Unformatted text preview: the other unit is lost. Round the uniform realizations of the lead times to the nearest integers. Conduct 5 independent replications, each starting (in day 1) with 18 units in inventory and simulating 8 weeks (56 days). Each replication terminates at the end of day 56; so you can ignore outstanding orders. ±rom each replication collect estimates for the average amount of inventory at hand at the start of each day (after potential order deliveries) and the fraction of days a demand shortage occurs. Then use the averages from the 5 replications to compute an approximate 90% conFdence interval for the mean inventory level at the start of a day and the fraction of time a shortage occurs during the time period under study. Make sure that you can perform such a simulation by hand. Note: The site http://www.bus.ualberta.ca/aingolfsson/simulation contains many Excel spreadsheets with queueing simulations....
View
Full Document
 Spring '08
 ALEXOPOULOS
 Monte Carlo method, Computer simulation, mean inventory level

Click to edit the document details