The aim of the course is to understand control system design based on the approaches of optimal and predictive control to processes described by both parametric and non-parametric model. Based on the practical examples the quadratic optimal control, model predictive control with prescribed dynamics, predictive deadbeat control and predictive control under constraints (saturation etc.) are demonstrated. The course topics cover

1. Solving linear (LP) and quadratic (QP) programming problems
2. Optimal control problem.
3. Gradient-based optimization methods.
4. Bellman principle of optimality, dynamic programming.
5. Linear quadratic regulator, algebraic (matrix) Riccati equation.
6. Development of optimal control design based on linear matrix inequalities (LMI)
7. Model Predictive Control (MPC) algorithm – definition and synthesis.
8. Receding Horizon Control (RHC) – comparison between the control in finite and infinite horizon.
9. MPC design using output prediction based on plant state estimation.
10. MPC design strategy with pole placement, deadbeat control.
11. MPC design constrained on the control input (saturation etc.).
12. Multivariable MPC: challenges from industry.
13. Optimal and predictive control problems: Case studies

Ogata K.: Modern Control Engineering. Prentice Hall, Boston, 2002.
Camacho E. F. and Bordons A. C.: Model Predictive Control. Springer-Verlag, London, 2004.
Goodwin, G. C., Graebe, S. F., and Salgado, M. E.: Control System Design, Pearson, 2001.
Kwakernaak H. and Sivan R.: Linear optimal control systems. Wiley &Sons, Inc., New York, 1972

Texts for distance study:
Before commencing on the course „Optimal and predictive control systems“ the electronic materials to lectures are available at https://moodle-vyuka.cvut.cz/.