High-Performance Model Predictive Control for Medium-Voltage Drives
Event details
| Date | 20.05.2011 |
| Hour | 11:15 |
| Speaker | Dr Tobias Geyer, University of Auckland |
| Location |
ELA2
|
| Category | Conferences - Seminars |
Model predictive control (MPC) is rapidly emerging as a new methodology to achieve high-performance closed-loop control of ac drives. MPC relies on an internal model, predictions, an objective function, an optimization stage and the so called receding horizon policy. By combining the inner current control loop with the modulation stage, the gating signals can be directly generated by MPC in one computational stage. MPC computes switching patterns that simultaneously reduce the inverter’s switching losses and the current and torque distortions. At the same time, the drive’s controlled variables, which are either the currents or the torque and stator flux, are kept around their references or within specified bounds. This talk provides an introduction to MPC for ac drives and a review of the different emerging methods. It is shown how the related computational challenges, which are due to the combinatorial explosion of the number of admissible switching sequences, can be addressed. As a result, at low switching frequencies, the resulting steady-state behaviour is similar to the one obtained with optimized pulse patterns. During transients, however, very fast current and torque response times are achieved, similar to deadbeat control. Simulations and selected experimental results will be presented to illustrate these performance characteristics, using three- and five-level medium-voltage drives with induction machines as illustrative examples.
Practical information
- General public
- Free