Comparative study of two predictive direct power control algorithms for three-phase AC/DC converters

In this paper two predictive direct power control (P-DPC) algorithms developed by authors for a three- phase AC/DC converter are presented and compared. In the first, a variable switching frequency (VSF) algorithm, behavior of active and reactive power is predicted for all available voltage vectors generated by the AC/DC converter and the voltage vector which minimizes a cost function is selected for the next sampling period. The second, a constant switching frequency (CSF) algorithm, in every sampling period selects the voltage-vectors´ sequences and computes their application instants in order to minimize tracking error of active- and reactive-power. Experimental results measured on 5 kVA laboratory AC/DC converter show performance of both P-DPC algorithms in stationary and transient states.