Difference in Results Between Initial Model Run and Optimization Scenario

Dear FlexSim Community,

I recently ran my model using our predefined settings. Following this, I conducted an optimization process with 300 scenarios, and the optimization identified Scenario 300 as the best-performing scenario. By coincidence, the settings of Scenario 300, including parameters such as source quantity and batch size, are exactly the same as those in my initial model run.

Despite having identical settings, the results between the two runs are different. The quantity of finished products, the number of completed product types, and the cycle time do not match. Given that all input conditions remain unchanged, I would like to understand the underlying reason for these discrepancies.

Could someone help explain why this happens and provide a reasonable logical explanation to this situation?

Thank you in advance for your time and assistance.

Software Version:

FlexSim 22.2.4

· 3

5 |100000

Attachments: Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

Joerg Vogel commented · Feb 17 at 07:06 AM

@AmyTsai, is there any random generator in your model?

0 ·

Felix Möhlmann commented · Feb 17 at 07:34 AM

Without seeing the model I can think of three possible causes:

1. Is the optimizer runnning multiple replications per scenario and you are comparing averages to the results of a single run?

2. Similarly, what is the highlighted option below set to on your end? Maybe the live model is using the random number stream belonging to a different replication rank than what the optimizer is using?

3. Your model is not repeatable (possibly depends on the drawing of the 3d view in some way). Do you get the same results if you run the model live but at different speeds or without the 3d view being open at all?

0 ·

1739777535157.png (3.2 KiB)

AmyTsai Felix Möhlmann commented · Feb 18 at 12:58 AM

Thank you so much for taking the time to provide these suggestions. I truly appreciate your insights and the possible explanations you have outlined.

I will review the points you mentioned and check whether any of these factors are influencing the differences in results. Your guidance is very helpful, and I really appreciate your support in helping me better understand this issue.

Thanks again for your time and assistance!

0 ·