Question

Q 2b18.

Ms. Case must determine a replication length and run time strategy for her simulation. She believes that a 30-day period, after the identification of the transient, is sufficient for her purposes and makes some sense, given the historical data she has available for a similar system that was installed by the Ontario Liquor Control Board in their warehouse in Whitby, Ontario.

Nico has completed a pilot run of 10 replications of 30 simulated days, using the run strategy of batch means. She receives the following output:

   

Rep	Throughput
1	3500
2	3851
3	4522
4	4756
5	4872
6	4020
7	4618
8	4968
9	4632
10	4548

a) Use the data from the pilot runs to prove to Nico that a 30-day replication length is insufficient.

b) Provide an algorithm (and advice) for determining a proper run length for this model.

c) Compare the run strategies of "batch means" and "replication/deletion". How are these implemented? What are the advantages/disadvantages of each?

d) Assume that Ms. Case wishes to get around the problem in part (a) by implementing a "replication/deletion" strategy with a run length of 30 days. Would this solve the problem? Why or why not?

NOTE: This is for Industrial Engineering stream for Simulation & Modelling Analysis subject. Since there is no options for IE in drop down menu, it is asked under ME stream. Please help in solving all a, b, c, & d sets.

Answer 1