Hi,
I have a customer (DGA, Directorate General of Armaments) interested in DoE associated with computer simulations. As they have temporal data, I showed them the FDoE functionality. They seem quite enthusiastic and they asked me a question about the target function.
They are not so much interested in defining a target function, but rather in specifying two functions, one defining the upper limit target curve and the other the lower limit target curve. The objective would therefore not be to reach a target curve but to remain absolutely within this envelope, any variations within this envelope is allowed.
Imagine a missile whose trajectory is defined by upper and lower limits.
Is it possible to do this with the FDE platform?
Thanks a lot for your help.
@Steph_Georges I don't think FDE can do this. Probably the best you could do is to create an "average curve" that is between the upper and lower curve and use that as the target to try to optimize for.
If you could, create a wishlist item on the Internal Wish List so we can capture this and track it.
Hi Stephane,
not sure this is of help here, but Astrid Ruck from Autoliv had a presentation on a gage R&R study on curved data. It might be helpful:
https://community.jmp.com/t5/Discovery-Summit-Munich-2020/Measurement-Systems-Analysis-for-Curve-Data-2020-EU-30MP-447/ta-p/243880
here is the corresponding white paper: https://www.jmp.com/en_us/whitepapers/jmp/measurement-systems-analysis-for-curve-data.html
A follow up presentation on this has been done by Colleen McKendry (https://community.jmp.com/t5/Discovery-Summit-Europe-2023/Measurement-Systems-Analysis-for-Curve-Data-Using-Functional/ta-p/572656 )
I like the idea from @SamGardner of an average curve as ideal curve.
