Course 7: Advanced Optimization: Techniques and Industrial Applications

(Syllabus)

Last Update: July 7,  2020

Course Objective: The objective of this course is to provide the students with knowledge on the application of various optimization techniques which can help making decisions for practical problems in industries. Modeling concepts and applications of linear, integer, nonlinear, and dynamic programming as well as network models are addressed.  Meta-heuristic techniques are also discussed to obtain good solutions for large scale practical problems in a reasonable computational time. Optimization model and its applications are demonstrated for solving problems in Industry 4 era.

Learning Outcomes: The students on the completion of this course would be able to:

• • • Formulate mathematical programs for practical problems in production and supply chain systems (Create).
    • Apply appropriate optimization techniques and write codes of optimization models using professional optimization software (i.e., MATLAB, LINGO, or MPL software) to solve single-objective practical problems in production systems, supply chain systems and specific operational problems (Create).
    • Find appropriate trade-off solutions for multiobjective decision making problems in production systems, supply chain systems and specific operational problems (Create).
    • Use meta-heuristic techniques to solve large scale NP-hard combinatorial problems for both single and multiple objective decision making problems where analytical methods cannot be used (Create)
    • Conduct sensitivity analysis to examine the robustness of the solutions resulting from optimization models in order to ensure that appropriate solutions will be deployed in real world situations where input parameters are uncertain and cannot be estimated precisely (Evaluate).
    • Understand how to apply digital technology for automated data-driven and in real-world optimization model. (Apply).

Prerequisite: Operations Research

 Course Outline:

 

Week	Module	Topic	Learning  Materials	Teaching Materials	Note
1	Module 1: Theory of mathematical programming for convex optimization	1.1 Basic Modeling Concepts	MSIE-07-L-M1S1-01	MSIE-07-T-M1S1-01
2		1.2 Linear Programming	MSIE-07-L-M1S2-01	MSIE-07-T-M1S2-01
3		1.2 Linear Programming (cont.)
4		1.3 Integer Programming, Mixed Integer Programming, and Combinatorial Optimization	MSIE-07-L-M1S3-01	MSIE-07-T-M1S3-01
5		1.4 Non-linear Optimization	MSIE-07-L-M1S4-01	MSIE-07-T-M1S4-01
6		1.5 Dynamic Programming	MSIE-07-L-M1S5-01	MSIE-07-T-M1S5-01
7	Module 2: Heuristics and Metaheuristics	2.1 Concept of Heuristics and Metaheuristics	MSIE-07-L-M2S1-01	MSIE-07-T-M2S1-01
8		2.2 Population-based algorithms: GA, PSO, DE	MSIE-07-L-M2S2-01	MSIE-07-T-M2S2-01
9		2.2 Population-based algorithms: GA, PSO, DE (cont.)
10		2.3 Local Search Methods: ALNS and Tabu Search	MSIE-07-L-M2S3-01	MSIE-07-T-M2S3-01
11		2.4 Multiobjective optimization	MSIE-07-L-M2S4-01	MSIE-07-T-M2S4-01
12	Module 3: Optimization and Its Applications in Industry 4 Era	3.1 An Overview of Digital Technologies 3.2 Optimization (Opt) concept and Its Applications in Industry 4.0 Era	MSIE-07-L-M3S1-01 MSIE-07-L-M3S2-01	MSIE-07-T-M3S1-01 MSIE-07-T-M3S2-01
13		3.3 Optimization (Opt) Design in Industry 4.0	MSIE-07-L-M3S3-01	MSIE-07-T-M3S3-01
14		3.4 Real-Time Optimization	MSIE-07-L-M3S4-01	MSIE-07-T-M3S4-01
15		3.5 Case Study	MSIE-07-L-M3S5-01	MSIE-07-T-M3S5-01

Laboratory Sessions: None

Learning Resources:

Textbooks: No designated textbook, but class notes and handouts will be provided.

Reference Books:

1. 1. Operations Research: An Introduction, 9th Edition, Hamdy A. Taha, Pearson Education, 2013.
   2. Linear Programming and Extensions, George Dantzig, Princeton University Press, 2016.
   3. Model Building in Mathematical Programming, 4th Edition, H. Paul Williams, Wiley, 1999.
   4. Operations Research: A Practical Introduction (Operations Research Series) 1st Edition, Michael W. Carter, Camille C. Price, CRC press, 2000.

Journals and Magazines:

• 1. European Journal of Operational Research, Elsevier
  2. International Journal of Production Research, Taylor and Francis
  3. International Journal of Production Economics, Elsevier
  4. Journal of the Operational Research Society, Palgrave Macmillan
  5.  Management Science, Informs

Teaching and Learning Methods:

This is a project-based learning (PBL). During lecture sessions, class discussions and group projects will be conducted. Group projects are included in all modules in order to make students are able to solve the real-world optimization models. Students will practice various skills included developing mathematical programming, heuristics, and metaheuristics. Most importantly, students will be able to apply digital technology for automated data-driven used in the real-world optimization models.

Time Distribution and Study Load:

Lectures: 45 hours

Self-Study: 30 hours

Group Project: 15 hours

Discussion: 10 hours

Evaluation Scheme:  

An “A” would be awarded if a student can demonstrate clearly skills in developing mathematical programming, heuristics and metaheuristics to solve real-world optimization models. Additionally, a student can show clearly understanding how to apply digital technology for automated data-driven used in real-world optimization models.

An “B” would be awarded if a student can show good progress in developing mathematical programming, heuristics and metaheuristics to solve real-world optimization models. Additionally, a student can show good understanding how to apply digital technology for automated data-driven used in real-world optimization models.

An “C” would be awarded if a student can show reasonable progress in developing mathematical programming, heuristics and metaheuristics to solve real-world optimization models. Additionally, a student can show reasonable understanding how to apply digital technology for automated data-driven used in real-world optimization models.

An “D” would be awarded if a student shows lack of improvement  in developing mathematical programming, heuristics and metaheuristics to solve real-world optimization models. Additionally, a student shows lack of improvement how to apply digital technology for automated data-driven used in real-world optimization models.