Method and system for detecting a failure or performance degradation in a dynamic system such as a flight vehicle

• US 6,643,569 B2
• Filed: 03/28/2002
• Issued: 11/04/2003
• Est. Priority Date: 03/30/2001
• Status: Expired due to Fees

• Current Assignee: Board of Regents of the University of Michigan (University of Michigan)
• Original Assignee: Board of Regents of the University of Michigan (University of Michigan)
• Inventors: Ribbens, William B., Miller, Robert H.
• Primary Examiner(s): Camby, Richard M.
• Application Number:
  US10/109,028
• Publication Number:
  US 20020193920A1
• Time in Patent Office: 586 Days
• Field of Search: 701/1, 701/29, 701/30, 701/31, 701/36, 701/32, 701/33, 701/34, 701/35
• US Class Current: 701/29
• CPC Class Codes: G05B 23/0245 based on a qualitative model, e.g. rule based; if-then decisions
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 50/40 Business processes related to the transportation industry shipping G06Q10/083

Abstract

A method and system for detecting a failure or performance degradation in a dynamic system having sensors for measuring state variables and providing corresponding output signals in response to one or more system input signals are provided. The method includes calculating estimated gains of a filter and selecting an appropriate linear model for processing the output signals based on the input signals. The step of calculating utilizes one or more models of the dynamic system to obtain estimated signals. The method further includes calculating output error residuals based on the output signals and the estimated signals. The method also includes detecting one or more hypothesized failures or performance degradations of a component or subsystem of the dynamic system based on the error residuals. The step of calculating the estimated values is performed optimally with respect to one or more of: noise, uncertainty of parameters of the models and un-modeled dynamics of the dynamic system which may be a flight vehicle or financial market or modeled financial system.

Claims

1. A method for detecting a failure or performance degradation in a dynamic system having sensors for measuring state variables of the system and providing corresponding output signals in response to at least one system input signal, the method comprising:

calculating estimated gains of a filter and selecting an appropriate linear model for processing the output signals based on the at least one system input signal, the step of calculating utilizing at least one model of the dynamic system to obtain estimated signals;

calculating output error residuals based on the output signals and the estimated signals; and

detecting at least one hypothesized failure or performance degradation of a component or subsystem of the dynamic system based on the error residuals, wherein the step of calculating the estimated gains is performed optimally with respect to one or more of;

noise, uncertainty of parameters of the at least one model and un-modeled dynamics of the dynamic system.

2. The method as claimed in claim 1 wherein the step of calculating estimated gains is performed continually.

3. The method as claimed in claim 1 wherein the dynamic system is a closed-loop dynamic system.

4. The method as claimed in claim 1 further comprising generating a signal for each hypothesized failure or performance degradation and storing each signal in a database for subsequent retrieval.

5. The method as claimed in claim 1 further comprising generating a signal for each hypothesized failure or performance degradation and processing each signal to diagnose the at least one hypothesized failure or performance degradation.

6. The method as claimed in claim 1 further comprising generating a signal for each hypothesized failure or performance degradation and processing each signal to obtain a reconfiguration signal.

7. The method as claimed in claim 6 wherein the dynamic system has a controller and wherein the method further comprises reconfiguring the controller based on the reconfiguration signal to compensate for the at least one hypothesized failure or performance degradation.

8. The method as claimed in claim 7 wherein the at least one hypothesized failure or performance degradation is a failure or degradation of one of the sensors.

9. The method as claimed in claim 8 wherein the reconfiguration signal inserts an estimated or compensated value of the output signal of the failed or degraded sensor into the controller.

10. The method as claimed in claim 6 wherein the dynamic system has a controller and at least one actuator and wherein the method further comprises reconfiguring the controller based on the reconfiguration signal to compensate for a change of the at least one actuator.

11. The method as claimed in claim 1 wherein the dynamic system is a flight vehicle and the sensors include flight control sensors.

12. The method as claimed in claim 1 wherein the dynamic system is a physical system characterized by a nonlinear dynamic model having parameters and wherein changes in the dynamic system are manifest by parameter changes in the nonlinear dynamic model.

13. The method as claimed in claim 1 wherein the error residuals are propagated in a unique direction in output detection space for a given hypothesized failure or performance degradation.

14. The method as claimed in claim 1 wherein the step of calculating estimated gains includes the step of controllably selecting parameters of the at least one model.

15. The method as claimed in claim 1 wherein the step of detecting detects intermittent faults.

16. The method as claimed in claim 13 wherein the step of detecting is based on magnitude and direction of the error residuals in the detection space.

17. The method as claimed in claim 1 wherein the at least one model includes non-dimensional variables and wherein the step of detecting includes the step of converting from the non-dimensional variables to dimensional variables to obtain re-scaled error residuals and wherein the step of detecting is also based on the re-scaled error residuals.

18. The method as claimed in claim 1 wherein a plurality of mathematical models are utilized to model the dynamic system and wherein the step of calculating estimated gains includes the step of selecting one of the plurality of mathematical models.

37. The method as claimed in claim 1 wherein the dynamic system is a financial market or modeled financial system.

19. A detection system for detecting a failure or performance degradation in a dynamic system having sensors for measuring state variables of the dynamic system and providing corresponding output signals in response to at least one system input signal, the detection system comprising:

means for calculating estimated gains of a filter and selecting an appropriate linear model for processing the output signals based on the at least one system input signal, the means for calculating utilizing at least one model of the dynamic system to obtain estimated signals;

means for calculating output error residuals based on the output signals and the estimated signals; and

means for detecting at least one hypothesized failure or performance degradation of a component or subsystem of the dynamic system based on the error residuals, wherein the means for calculating the estimated gains calculates optimally with respect to one or more of;

noise, uncertainty of parameters of the at least one model and un-modeled dynamics of the dynamic system.

20. The detection system as claimed in claim 19 wherein the estimated gains are calculated continually.

21. The detection system as claimed in claim 19 wherein the dynamic system is a closed-loop dynamic system.

22. The detection system as claimed in claim 19 further comprising means for generating a signal for each hypothesized failure or performance degradation and a database for storing each signal for subsequent retrieval.

23. The detection system as claimed in claim 19 further comprising means for generating a signal for each hypothesized failure or performance degradation and means for processing each signal to diagnose the at least one hypothesized failure or performance degradation.

24. The detection system as claimed in claim 19 further comprising means for generating a signal for each hypothesized failure or performance degradation and means for processing each signal to obtain a reconfiguration signal.

25. The detection system as claimed in claim 24 wherein the dynamic system has a controller and wherein the detection system further comprises means for reconfiguring the controller based on the reconfiguration signal to compensate for the at least one hypothesized failure or performance degradation.

26. The detection system as claimed in claim 25 wherein the at least one hypothesized failure or performance degradation is a failure or degradation of one of the sensors.

27. The detection system as claimed in claim 26 wherein the reconfiguration signal inserts an estimated or compensated value of the output signal of the failed or degraded sensor into the controller.

28. The detection system as claimed in claim 24 wherein the dynamic system has a controller and at least one actuator and wherein the detection system further comprises means for reconfiguring the controller based on the reconfiguration signal to compensate for a change of the at least one actuator.

29. The detection system as claimed in claim 19 wherein the dynamic system is a flight vehicle and the sensors include flight control sensors.

30. The detection system as claimed in claim 19 wherein the dynamic system is a physical system characterized by a nonlinear dynamic model having parameters and wherein changes in the dynamic system are manifest by parameter changes in the nonlinear dynamic model.

31. The detection system as claimed in claim 19 wherein the error residuals are propagated in a unique direction in output detection space for a given hypothesized failure or performance degradation.

32. The detection system as claimed in claim 19 wherein the means for calculating estimated gains includes means for controllably selecting parameters of the at least one model.

33. The detection system as claimed in claim 19 wherein the means for detecting detects intermittent faults.

34. The detection system as claimed in claim 31 wherein the means for detecting detects based on magnitude and direction of the error residuals in the detection space.

35. The detection system as claimed in claim 19 wherein the at least one model includes non-dimensional variables and wherein the means for detecting includes means for converting from the non-dimensional variables to dimensional variables to obtain re-scaled error residuals and wherein the means for detecting detects based on the re-scaled error residuals.

36. The detection system as claimed in claim 19 wherein a plurality of mathematical models are utilized to model the dynamic system and wherein the means for calculating estimated gains includes means for selecting one of the plurality of mathematical models.

38. The detection system as claimed in claim 19 wherein the dynamic system is a financial market or modeled financial system.