METHOD OF MULTI-LEVEL FAULT ISOLATION DESIGN

• US 20090248363A1
• Filed: 04/01/2008
• Published: 10/01/2009
• Est. Priority Date: 04/01/2008
• Status: Active Grant

• Current Assignee: Rtx Corporation
• Original Assignee: United Technologies Corporation (Raytheon Technologies Corporation d/b/a RTX)
• Inventors: Butler, Steven W.
• Granted Patent: US 8,271,233 B2
• Time in Patent Office: Days
• US Class Current: 702/185
• CPC Class Codes: F01D 21/003 Arrangements for testing or measuring for measuring vibrations G01H
• CPC Class Codes: F01D 21/02 Shutting-down responsive to overspeed
• CPC Class Codes: F01D 21/12 responsive to temperature
• CPC Class Codes: F01D 21/14 responsive to other specific conditions
• CPC Class Codes: F01D 21/20 Checking operation of shut-down devices
• CPC Class Codes: F02K 3/06 with front fan

Abstract

A fault isolation method and system includes tailored fault isolators for each grouping of similar data to differentiate between one or several potential faults within any group.

Claims

1. A method of isolating a fault condition from many potential fault conditions comprising:

generating a group of potential faults based on a plurality of measured parameters;

selecting a first isolator from a plurality of isolators based on performance of the first isolator in differentiating between potential faults within the group of potential faults;

redefining the group of potential faults to include fewer potential faults than the previous group of potential faults based on criteria applied by the first isolator; and

reporting a fault condition responsive to the redefined group of potential faults including one potential fault.

2. The method as recited in claim 1, including the step of selecting a second isolator different than the first isolator from the plurality of isolators responsive to the redefined group including more than one potential fault, wherein the second isolator is selected based on performance of the second isolator in differentiating between potential faults in the redefined group of potential faults.

3. The method as recited in claim 1, wherein the group of potential faults is generated by clustering individual potential faults together responsive to similarities in fault behavior within a defined range.

4. The method as recited in claim 3, wherein similarities in fault behavior are quantified by measuring differences between fault classes within a confusion matrix.

5. The method as recited in claim 3, wherein the group of potential faults is accepted responsive to a difference discernable by the first isolator between at least two potential faults being within a desired acceptance criteria.

6. The method as recited in claim 5, wherein the acceptance criteria comprises a desired frequency of obtaining a correctly identified fault condition.

7. The method as recited in claim 5, wherein the acceptance criteria comprises a desired frequency of obtaining a correctly consolidated group.

8. The method as recited in claim 1, including the step of performing a maintenance operation on a mechanical system corresponding with the reported fault condition.

9. A fault isolator for determining a fault condition of a system comprising:

a receiver for receiving parameter data indicative of system performance;

a controller receiving data from the receiver, the controller including, a first isolator for differentiating between at least two potential faults from a group of potential faults indicated by the received data;

a second fault isolator different than the first isolator for differentiating between potential faults not discernable by the first isolator; and

a communication device for communicating the fault condition determined by criteria of the first and second isolators.

10. The fault isolator as recited in claim 9, wherein the first isolator comprises a model of the system utilized for discerning between at least two of the potential faults from the group of potential faults.

11. The fault isolator as recite in claim 10, wherein the second isolator comprises a model of the system different than the model of the system utilized for the first isolator.

12. The fault isolator as recited in claim 9, wherein each of the first fault isolator and the second fault isolator are selected based on quantified performance of determining fault conditions based on a sample set of data indicative of system performance indicative of known fault conditions.