Remaining life prediction for individual components from sparse data

• US 20070239407A1
• Filed: 01/12/2007
• Published: 10/11/2007
• Est. Priority Date: 01/12/2006
• Status: Active Grant

• Current Assignee: Jentek Sensors, Inc.
• Original Assignee: Jentek Sensors, Inc.
• Inventors: Zilberstein, Vladimir, Sheiretov, Yanko, Weiss, Volker, Goldfine, Neil
• Granted Patent: US 8,768,657 B2
• Time in Patent Office: Days
• US Class Current: 703/2
• CPC Class Codes: F05B 2270/109 to prolong engine life
• CPC Class Codes: F05B 2270/404 active, predictive, or anticipative
• CPC Class Codes: G01B 7/34 for measuring roughness or irregularity of surfaces
• CPC Class Codes: G01N 27/82 for investigating the presence of flaws
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Predicting the remaining life of individual aircraft, fleets of aircraft, aircraft components and subpopulations of these components. This is accomplished through the use of precomputed databases of response that are generated from a model for the nonlinear system behavior prior to the time that decisions need to be made concerning the disposition of the system. The database is calibrated with a few data points, to account for unmodeled system variables, and then used with an input variable to predict future system behavior. These methods also permit identification of the root causes for observed system behavior. The use of the response databases also permits rapid estimations of uncertainty estimates for the system behavior, such as remaining life estimates, particularly, when subsets of an input variable distribution are passed through the database and scaled appropriately to construct the output distribution. A specific example is the prediction of remaining life for an aircraft component where the model calculates damage evolution, input variables are a crack size and the number of cycles, and the predicted parameters are the actual stress on the component and the remaining life.

Claims

1. A method for estimating an output value distribution for a system comprising:

providing a database of system responses, the database relating a plurality of input variables to at least one output value;

providing the input variables to the system with an uncertainty distribution for at least one variable;

dividing the uncertainty distribution into portions;

determining a weighting value for each portion; and

calculating an output value distribution by passing the input variables and each weighted value of the uncertainty distribution through the database.

2. The method as claimed in claim 1 wherein the database is precomputed using a model for the system prior to its use in estimate the output value distribution.

3. The method as claimed in claim 1 wherein the input distribution is provided in terms of a quantitative distribution function.

4. The method as claimed in claim 3 wherein the input distribution is provided in terms of a probability distribution function.

5. The method as claimed in claim 1 wherein calculating an output value distribution involves passing a fraction of the input variables and input distribution into the database.

6. The method as claimed in claim 5 further comprising using a Latin hypercube methodology to determine the output value distribution, by passing a substantially reduced number of inputs through the database to construct the output distribution.

7. The method as claimed in claim 1 wherein the input variables are determined using an eddy current sensor.

8. The method as claimed in claim 1 wherein at least one of the input variables is crack size.

9. The method as claimed in claim 1 wherein at least one of the input variables is number of cycles.

10. The method as claimed in claim 1 wherein at least one of the system responses is remaining life.

11. A method for rapid decision making for a nonlinear system comprising:

generating a database of system responses from a model, with the responses having a nonlinear dependence on at least two input variables;

storing the database for future use;

calibrating the database with empirical information about the system;

obtaining an input value for an input variable;

converting the input value into a system response value using the database and a multivariate inverse method; and

using the system response value to make a decision.

12. The method as claimed in claim 11 wherein the empirical information is sensor based.

13. The method as claimed in claim 12 wherein the sensor is an eddy current sensor.

14. The method as claimed in claim 11 wherein the system response value is a prediction for future system behavior.

15. The method as claimed in claim 11 wherein the value for at least one input variable is unknown and a multivariate inverse method is used for estimation of the value of the unknown input variable.

16. The method as claimed in claim 15 wherein the unknown input is a residual stress in a metal.

17. The method as claimed in claim 15 wherein the unknown input is an applied load amplitude for a component.

18. The method as claimed in claim 11 further comprising:

using a distribution of values for an input variable with a multivariate inverse method to obtain an uncertainty for the system response value.

19. The method as claimed in claim 18 wherein the distribution is assumed.

20. The method as claimed in claim 18 wherein the distribution is measured.

21. The method as claimed in claim 11 wherein the system is a shot peened metal.

22. The method as claimed in claim 11 wherein the system response is an image of effective material properties and the input variables are quantitative features derived from this response.

23. The method as claimed in claim 11 further comprising:

tracking the system response value for multiple input values.

24. The method as claimed in claim 11 wherein the model is a damage evolution model.

25. The method as claimed in claim 11 wherein at least one of the input variables is crack size.

26. The method as claimed in claim 11 wherein at least one of the input variables is number of cycles.

27. The method as claimed in claim 11 wherein at least one of the system responses is remaining life.