IOE 574. Simulation Analysis
Fall 2007
Tue-Thur 12:00-1:30 in Room IOE 2717
Revised 12-6-07
Instructor: Mark
P. Van Oyen, Assoc. Prof. of IOE, Room 2853
I prefer you send email - vanoyen"at"umich.edu
Office Hours: Monday 1:00-2:00pm and Fri 3:00-4:00pm; other times by appointment
Detailed Description: A course in discrete-event simulation for graduate students. The course covers system modeling, simulation design (including mechanisms for efficiency), analysis of output, and programming in general-purpose languages (such as Visual Basic for Applications in Excel or C++). Proper design and analysis of discrete-event simulation experiments is emphasized. Applications are drawn primarily from manufacturing, service systems, and healthcare. Includes fundamentals as well as the more advanced concepts that allow students to model and to analyze systems using custom simulations at a deeper level than courses based on off-the-shelf simulation packages. Topics include stochastic models for simulation, statistical methodology for designing simulations and output analysis, random variable and process generation, and efficiency improvement techniques.
Note: IOE 474 is a more basic course that emphasizes the use of ProModel as a simulation package. It is not necessary to have IOE 474 prior to IOE 574.
IOE grad students may choose to take a qualifying exam on the topic of this course.
Text: A. M.
Law 2007. Simulation Modeling and Analysis, 4th Edition,
McGraw-Hill.
(We previously used the 3rd edition, which has ISBN: 0070592926
Note that it is not necessary to have "Expertfit" software included with the book. We will have Statfit available via CAEN.
Prerequisites:
Familiarity with basic probability, statistics, and Poisson processes; computer programming in some programming language (for example, C, Java, VBA, or Matlab); Background in Markov chains and queueing is helpful, but not required; Graduate standing is recommended, but undergraduates with IOE 316 and a statistics course have the essential background. IOE 515 is listed as a prerequisite in the catalog, but it is not required if you have the necessary relevant background.
Grading: (tentative)
Simulation homework & projects:
45%
Midterm: 25%
Final Examination: 30%
The
simulation "homework projects" (plan on 2 or 3 of them) may be done individually or in a team of 2 students.
All other homework is to be done individually. When a project is done as a team,
the group writeup must be submitted as a single copy bearing the names of all
group members (and both members of the group generally share the same grade).
The midterm and final examination will be open book and notes.
Course Schedule (tentative):
|
Mod ule |
Topic |
(L&K = Law & Kelton) |
Other |
|
1 |
Discrete Event Simulation modeling |
L&K Chapter 1
Get familiar with VBA (see Intro to VBA & Excel .ppt and
lec2-formulationSimModels.pdf = lecture notes
|
Note that the Ctools website https://ctools.umich.edu/portal has Assignment functionality to upload .xls or .c attachments to the website, and I can view your work through that mechanism.
Intro to Queueing: 574queue-lec-A.ppt 574queue-lec-B.ppt
|
|
2 |
Programming simulations |
L&K Ch 2
Tue Sept. 25 at start of class. |
Problem discussed in class: InClassPracticeFormulation.doc |
|
3 |
Developing Simulations
& using “Simlib” |
L&K Chapter 5 - Building Credible Sim's.
Lecture Notes: 9-Intro to RV Generation & Event Scheduling.pdf
|
More complex Modeling_Complex_Systems.xls (contains "TandemWalk" example from class)
|
|
4 |
Input modeling Simulation as a stochastic process |
L&K Chapter 6 L&K Chapter 4.3 and 4.6
|
Read L&K Chapter 4 only to the extent you need the review.
Lecture notes: 13-Poisson.pdf
|
|
5 |
Random-number generation |
L&K Chapter 7.1-7.3.2
|
Lecture notes: 14-Generating Random Variates.pdf; |
| 6 | Choosing distributions for Modeling | EvaluatingFit.ppt | |
|
MIDTERM TEST |
Tuesday Nov. 6 in class, proctored. (open notes and book) |
||
|
7 |
Output analysis |
16-Sim-as-Stoch-Proc-B.pdf posted 11-11 = Stochastic Process notes - UPDATED.
L&K Chapter 9 assignment: hwk5.pdf due 11-19. MeanPlots.ppt = example from class 17-9_5_2 Replication-Deletion.pdf (posted 12-6-07) |
Modeling_Complex_Systems_BLN_MVO.xls
|
|
8 |
Comparison, ranking, and
selection of alternatives |
L&K Chapter 10
Brief intro to comparisons: Comparisons Example.ppt
Discussed assignment on Batch means (lag-1 process, but it was not assigned
19-Comparisons.pdf (updated 12-12-07) |
Arrival Data for B's: |
|
|
Variance reduction |
L&K Chapter 11 through section 11.3
Discussion of Random Variate Generation and streams: 13B-RanNumberGenerators.pdf posted 12-12-07)
|
Related Presentation:
|
| 10 |
Application of Simulation to Research |
Simulation and its role in research. | |
|
|
Thursday, December 13 4:00 pm for 2 hours,
open book and notes.
The details of Projects 2 and 3 will not be tested. |
For those doing the class project, the project will be due any time on or before December 20. |

Interpretation of the Honor Code for this class:
You may obtain help from others in doing the homework and projects (with the exception that some assignments may explicitly disallow teamwork), but the purpose of teamwork is to help you to fully comprehend the material. While you may discuss assignments with multiple people, actual completion of an assignment must be limited to a team of at most two (2) members. A homework assignment or project that is prepared as a team should be submitted as a single copy containing the names of all group members (and all members will share the grade given).
Supplementary Resources - For Your Information Only:
sim-probability-facts.pdf = Helpful details on computing sample variance, arrival counting processes, parameterizing a Poisson process, timing diagrams for queues, and generating pseudo-random numbers (for illustrative purposes, not practical)
Tutorial: intro to statistical issues and Stat::Fit: Making Models Statistically Valid.ppt
Inventory Management (basic intro.): Chapter02-SSmodel.ppt (see the latter half, emphasizing The Single Product (Q,r) Model)
Tips for Success with Simulation: Sim Pro Model Introduction.ppt
C and Fortran code for Law and Kelton is available from http://www.mhhe.com/engcs/industrial/lawkelton/