How to run a DOE experiment with different standard deviations?

Question

question

Liam Chiu asked Sep 20, '22 Jacob W2 commented Oct 4, '22

How to run a DOE experiment with different standard deviations?

HI,

I am trying to do a DOE experiment in a model.

There are different processing times in each tool in the model, and I want to use three different standard deviations (5, 10, 15) to do experiment. I want to know how I can set it up to run different scenarios. The kpi I want to get is the output of each buffer under different scenarios (5, 10, 15).

Currently I can only use different models separately, I would like to know if there are other ways to experiment?

Thank you for your advance. :)

https://drive.google.com/file/d/1SBb9TF2x3PmgV6yaPdhAsbNRO70JXrv0/view?usp=sharing

Software Version:

FlexSim 20.1.3

scenarios doe

· 4

Joerg Vogel commented · Sep 20, 2022 at 04:12 PM

Use Parameters for number values. Or labels. In experimenter you can exchange the values then.

normal(myMeanValue, myDevValue,getstream(current))
// with labels
normal(current.myMeanValue, current.myDevValue,getstream(current))

instead of current reference you can choose a different object, Then you must not add to many object parameters into your experimenter.

0 ·

Liam Chiu Joerg Vogel commented · Sep 20, 2022 at 04:24 PM

hi @Joerg Vogel

Thanks for your advice

I'm also wondering, if I have hundreds of machines, is there any way to change the standard deviation for all of them at once? Thanks~

0 ·

1663690787095.png (20.3 KiB)

1663690824419.png (35.6 KiB)

1663690912396.png (9.1 KiB)

Joerg Vogel Liam Chiu commented · Sep 20, 2022 at 09:20 PM

There has been introduced a new feature called property table in FlexSim 20.2.0, which does exactly your request. But before this version you can only copy variables, labels, values by edit tool (Edit or View main menu item) called “ edit selected objects” or you write a script, which step through members of a group and change their attributes.

0 ·

Jacob W2 ♦ commented · Oct 04, 2022 at 05:28 PM

Hi @Wei_cheng Chiu, was Jason Lightfoot's answer helpful? If so, please click the "Accept" button at the bottom of their answer. Or if you still have questions, add a comment and we'll continue the conversation.

If we haven't heard back from you within 3 business days we'll auto-accept an answer, but you can always unaccept and comment back to reopen your question.

0 ·

Answer 1 · 2022-09-28T13:54:47Z

Jason Lightfoot answered Sep 28, '22 Joerg Vogel commented Sep 29, '22

For the process time you can use something like :

Math.max(0,normal(30,Model.parameters.StdDev,getstream(current)))

1664373240098.png (122.2 KiB)

· 1

Joerg Vogel commented · Sep 29, 2022 at 12:22 AM

This approach is a filter to prevent negative values from a statistical distribution. Math.max(0,’functional value, which becomes possible negative’).

1 ·

question