Model predictive control of power converters: An introduction

Tobias Geyer

ABB Corporate Research

Tuesday, February 20, 2018, 11:00

Room 02-012, Georges-Köhler Allee 102, Freiburg 79110, Germany

Abstract: High power converters are operated at low switching frequencies per fundamental frequency, require fast control loops and must often adhere to tight limits on their harmonic distortions. The typically used linear control loops based on averaging and a separate modulation stage struggle to fully exploit the performance potential of such converters. To address this, model predictive control (MPC) is emerging, which achieves high-performance closed-loop control of power electronic systems.

This lecture provides an introduction to MPC for power electronics and discusses two industrially-relevant MPC methods in detail. The first method ensures fast control
responses at very low switching frequencies and low harmonic distortions. The second method improves the low-voltage ride through of large line commutated inverter drive systems. Both methods have been running in the field successfully for several years. The lecture concludes with an assessment of the available MPC methods for power converters and points out future research directions.

The slides of the talk can be found here.