Ripple-Reduced Model Predictive Direct Power Control for Active Front-End Power Converters with Extended Switching Vectors and Time-Optimised Control

Hui Fang,Zhenbin Zhang,Xiaoyun Feng,Ralph Kennel
DOI: https://doi.org/10.1049/iet-pel.2015.0857
IF: 2
2016-01-01
IET Power Electronics
Abstract:Model predictive control (MPC) is an attractive technique for power electronics and drives. It requires no extra modulators and considers the non-linear nature of the power converter, so remarkable control performance can be achieved. However, since the conventional MPC scheme uses only one switching state during the whole control interval, ripples of the control variables are more evident than that of the classical modulator based techniques. This study proposed a ripple-reduced model predictive direct power control (RR-MPDPC) scheme for the grid-tied active front end using a fully field programmable gate array-based platform. The novelty of the proposed RR-MPDPC is that through extended switching vectors and time-optimised control, more freedoms and precise tracking possibility are realised in the predictive controller. The proposed control scheme is compared with the conventional MPDPC and the recently reported duty optimal MPDPC scheme (DutyOpt-MPDPC) through experimental data. Experimental results confirm that better performance is achieved using the proposed RR-MPDPC scheme.
What problem does this paper attempt to address?