Objectives
- Nonlinear optimization – MATLAB implementation
- Optimization approaches: dynamic programming, Calculus of Variations
- Linear quadratic and H∞ compensators – stochastic and deterministic
- Investigate key basic control concepts and extend to advanced algorithms (MPC)
- Will focus on both the technique/approach and the control result
Approximate Number of Lectures per Topic
Keywords
LQR = linear-quadratic regulator
LQG = linear-quadratic Gaussian
MPC = model predictive control
Number of lecture topics.| NUMBER OF LECTURES | TOPICS |
|---|
| 2 | Nonlinear optimization |
| 3 | Dynamic programming |
| 2 | Calculus of variations – general |
| 3 | Calculus of variations – control |
| 5 | LQR/LQG - stochastic optimization |
| 3 | H∞ and robust control |
| 2 | On-line optimization and control (MPC) |
Grades
Grading criteria.| ACTIVITIES | PERCENTAGES |
|---|
| Homework: problem sets every other Thursday due 2 weeks later (usually) at 11 am | 20% |
| Two midterms: both are in class, and you are allowed 1 sheet of notes (both sides) for the first, 2 sheets for the second | 25% each |
| Final exam | 30% |
Prerequisites
- Course assumes a good working knowledge of linear algebra and differential equations. New material will be covered in depth in the class, but a strong background will be necessary.
- Solid background in control design is best to fully understand this material, but not essential.
- Course material and homework assume a good working knowledge of MATLAB.
Policies
- You are encouraged to discuss the homework and problem sets. However, your submitted work must be your own.
- Late homework will not be accepted unless prior approval is obtained from Professor How. Grade on all late homework will be reduced 25% per day. No homework will be accepted for credit after the solutions have been handed out.