Excell add-ins: In this chapter and the next, we will be using the three Excel add-ins TreePlan, RiskSim and SensIt. Please install these add-ins at your earliest convenience. Installation instructions are provided in this file.
There are several published papers on simulation with links on the book web site. For example, one paper describes how simulation was used as Taco Bell. Another talks about managing credit lines and prices for Bank One Credit Cards. Yet another one describes the application of simulation in General Motors OnStar Project. You can access these and others using your Access Code that came with the textbook.
Simulation with Excel: Here is the Excel file for this problem. Note that the worksheet named RL = 50 simulates the problem for only 50 replications. The other one named RL = Long does it for 1000 replications.
t-table: This table is needed to find confidence intervals for the true mean when variance is no known.
A simple example with RiskSim: In this example, we are simulating the total revenue obtained per day when the number of customers shopping in our store is Poisson with mean 70 and each spending a uniform amount of money with range [20,40].
Simulation of the New Product Marketing Problem with RiskSim: We can solve this problem much more easily using RiskSim. Here's the very simple Excel file for this method.
We solve this inventory problem with RiskSim which calculates the expected profit easily.
Here is an Excel file that generates random numbers using the mixed congruential method.
Note: Check the Help feature in Excel for the RAND() function to see how Excel does these things.
Here is another Excel file that generates exponential random variates using Excel's RAND() function.
RiskSim's simulation of uniform, exponential and triangular random variables.
Here's an example of discrete event simulation (manually done) over a 12 minute interval.
We can simulate more general queues using the Excel file Queueing_Simulator.xls. In order for this to work, you should do the following: Excel Options > Trust Center > Trust Center Settings > Macro Settings > Enable All Macros.
We will first discuss a simple case of the M/M/1 queue (Innis Library case) with mean 1/λ = 1/100 for interarrivals, and mean 1/μ = 1/120 for service time. The simulated results will agree what we found from our formulas. Here's the Excel file. The theoretical results for this problem were L = 5, Lq = 4.17, W= 3 min (0.05 hr) and Wq = 2.5 min (0.04 hr).
We will also discuss an example with G/E(4)/5 where G is Uniform(0,6) and 1/μ = 10 for E(4), i.e., k = 4. This would be impossible to analyze using any formulas.