An improved low-computation model predictive direct power control strategy for three-level four-leg active power filter

Model predictive direct power control (MPDPC) strategy offers numerous benefits over the traditional strategies in active filters (APF). However, real-time execution duration (RTED) reduction of model multilevel inverters is a big challenge. In this paper, an improved low-computation for three-level four-leg APF (TL-FL-APF) presented. First, based on generalized instantaneous reactive theory, new TL-FL-APF established. Then, order to reduce RTED, classical multiple predictions are converted into single voltage prediction through principle deadbeat control, and optimized vector range reduced from 81 4–18 by method spatial stratification using γ-component reference vector. Finally, simulation experimental results demonstrate that proposed MPDPC has excellent performance.