[dtobonme]

Hi,

How long must be the prognostic horizon to estimate the RUL? And can you explain how to use the “leave-one-out cross-validation” to evaluate the performance of the method ? because i don’t see how to use this technique to evaluate multistep-ahead predictions necessary to estimate the future wear or failure state.

Thank you,

Hi,

How long must be the prognostic horizon to estimate the RUL? And can you explain how to use the “leave-one-out cross-validation” to evaluate the performance of the method ? because i don’t see how to use this technique to evaluate multistep-ahead predictions necessary to estimate the future wear or failure state.

Thank you,

[eklund]

The RUL estimate is implicit. You generate an estimated maximum wear for any flute, and when it crosses a fixed (and unknown to you) threshold, the tool is finished.

Leave one out (LOO) is a method of cross validation:
http://en.wikipedia.org/wiki/Cross-validation_%28statistics%29

There are at least two ways to partition the data. You might train on two runs and evaluate on the third. Alternatively, you might train on two flutes within a run and evaluate on the third flute within a run.

But you are not obliged to use cross validation! If you have an approach you think will work better, use that…

[vsotiris]

Regarding RUL,

Can you clarify what you mean by: “The RUL estimate is implicit”.

I am under the impression that RUL is the number of cuts remaining until failure, given a fixed failure threshold, and given survival up until a certain earlier cut.

For example, given observed data from a test tool, up until say the 100th cut, then ask the question: what is the remaining useful life? Or alternatively, how many more cuts until failure?

Thanks

[Author]

Posts