Three-phase pulse width modulation (PWM) rectifiers are crucial in industrial, civil, and transportation applications. Due to the diversity of grid loads, the three-phase input voltage may be unbalanced, which consequently leads to unequal three-phase current sharing and increased current harmonic distortion, severely risking the overheating of cables and circuit breakers. When the three-phase voltage unbalance reaches 20%, the three-phase current deviation can reach 13.7%, and the 3rd harmonic content of the current increases by 3.8%. Therefore, the focus of this paper is to control the current uniformly and reduce harmonics under three-phase voltage unbalance conditions. Based on the Recursive Gauss-Newton (RGN) algorithm, this paper proposes two algorithms: a Negative Sequence Component Elimination (NSCE) algorithm, which is capable of extracting the effective positive-sequence voltage component during voltage unbalance for precise control, and an Adaptive Periodic Current Fluctuation Suppression (APCFS) algorithm. The APCFS method serves the dual purpose of achieving uniform current control and harmonic reduction, while mitigating the harmonics introduced by the dead-time effect. The current deviation is reduced from 13.7% to 2%, and the total harmonic distortion (THD) is decreased from 5.18% to 2.07%. The efficacy of these proposed algorithms is comprehensively validated through theoretical analysis, current harmonic comparison, and transient response tests, including load step and AC voltage step changes. Experimental results obtained from a 10-kW prototype confirm the theoretical analysis and demonstrate that the combined NSCE and APCFS algorithms effectively resolve the negative effects arising from both three-phase voltage unbalance and dead-time.
Building similarity graph...
Analyzing shared references across papers
Loading...
Hongda Song
Li Liu
Xiaolei Che
IEEE Access
Jilin University
Building similarity graph...
Analyzing shared references across papers
Loading...
Song et al. (Thu,) studied this question.
www.synapsesocial.com/papers/69fd7ddcbfa21ec5bbf06214 — DOI: https://doi.org/10.1109/access.2026.3683907