DEE-34406 Model Predictive Control of Power Electronic Systems, 5 cr
Additional information
Suitable for postgraduate studies.
Person responsible
Petros Karamanakos
Lessons
| Implementation | Period | Person responsible | Requirements |
| DEE-34406 2019-01 | 1 - 2 |
Petros Karamanakos |
Final written exam; homework helps to pass the course |
Learning Outcomes
The concept of averaging is commonly applied to power electronics systems to conceal the switching aspect from the control problem. This enables the use of linear control loops, but also limits the achievable performance during transients. To achieve the highest possible performance for power converters, averaging is to be avoided and the traditionally used current control loop and modulator should be replaced by one single control entity. To this end, model predictive control (MPC) is particularly promising, since it can address switched nonlinear systems, constraints, and systems with multiple inputs and outputs. This course reviews MPC methods that fully exploit the performance potential of power converters, by ensuring fast control during transients and low harmonic distortions during steady-state operation. Such MPC methods are particularly suitable when operating at low pulse numbers (medium-voltage power electronics and traction converters) or when considering complicated systems such as converter systems with filters and modular multi-level converters. The objective of the course is to bridge the gap between modern control methods and power electronics systems, and to provide an understanding of MPC methods for power electronics systems. This includes MPC methods both without and with a modulator. An emphasis is put on computational methods to solve the underlying (integer) optimization problems. Exercises based on industrial power electronics systems are used to consolidate the presented control theoretic concepts and to assist the learning process. Matlab/Simulink simulations further improve the understanding of the control methods, their benefits and intrinsic challenges.
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Modeling of power electronic systems (derivation of state-space models) | Direct MPC with reference tracking (long horizons) | Optimized pulse patterns (OPPs) |
| 2. | Direct MPC with reference tracking-short horizons (formulation of the current control problem) | Branch-and-bound optimization techniques | Fast control of OPPs |
| 3. | Mathematical programming (enumeration technique to solve the short-horizon problem) | Model predictive direct torque control (MPDTC) | Model predictive pulse pattern control (MP3C) |
| 4. | Delays and their compensation | Derivatives of MPDTC (e.g., Model predictive direct current control-MPDCC) | Indirect MPC for modular multilevel converters |
| 5. | Applications to drive systems consisting of a three-level neutral point clamped (NPC) inverter and an induction machine, inverters with LC filters |
Prerequisites
| Course | Mandatory/Advisable | Description |
| DEE-33116 Power Electronics Converters | Advisable |
Additional information about prerequisites
* Control systems (such as ASE-1251 Järjestelmien ohjaus or ASE-1258 Introduction to Control, and ASE-2110 Systeemit ja säätö or ASE-2116 Systems and Control)
* Power electronic systems (modeling and analysis) (DEE-33010 Sähkökoneet and DEE-33030 Sähkömoottorikäytöt)
* Matlab
Correspondence of content
There is no equivalence with any other courses