Normal distribution with Multi-objective Design Study?

[mloew]

Has anyone figured out how to populate a table in the Multi-objective Design Study with a normal distribution? As far as I can tell, one can only get a uniform distribution from a Monte Carlo simulation. I would like to do some simple statistical studies without using a add-on product like CETOL.

Any suggestions would be appreciated. Best regards,

Matthew Ian Loew

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora

 

[darcyo]

from my experience with multi-opjective design studies, the only options to filter (or derive) the database are to:
1. keep goal between min and max values
2. using pareto filter, to obtain the max, or min for the populated database.

as far as a normal distribution, i am assuming that you mean to determine a number of variables that will produce the same goal value. if this is what you mean, then no - there isn't a way, since no two goal values have exactly the same value. the only way is to derive a table from teh master table between a high and low bound, which are very close in value. you might have to create a rather large database initially, (200+) variables, in order to get any kind of useful result.

let me know if i'm way off, and i will try again.

ps- MLoew - check out the post that i left on the bicycle engineering fourm - you might be interested.

 

[mloew]

Darcyo,

Thanks for the reply, but as you feared, not quite what I was looking for. :-( When Pro/E populates the space for each variable in a design study, the distribution is pretty uniform as a result of the Monte Carlo simulation. If the distribution followed a normal distribution (also called a Gaussian distribution and commonly referred to as a bell-shaped curve) the data points are concentrated near the mean and tapers off towards the tails. See

http://mathworld.wolfram.com/NormalDistribution.html

for a more detailed explanation. This is the type of distribution one commonly expects from a manufacturing process. One would know the variance (sigma²) from the manufacturing process.
I am sure I could generate the population if I could import the points from Excel. But I know this can not be done. Any tips? Best regards,

Matthew Ian Loew

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora