An Enhanced Model-Free Predictive Control to Eliminate Stagnant Current Variation Update for PWM Rectifiers

Abstract

In model-free predictive control (MFPC), high active power spikes and grid current notches are induced when applied to pulsewidth modulation (PWM) rectifiers. To overcome these problems, we develop an enhanced MFPC to eliminate the stagnant current variation update for PWM rectifiers. In the proposed MFPC, only four variables are estimated to remove the stagnant current variation update, and the grid current variations are obtained directly from these four estimated variables. Therefore, there is no need to identify any voltage vector combination, and a lookup table to store grid current variations is not needed in contrast to the previous method. Thanks to the elimination of the stagnant current variation update, the prediction accuracy of the proposed MFPC is significantly improved without complexity compared with that of the conventional MFPC. As a result, the active power spikes and grid current notches are eliminated with the proposed MFPC. Both simulation and experiment are implemented to verify the effectiveness of the proposed MFPC.