Predictive Control With Novel Virtual-Flux Estimation for Back-to-Back Power Converters

Abstract

This work proposes and experimentally verifies a novel virtual flux (VF) estimation method for the predictive control of back-to-back voltage-source power converters with a space vector modulator (SVM). The novelty of the proposed approach is the development of a time domain initial bias compensation virtual flux (IBC-VF) estimation method and its combination with a predictive control scheme with constant switching frequency. Theoretical analysis and implementation procedures of the proposed method are developed in detail. Compared to two conventional bandpass filter-based VF estimation methods, the proposed IBC-VF method shows a much faster dynamic response. The whole IBC-VF-based predictive control scheme is realized in the $\alpha\beta$ frame and so does not require Park's transformation. The proposed control scheme is implemented on a commercial off-the-shelf field-programmable-gate-array-based platform. By using a single-cycle timed loop (SCTL) technique, the whole computation is finished within 2.6 $\mu\hbox{s} $ , making the compensation of the calculation time obsolete. Both simulation and experimental results emphasize the effectiveness of the proposed scheme.