Cancel
Showing results for 
Search instead for 
Did you mean: 

Different model behaviour in version 11 and 10.1

Experimenter
Experimenter

Hello altogether,

 

For my thesis, I want to use a model which was programmed with Plant Simulation 10.1 in Plant Simulation 11.

I opened the model in version 11, updated all libraries the program wanted me to and saved the model.

Now, when I do a simulation run, the timeshare "waiting" of my stations is way higher than in version 10.1. I didn't change anything in the model itself, so I guess it has something to do with the updated version?!

 

Does anyone have a similar problem or an idea what changed from version 10.1 to 11 that causes these differences?

 

Thanks so much for your help!

11 REPLIES

Re: Different model behaviour in version 11 and 10.1

Phenom
Phenom

Hi,

 

have a look in the V11 help under  "what's new " -"Eminent Changes in Plant Simulation 11"

 

There you will read , that Plant Simulation now automatically uses a dedicated random number stream for each material flow object

 

If your model is using random numbers, this might be the cause

Re: Different model behaviour in version 11 and 10.1

Experimenter
Experimenter

Hi!

Thanks so much for your quick response! Unfortunately, the problem is still the same, when I don't use any random numbers in the model. Smiley Frustrated
Do you have any other ideas?

 

Thanks so much for your help!

Re: Different model behaviour in version 11 and 10.1

Siemens Phenom Siemens Phenom
Siemens Phenom

Hi aedel,

 

are you sure that you don't use random numbers?

 

As soon as you use a distributed time or an availability less than 100% you depend on the changed random number generation.

 

If this isn't the case, can you post your model?

 

Regards,

Ralf

Regards,

Ralf
------------------------------------------------------------------------------------------------------
Did you like the answer? Then click the Thumbs Up button.
Did the answer solve your problem? Then accept the answer as solution.

Re: Different model behaviour in version 11 and 10.1

Experimenter
Experimenter

Hi,

 

I took all random numbers out of the model to test the behavior then and it was the same problem. The original model has random numbers.

I'm programming the model for a company so sadly, I'm not allowed to post the model, because of data security reasons.

 

Thanks!

Re: Different model behaviour in version 11 and 10.1

Phenom
Phenom

Hi

 

I suppose the quickest way to find out the apparently different behaviour  of your model ( w/o random numbers) between v10 and v11 is to check at what event there is a difference

between the eventlists v10 and v11. ( this is easy to check)

 

I think this would be interesting for all of us

Re: Different model behaviour in version 11 and 10.1

Siemens Phenom Siemens Phenom
Siemens Phenom

Hi aedel,

 

could you ask your customer if it is okay to send us the model per email?

You can then send it to plantsimulation.studentlicense.plm@siemens.com.

 

Or you contact our GTAC support (https://support.industrysoftware.automation.siemens.com/gtac.shtml).

 

Regards,

Ralf

Regards,

Ralf
------------------------------------------------------------------------------------------------------
Did you like the answer? Then click the Thumbs Up button.
Did the answer solve your problem? Then accept the answer as solution.

Re: Different model behaviour in version 11 and 10.1

Experimenter
Experimenter

Hey altogether,

 

sorry, that I didn't answer for so long.

I wasn't able to find our what's causing the difference between the two software versions.

Sadly I can't send you the mothel either :/

 

I now spend the last days to redo my model in software version 10.1. and another problem has come up.

I'm measuring the OEE of the system. From the data I put in, the model should have an OEE of around 65%. Sure there are differences because of the sochastic influences but when I do 10 simulation runs with the same input parameters, the first one works out as i would expect it to (around 65), but the nine following ones are all around 70%.

 

What is different about the first run of an experiment?

 

Thank you so much for your help!

Re: Different model behaviour in version 11 and 10.1

Siemens Phenom Siemens Phenom
Siemens Phenom

Hi Aedel,

 

two observations of an experiment are significantly different. Experience shows that the analysis of the reason is complicated and time consuming.

 

Go to the Advanced settings (menu Tools) of the Experimentmanager. On the tab Validation you can set the parameters of the observation. Because of the changed random number the material flow simulation in the versions 10.1 and 11.0 will differ.

 

Prepare a plot of the output value using the basic object Diagram. Perhaps it is shown that the two curves have a similar shape. You do not have to worry about the outlier.

Otherwise, you see the moment when something has changed.

 

Regards,

Peter

Re: Different model behaviour in version 11 and 10.1

Experimenter
Experimenter

Hi Peter,

 

thank you for your quick response!

The differences I was talking about weren't between software version 10.1 an 11 but within software version 10.1.

 

I start an experiment with 10 observations and get the following results:

 

observation 1: 0.6476
observation 2: 0.6993
observation 3: 0.6918
observation 4: 0.6940
observation 5: 0.6791
observation 6: 0.7014
observation 7: 0.6964
observation 8: 0.6992
observation 9: 0.6604
observation 10: 0.6918

 

So somehow, there is a gap betwen the first observation and the nine following ones. The differences within observations 2 to 9 I'd say, are caused by the normal stochastic influences. But why aren't they around the 65% of the first observation?

 

Is there anything different about the setting up process for the first observation in comparison to the one of the others?

 

Thanks again!