Granted Patent B2

US 11,262,272 · App. 16/156,995 · Granted Mar 1, 2022

Inventor: Deokwoo Jung (Mountain View, CA)

Assignee: Palo Alto Research Center Incorporated

G01M5/0033

Abstract

Claims (55)

1. A method for estimating useful life of a load-bearing structure, the method comprising:

performing degradation measurement on the load-bearing structure to obtain degradation data for a plurality of time periods;

applying a constraint convex regression model to the degradation data to fit the degradation data to a convex function comprising one or more function parameters, which comprises:

for each time period, estimating the function parameters of the convex function based on degradation data associated with the time period;

calculating variance of each function parameter based on corresponding estimated function parameters for the plurality of time periods;

determining whether the calculated variance of a function parameter exceeds a predetermined threshold; and

in response to determining that the calculated variance of the function parameter exceeds the predetermined threshold, performing new degradation measurement to obtain new degradation data;

estimating a total useful life (TUL) of the load-bearing structure based on outputs of the constraint convex regression model; and

predicting a remaining useful life (RUL) based on the TUL and a current time.

2. The method of claim 1 , wherein the load-bearing structure comprises a load-bearing cable, and wherein performing the degradation measurement comprises one or more of:

measuring an electrical resistance;

measuring a thermal resistance; and

measuring a magnetic resistance.

3. The method of claim 1 , further comprising:

performing additional degradation measurement for a number of subsequent time periods; and

updating the estimated TUL based on the additional degradation measurement.

4. The method of claim 3 , further comprising:

using a particle-filtering technique to estimate a probability distribution of the estimated TUL based on the outputs of the constraint convex regression model and the additional degradation measurement.

5. The method of claim 4 , wherein the particle-filtering technique comprises Kalman filtering.

6. The method of claim 1 , wherein the convex function comprises an asymptotic function.

7. An apparatus for estimating useful life of a load-bearing structure, comprising:

one or more sensors embedded in the load-bearing structure, wherein the sensors are configured to obtain degradation data for a plurality of time periods;

a constraint convex regression modeling module configured to apply a constraint convex regression model to degradation data to fit the degradation data to a convex function comprising one or more function parameters, wherein applying the constraint convex regression model comprises:

for each time period, estimating the function parameters of the convex function based on degradation data associated with the time period;

calculating variance of each function parameter based on corresponding estimated function parameters for the plurality of time periods;

determining whether the calculated variance of a function parameter exceeds a predetermined threshold; and

in response to determining that the calculated variance of the function parameter exceeds the predetermined threshold, performing new degradation measurement to obtain new degradation data;

a total useful life (TUL) estimation module configured to estimate a TUL of the load-bearing structure based on outputs of the constraint convex regression model; and

a remaining useful life (RUL) prediction module configured to predict an RUL based on the TUL and a current time.

8. The apparatus of claim 7 , wherein the load-bearing structure comprises a load-bearing cable, and wherein the degradation data comprises one or more of:

an electrical resistance;

a thermal resistance; and

a magnetic resistance.

9. The apparatus of claim 7 , wherein the one or more sensors are further configured to obtain additional degradation data for a number of subsequent time periods, and wherein the TUL estimation module is further configured to update the estimated TUL based on the additional degradation data.

10. The apparatus of claim 9 , further comprising a particle-filtering module configured to use a particle-filtering technique to estimate a probability distribution of the TUL based on the outputs of the constraint convex regression model and the additional degradation data.

11. The apparatus of claim 10 , wherein the particle-filtering technique comprises Kalman filtering.

12. The apparatus of claim 7 , wherein the convex function comprises an asymptotic function.

13. A non-transitory computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for estimating useful life of a load-bearing structure, the method comprising:

performing degradation measurement on the load-bearing structure to obtain degradation data for a plurality of time periods;

applying a constraint convex regression model to the degradation data to fit the degradation data to a convex function comprising one or more function parameters, which comprises:

for each time period, estimating the function parameters of the convex function based on degradation data associated with the time period;

calculating variance of each function parameter based on corresponding estimated function parameters for the plurality of time periods;

determining whether the calculated variance of estimated function parameter exceeds a predetermined threshold; and

in response to determining that the calculated variance of the function parameter exceeds the predetermined threshold, performing new degradation measurement to obtain new degradation data;

estimating a total useful life (TUL) of the load-bearing structure based on outputs of the constraint convex regression model; and

predicting a remaining useful life (RUL) based on the TUL and a current time.

14. The non-transitory computer-readable storage medium of claim 13 , wherein the load-bearing structure comprises a load-bearing cable, and wherein performing the degradation measurement comprises one or more of:

measuring an electrical resistance;

measuring a thermal resistance; and

measuring a magnetic resistance.

15. The non-transitory computer-readable storage medium of claim 13 , wherein the method further comprises:

performing additional degradation measurement for a number of subsequent time periods; and

updating the TUL estimation based on the additional degradation measurement.

16. The non-transitory computer-readable storage medium of claim 15 , wherein the method further comprises using a particle-filtering technique to estimate a probability distribution of the TUL based on the outputs of the constraint convex regression model and the additional degradation measurement, and wherein the particle-filtering technique comprises Kalman filtering.

17. The non-transitory computer-readable storage medium of claim 13 , wherein the convex regression function comprises an asymptotic function.

Assignments (10)

Continuity (1)