SYSTEM AND METHOD FOR EQUIPMENT REMAINING LIFE ESTIMATION

• US 20080208487A1
• Filed: 02/23/2007
• Published: 08/28/2008
• Est. Priority Date: 02/23/2007
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Bonissone, Piero Patrone, Yan, Weizhong, Goebel, Kai Frank, Eklund, Neil Holger White, Xue, Feng
• Granted Patent: US 7,548,830 B2
• Time in Patent Office: Days
• US Class Current: 702/34
• CPC Class Codes: G05B 23/0283 Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06
• CPC Class Codes: G06N 3/043 based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
• CPC Class Codes: G06N 3/126 Evolutionary algorithms, e.g. genetic algorithms or genetic programming
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• 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

Abstract

A method to reduce uncertainty bounds of predicting a remaining life of a probe using a set of diverse models is disclosed. The method includes generating an estimated remaining life output by each model of the set of diverse models, aggregating each of the respective estimated remaining life outputs via a fusion model, and in response to the aggregating, predicting the remaining life, the predicting having reduced uncertainty bounds based on the aggregating. The method further includes generating a signal corresponding to the predicted remaining life of the probe.

Claims

1. A method to reduce uncertainty bounds of predicting a remaining life of a probe using a set of diverse models, the method comprising:

generating an estimated remaining life output by each model of the set of diverse models;

aggregating each of the respective estimated remaining life outputs via a fusion model;

in response to the aggregating, predicting the remaining life, the predicting having reduced uncertainty bounds based on the aggregating; and

generating a signal corresponding to the predicted remaining life of the probe.

2. The method of claim 1, wherein the generating an estimated remaining life output comprises:

making available to each model in the set of diverse models data relating to a behavior of the probe so that each model generates the estimated remaining life output.

3. The method of claim 1, wherein the aggregating comprises:

making available to the fusion model data relating to a behavior of the probe; and

making available to the fusion model state information corresponding to each model of the set of diverse models.

4. The method of claim 1, wherein the predicting comprises:

predicting the remaining life of a turbine engine.

5. The method of claim 1, further comprising:

training each model in the set of diverse models using bootstrap data validation.

6. The method of claim 1, further comprising:

defining at least one of a parameter and a structure of the fusion model.

7. The method of claim 6, wherein the defining comprises:

pre-computing at least one of the parameter and the structure of the fusion model prior to the generating the estimated remaining life output.

8. The method of claim 6, further comprising

tuning and maintaining at least one of the parameter and the structure via an optimization wrapper.

9. The method of claim 8, wherein the tuning and maintaining comprises:

receiving historical data regarding at least one of the parameter and the structure;

encoding at least one of the parameter and the structure for an evolutionary algorithm;

creating a performance metric for the estimating;

tuning at least one of the parameter and the structure using the optimization wrapper to optimize the performance metric; and

storing the tuned at least one of the parameter and the structure for subsequent receiving.

10. The method of claim 6, wherein the defining comprises:

defining the parameter of the fusion model, the parameter comprising a weight corresponding to each respective model of the set of diverse models.

11. The method of claim 10, wherein the defining the parameter comprises:

making available input conditions describing more than one region of a feature space; and

defining a local weight corresponding to each of the more than one regions of the feature space.

12. The method of claim 11, further comprising:

describing a region of the more than one regions via at least one of a decision tree, a grid, and a fuzzy partition.

13. The method of claim 6, wherein the defining the structure of the fusion model comprises:

defining at least one of an intersection fusion operator, a compensatory fusion operator, and a union fusion operator.

14. The method of claim 6, wherein the defining the structure of the fusion model comprises:

defining a temporal operator comprising at least one of a forgetting factor and a moving window.

15. A program storage device readable by a computer, the device embodying a program or instructions executable by the computer to perform the method of claim 1.

16. A system to estimate a remaining life of a probe using a set of diverse models, the system comprising:

a processor;

a computational model application for executing on the processor, the computational model application performing a method, comprising;

generating an estimated remaining life output by each model of the set of diverse models;

aggregating each of the respective estimated remaining life outputs via a fusion model; and

in response to the aggregating, predicting the remaining life, the predicting having reduced uncertainty bounds based on the aggregating;

wherein the processor is responsive to the computational model application to generate a signal corresponding to the predicted remaining life of the probe.

17. The system of claim 16, wherein the generating by the computational model application comprises:

making available to each model in the set of diverse models data relating to a behavior of the probe so that each model generates the estimated remaining life output.

18. The system of claim 16, wherein the aggregating by the computational model application comprises:

making available to the fusion model data relating to a behavior of the probe; and

making available to the fusion model state information corresponding to each model of the set of diverse models.

19. The system of claim 16, wherein:

the probe is a turbine engine.

20. The system of claim 16, wherein the computational model application further performs:

training each model in the set of diverse models using bootstrap data validation.

21. The system of claim 16, wherein the computational model application further performs:

defining at least one of a parameter and a structure of the fusion model.

22. The system of claim 21, wherein the defining comprises:

pre-computing at least one of the parameter and the structure of the fusion model prior to the generating the estimated remaining life output.

23. The system of claim 21, wherein the computational model application further performs:

tuning and maintaining at least one of the parameter and the structure via an optimization wrapper.

24. The system of claim 21, wherein the defining by the computational model application comprises:

defining the parameter of the fusion model, the parameter comprising a weight corresponding to each respective model of the set of diverse models.

25. The system of claim 24, wherein:

the defining the parameter by the computational model application comprises;

making available input conditions describing more than one region of a feature space; and

defining a local weight corresponding to each of the more than one regions of the feature space.

26. The system of claim 25, wherein the computational model application further performs:

describing a region of the more than one regions via at least one of a decision tree, a grid, and a fuzzy partition.

27. The system of claim 21, wherein the defining the structure of the fusion model by the computational model application further comprises:

defining at least one of an intersection fusion operator, a compensatory fusion operator, and a union fusion operator.

28. The system of claim 21, wherein the defining the structure of the fusion model by the computational model application further comprises:

defining a temporal operator comprising at least one of a forgetting factor and a moving window