Simulation parameters settings

[Vajooo]

Hello everybody,
i'm dealing with a complex series-parallel system consisting of blocks with different MTBF, ranging from 24834 to 300000 hours (I suppose for each block an exponential distribution). I have the problem of facing with always different simulation results, as i change simulation time and number of simulations parameters.
Is there a common rule and/or best practise for setting this two parameters in such an optimum way that the results of the simulation can be accurate?

[prakash]

Hi Vajooo
Changing Simulation Parameters such as Seed and Increments will change the output very little but changing End Time and No. of Simulation would normally produce different results.

Hope this helps.:)

Thanks

Prakash

[simulation11]

Sorry I can not help you now, but you can wait for the others' help.:(

simulation rachat credit (http://simulationrachatcredit.com)

[meow44]

simulation rachat de credit (http://simulationrachatdecredit.org)
this is a very informative post. thanks so much!

[Pantelis]

My general rule of thumb is to simulate 10^(sig_figs*+1).

Sig_Figs being the desired accuracy in terms of significant figures.