Students will obtain in the course knowledge of the concept of Model Predictive Control (MPC), they will know how to formulate MPC problems as convex optimization problems, they will be able to choose a proper type of objective function and constraints, and and they will know how to solve MPC numerically and implement it in closed loop.
Prerequisites for registration:
The course is divided into three main parts. The first one introduces the concept of model predictive control (MPC) and shows its analogies to optimal control. Second part describes mathematical fundaments required to formulate MPC problems as convex optimization problems. The final part discusses various formulations of MPC, including regulation towards non-zero references, removal of regulation offsets, and output regulation.
Recommended or required reading:
MACIEJOWSKI, J M. Predictive Control with Constraints. Harlow : Prentice Hall, 2002. 331 s. ISBN 0-201-39823-0.
RAWLINGS, J B. – MAYNE, D Q. Model Predictive Control: Theory and design. Madison : Nob Hill Publishing, 2009. 533 s. ISBN 978-0-975-93770-9.
Planned learning activities and teaching methods:
lecture 1 hour per week, practice 2 hours per week
Assesment methods and criteria:
Evaluation is based 50% on the work during the semester, 50% from the final project defense. The rating is based on the standard FCHPT scale.
Language of instruction:
Institute of Information Engineering, Automation and Mathematics was established in 1.1.2006 from two departments: Department of Information Engineering and Process Control and Department of Mathematics.