Fast-Processing Implementation of Current-Ripple-Losses-Optimized Common-Mode Voltage Reduction PWM

In this paper, a fast-processing implementation method of the current-ripple-loss-optimized common-mode voltage reduction pulsewidth modulation (CRLO-CMVRPWM) is studied. With the novel sector definitions and the simplified method to solve the constraint nonlinear programming model, the compute burden of the proposed method is reduced by 82.8% and 50.5% for the case where the modulation index is below and above 4√ 3 /9, respectively. In addition, the effects of the optimized zero-sequence voltage errors caused by the adopted linear fitting strategy are proven to be negligible by studying the harmonic characteristics difference between the fast-processing and original CRLO-CMVRPWM. Simulated and experimental results are given to validate the effectiveness.