Syllabus

ME6105 Syllabus for Fall 2017

Instructor: Chris Paredis

Room: MARC 256
Phone: 404-894-5613 (office)
Email: chris.paredis@me.gatech.edu
Web: http://www.mbsec.gatech.edu/users/cparedis
Office Hours:  1/2 per week per team -- team schedule will be established during the semester
Consult my calendar (check because the schedule may change due to travel)

Lectures

Room:  Love 184
Time:  Tue-Thu 4:30-5:45

Prerequisites

Graduate standing in engineering or related discipline;  Undergraduate seniors with permission of the instructor.

Prerequisites by topics: The students are expected to be familiar with the following topics at an undergraduate level:

  • Statistics and probability theory
  • Differential equations
  • Design topics: specifications, functional requirements, conceptual design, embodiment and detailed design.
  • The behavior of physical systems: thermal systems, electrical systems, dynamics of mechanical systems, hydraulics (to complete their projects, the students will need familiarity with at least a subset of these topics).
  • Computer-aided design and engineering (CAD/CAE)

Course Goals

To provide a theoretical foundation for the development
and use of models in system design. The models include both
decision models and analysis models

 

Upon completion of this course, the student should be able to:

  • frame decisions: objectives, alternatives, outcomes, preferences.
  • evaluate design alternatives by conducting simulation studies
  • select the appropriate modeling paradigm to support a design decision
  • select a solution algorithm that matches the characteristics of an analysis model
  • critically evaluate analysis results in the presence of uncertainty
  • model designer preferences -- risk aversion, time preferences, robustness
  • recognize the trade-offs between the costs and value of different simulation-based design processes

Course Materials

Web Page: All materials will be posted on the official course web-site: http://www.mbsec.gatech.edu/education/ME6105. In addition, homework and project submissions need to be submitted electronically via t-square: http://www.t-square.gatech.edu

Textbook:

"Fundamentals of Decision Making for Engineering Design and Systems Engineering," by George Hazelrigg.  Available only via the following web-site: http://www.engineeringdecisionmaking.com/ (Note that this book is self-published by the author and therefore not available anywhere else -- don't look for it on amazon...)

In addition, specific chapters or overview articles will be provided on-line in the Reading Material section.

References:

  • Making Hard Decision: An Introduction to Decision Analysis, R.T. Clemen, Duxbury Press, 1997. (ISBN: 0534260349)
  • Introduction to Systems Engineering, A.P. Sage, J.E. Armstrong Jr., Wiley & Sons, 2000. (ISBN: 0471027669)
  • Introduction to Modeling and Simulation of Technical and Physical Systems with Modelica, Peter Fritzson, Wiley-IEEE Press, 2011. (ISBN: 978-1118010686)
  • Continuous System Simulation, F.E. Cellier and E. Kofman, Springer, 2006. (ISBN: 0387261028)
  • Simulation with Arena, 3rd edition, W. Kelton, R. Sadowski, D. Sturrock, McGraw-Hill, 2003. (ISBN: 0072919817)

Topics

  1. Course Overview and Introduction
    • Modeling of energy-based systems
    • Object Oriented Modeling in Dymola
  2. Modeling the structure of design problems
    • Modeling the structure of design problems:  Influence diagrams
    • Modeling value functions in Value-Driven Design
    • Introduction to the economics of design: profit, cost, demand
  3. Modeling of energy-based systems
    • What is modeling and simulation?
    • Types of simulation models and tools
    • The Modelica Language
    • Solving differential (algebraic) equations
    • Debugging Modelica Models
  4. Value-driven design
    • Value functions and trade-offs
    • Preferences over time
    • Cost modeling
    • Demand modeling
  5. Modeling uncertainty
    • Probability basics
    • Uncertainty in design
    • Representation and elicitation of uncertainty
    • Bayes Theorem
    • Computing with uncertainty information
  6. Design optimization under uncertainty
    • Utility theory
    • Optimization in design
    • Surrogate modeling
    • Numerical solution of design problems under uncertainty
  7. Wrap-up
    • Comparison of value-driven design and utility theory to other design methods
    • Multi-level design and decision delegation
    • Game theory and its relevance in design and systems engineering
    • Information Modeling for Systems Engineering -- SysML

Software

The homework assignments will involve two software packages: Dymola and ModelCenter.  Detailed installation instructions can be found in the Reading Materials section of the course web-site (login required.  I will share this password during the first lecture, and list it in the "In-Class Notes" section on t-square).

  • Dymola: An academic license is available and the software will be distributed to all students via download. The software can only be used in the context of this course.  Find more information about the Dymola software at http://www.3ds.com/products/catia/portfolio/dymola.  Complete installation instructions can be found here (requires password provided during the first lecture).  Several Modelica models can be found in the Project Folders from previous years.
  • ModelCenter: An academic license is available and the software will be distributed to all students via download.  The software can only be used in the context of this course. Find more information about the ModelCenter software at www.phoenix-integration.com.  To help you get started, we have created a getting started page (requires password provided during the first lecture)
  • All other software tools (e.g., Spline CDF excel sheet, demand model sheet, etc.) can be found in the Resources folder on t-square.

Projects

A major component of this course are the projects. You are allowed (encouraged) to work in teams of 3-4 students. You are encouraged to choose a topic for your project that is of particular interest to you and that is in an application domain in which you have prior expertise.  This will allow you to tailor the course to your specific interests and needs. If your thesis research includes a component that relates to modeling and simulation, you may be able to develop that component further as part of your course project.

The project is also closely tied to the homework assignments. For each of the homework assignments you will model for a different aspect of the same application as for your project. In the project you can then leverage your results of the homework assignments into a complete simulation-based design study.

Given the focus on Value-Based Design, you are strongly encouraged to select a project topic that relates to a consumer product.  This will facilitate the assignment on demand modeling.

Grading Scheme

There are no exams.  The grade will be based primarily on a comprehensive course project that is divided into 4 group homework assignments:

  • Homework Assignment G1:  Planning your simulation-based design study (10%)
  • Homework Assignment G2:  Energy-based modeling with Modelica (15%)
  • Homework Assignment G3:  Value-driven problem definition (15%)
  • Homework Assignment G4:  Uncertainty Analysis (15%)
  • Homework Assignment G5:  Design optimization (10%)
  • Final Project Presentation (5%)

In addition to the group assignments, there will be three individual assignments:

  • Homework Assignment I1:  Becoming familiar with object-oriented modeling in Dymola (10%)
  • Homework Assignment I2:  Bayesian probability theory (10%)
  • Homework Assignment I3:  Decision theory (10%)

Peer Evaluation

As part of each group homework assignment, you will be required to fill out a (very) brief, peer evaluation survey.  The survey consists of a numerical characterization of the relative contributions made by all teammates.  In addition, you will have an opportunity to leave a comment for the instructor.  The survey results will only be accessible to the instructor. 

Based on the results of these surveys, your homework grade will be adjusted.  This is simply to encourage all teammates to contribute equally to the success of their team.  The grade adjustment is computed automatically, but the teacher reserves the right to make further changes based on the anonymous comments provided by the team mates.  The total adjustment will be in the [-5%,+5%] range for each assignment.

Distance Learning

  • All lecture materials and assignments will be made available electronically through the course web-site.
  • All assignments are due 1 week after the on-campus deadlines (DL deadlines will be listed explicitly in the assignments)
  • All homework submissions should be uploaded to t-square
  • Graded assignments will be e-mailed back to students in pdf format
  • Office hours for DL students:  I like interaction.  Don't hesitate to call or contact me through skype (ID: chris.paredis).  When you have modeling problems with either Dymola or ModelCenter, make sure to e-mail your model to me before you call.

Additional Class Policies

Honor Code Policy: The members of the Georgia Tech Community believe that the fundamental objective of the Institute is to provide the students with a high quality education while developing in them a sense of ethics and social responsibility. We believe that trust is an integral part of the learning process and that self-discipline is necessary in this pursuit. We also believe that any instance of dishonesty hurts the entire community. It is with this in mind that we have set forth a Student Honor Code at Georgia Tech.

Office Hour Policy: We fully support an open-door policy for answering any questions you might have concerning this class. In fact, we appreciate students stopping by to ask questions. However, we hope that you: a) recognize that we have other duties as a faculty member and a student, b) respect our time, c) understand that we might not be able to see you immediately. Therefore, when stopping by outside office hours, call or send e-mail first.

Changes: When appropriate or necessary, the instructor may adjust, amend, or otherwise modify the information presented in the syllabus. Changes will be made in a manner to minimize disruption and in the interest of fostering learning. Every effort will be made to ensure that all changes are brought to the attention of students so as to minimize inconvenience.