METHOD FOR COMPUTING THE RELATIVE LIKELIHOOD OF FAILURES

• US 20100198771A1
• Filed: 02/05/2009
• Published: 08/05/2010
• Est. Priority Date: 02/05/2009
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Barton, Bradley John, Shao, Qingqiu Ginger, Kolbet, David Michael, Magnuson, Randy, Khalak, Asif, Dins, C. Arthur
• Granted Patent: US 8,224,765 B2
• Time in Patent Office: Days
• US Class Current: 706/52
• 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 for determining relative likelihood of a failure mode is provided. The method comprises receiving evidence observations of a monitored system from monitors connected in a many-to-many relationship to the failure modes, generating a fault condition including states of all failure modes that are connected to the monitors, and computing a relative probability of failure for each failure mode. The fault condition is generated for a reference model of the monitored system and is based on the received evidence observations. The relative probability of failure for each failure mode is based on a false alarm probability, a detection probability, and a ratio of prior probabilities of a candidate hypothesis to a null hypothesis of no active failure mode.

Claims

1. A method for determining relative likelihood of a failure mode, the method comprising:

receiving evidence observations of a monitored system from monitors connected in a many-to-many relationship to the failure modes;

generating a fault condition including states of all failure modes that are connected to the monitors, the fault condition being generated for a reference model of the monitored system based on the received evidence observations; and

computing a relative probability of failure for each failure mode based on a false alarm probability, a detection probability, and a ratio of prior probabilities of a candidate hypothesis to a null hypothesis of no active failure modes.

2. The method of claim 1, wherein computing a relative probability of failure for each failure mode comprises:

generating the ratio of the prior probabilities of the candidate hypothesis relative to the null hypothesis;

generating a trigger factor associated with triggered monitors in the fault condition;

generating a quiescent factor associated with quiescent monitors in the fault condition; and

determining the relative probability of failure based on the ratio of the prior probabilities of the candidate hypothesis to the null hypothesis, the trigger factor, and the quiescent factor.

3. The method of claim 2, wherein generating the trigger factor associated with the triggered monitors in the fault condition comprises:

generating a probability that the triggered monitors originated from the hypothesized fault condition normalized by a probability that the triggered monitors resulted from the null hypothesis.

4. The method of claim 2, wherein generating the quiescent factor associated with the quiescent monitors in the fault condition comprises:

generating a probability that the quiescent monitors originated from the candidate hypothesis normalized by a probability that the quiescent monitors resulted from the null hypothesis.

5. The method of claim 1, further comprising:

receiving additional evidence observations from the monitors;

generating an updated fault condition based on the received additional evidence observations; and

re-computing a relative probability of failure for each failure mode in the updated fault condition.

6. The method of claim 5, wherein re-computing a relative probability of failure for each failure mode comprises:

regenerating the ratio of the prior probabilities of each candidate in the updated fault condition to the null hypothesis;

regenerating a trigger factor associated with triggered monitors in the updated fault condition;

regenerating a quiescent factor associated with quiescent monitors in the updated fault condition; and

determining the relative probability of failure based on the ratio of the regenerated prior probabilities of the candidate hypothesis to the null hypothesis, the regenerated trigger factor, and the regenerated quiescent factor.

7. The method of claim 6, wherein regenerating the trigger factor associated with the triggered monitors in the updated fault condition comprises:

regenerating a probability that the triggered monitors originated from the hypothesized fault condition normalized by a probability that the triggered monitors resulted from the null hypothesis; and

wherein regenerating the quiescent factor associated with the quiescent monitors in the updated fault condition comprises;

regenerating a probability that the quiescent monitors originated from the candidate hypothesis normalized by a probability that the quiescent monitors resulted from the null hypothesis.

8. The method of claim 1, further comprising:

generating the reference model of the monitored system; and

implementing model-based software to generate the fault condition for the states-of-failure modes for the reference model of the monitored system.

9. A computer readable medium storing instructions to implement a method to:

generate a fault condition including states of all failure modes that are connected to the monitors, the fault condition being generated for a reference model of a monitored system based on evidence observations received from monitors; and

compute a relative probability of failure for each failure mode based on a false alarm probability, a detection probability, and a ratio of prior probabilities of a candidate hypothesis to a null hypothesis of no active failure modes.

10. The computer readable medium of claim 9, wherein the instructions further include instructions to implement a method to:

generate an updated fault condition based on additional evidence observations received from the monitors; and

re-compute a relative probability of failure for each failure mode in the updated fault condition.

11. The computer readable medium of claim 9, wherein the instructions to compute a relative probability of failure for each failure mode comprise instructions to implement a method to:

generate the ratio of the prior probabilities of the candidate hypothesis relative to the null hypothesis;

generate a trigger factor associated with triggered monitors in the fault condition; and

generate a quiescent factor associated with quiescent monitors in the fault condition.

12. A health management system configured to determine relative probabilities of failure in a monitored system, the health management system comprising:

sensors to monitor system states and to output evidence observations; and

at least one processor configured to;

generate a fault condition including states of all failure modes that are connected to the sensors based on the evidence observations received from the sensors for a reference model of the monitored system; and

execute algorithms configured to compute a relative probability of failure for each failure mode based on a false alarm probability, a detection probability, and a ratio of prior probabilities of a candidate hypothesis to a null hypothesis of no active failure modes.

13. The health management system of claim 12, wherein the monitored system includes system components, wherein at least one system component is communicatively coupled to at least one other system component, and wherein the sensors are configured to monitor the system components and the linking between two system components.

14. The health management system of claim 12, further comprising:

storage medium to store software executable by the at least one processor.

15. The health management system of claim 14, further comprising:

a failure-mode-relative-likelihood algorithm stored in the storage medium.

16. The health management system of claim 14, further comprising:

a reference model of the monitored system stored in the storage medium.