Showing results for 
Search instead for 
Do you mean 
Reply
Solved! Go to solution

Random Number Stream Logic

[ Edited ]

Hallo,

 

Question in Short: Does the random number stream always start from the first digit when the stochastic variable itself is changed as above, i.e., by assigning new parameters, such as a,b, or c in case of triangular -- or 'mu' and 'sigma' in case of normal? 

 

I am using version 8.2 and so have to manage random number streams myself.  I handled this by writing creation controls that allocate unique number streams to all stochastic variables in the model automatically.

 

However, the ProcTime is supposed to be MU specific and based on a custom MU variable 'Qty'. Basically, there are steel cages filled with parts and so I want to base the ProcTime on the specific number of parts in the steel cage.

 

To achieve this, I am currently doing the following: Assuming the distribution is triangular, I am calculating 'a', 'b', and 'c' based on 'Qty' and the fixed parameters for a,b,c ProcTimes per piece. Then, I allocate this these numbers to the corresponding SingleProc's stochastic triangular variable. Again, the stochastic variable already has a number stream assigned to it. The hope is that the random number stream will just continue with the new parameters.

 

If this is not the case, my current procedure isn't working and other ideas are very much appreciated.

 

Best regards,

 

Sebastian

1 REPLY
Solution
Solution
Accepted by topic author SebastianE
‎09-19-2016 05:15 AM

Re: Random Number Stream Logic

[ Edited ]

Plant Simultion does not reset the random number stream when you change the distribution parameters. So everything works like you expected, i.e. new random numbers will be generated.