequencing of the current event indicators and past event indicators to determine a likelihood of fault in the turbine engine.

8. The method of claim 7 further comprising the step of generating new event indicators in response to the timing and sequencing of the current event indicators and past event indicators, and wherein the new event indicators are used as the past event indicators in a next analysis cycle.

9. The method of claim 7 wherein the step of analyzing timing and sequencing of the current event indicators and past event indicators to determine a likelihood of fault in the turbine engine comprises analyzing using a Petri net.

10. The method of claim 9 wherein the Petri net includes a plurality of place nodes and transitions, wherein the Petri net updates tokens at the plurality of place nodes in response to the event indicators.

11. The method of claim 7 wherein the step of analyzing timing and sequencing of the current event indicators and past event indicators to determine a likelihood of fault in the turbine engine comprises:

determining if the engine speed event indicator is followed by the engine temperature event indicator within a first specified time period; and

determining if the start time event indicator is within a second specified time period of the engine temperature event indicator.

12. A program product comprising:

a) a fault detection program for detecting fault in a turbine engine, the program including;

an event determination program module, the event determination program module adapted to;

(i) receive sensor data from the turbine engine, wherein the sensor data comprises engine speed at peak exhaust gas temperature, exhaust gas temperature at peak exhaust gas temperature, and time to engine start;

(ii) analyze the sensor data to determine if specified events have occurred by comparing the engine speed at peak exhaust gas temperature to a first value to determine engine speed deviation from the first value, comparing the exhaust gas temperature at peak exhaust gas temperature to a second value to determine engine temperature deviation from the second value, and comparing the time to engine start to a third value to determine engine start time deviation from the third value; and

(iii) generate current event indicators responsive to the occurrence of the specified events by generating an engine speed event indicator responsive to the engine speed deviation exceeding a first threshold, generating an engine temperature event indicator responsive to the engine temperature deviation exceeding a second threshold, and generating a start time event indicator responsive to the engine start time deviation exceeding a third threshold; and

a discrete event system model program module, the discrete event system model program module adapted to (i) receive the current event indicators, (ii) receive past event indicators, and (iii) analyze timing and sequencing of the current event indicators and past event indicators to determine a likelihood of fault in the turbine engine; and

b) computer-readable signal bearing media bearing said program.

13. The program product of claim 12 wherein the discrete event system model further generates new event indicators in response to the timing and sequencing of the current event indicators and past event indicators, and wherein the new event indicators are fed back for use as the past event indicators in a next analysis cycle.

14. The program product of claim 12 wherein the discrete event system model comprises a Petri net, and wherein the event indictors comprise markings.

15. The program product of claim 14 wherein the Petri net includes a plurality of place nodes and transitions, wherein the Petri net updates tokens at the plurality of place nodes in response to the event indicators.