Analysis of Capacitor Charging Characteristics and Low-Frequency Ripple Mitigation by Two New Voltage-Balancing Strategies for MMC-Based Solid-State Transformers

Lei Zhang, Jiangchao Qin, Yuntao Zou, Qing Duan, Wanxing Sheng

Research output: Contribution to journalArticle

Abstract

This article investigates and compares various modulation methods and capacitor voltage-balancing algorithms of a modular multilevel converter for solid-state transformer applications. Characteristics of capacitor charging and discharging are analyzed for the existing single-step alternating voltage balancing and the conventional sorting algorithms with phase-shift (PS) modulation and nearest level control (NLC) methods. Based on the analysis, the single-step alternating algorithm leads to a low-frequency voltage ripple, while the conventional sorting algorithm introduces additional switching actions. To address these issues, two new voltage-balancing algorithms are proposed: a multistep alternating voltage-balancing algorithm, and a currentless sorting algorithm. The performance of the proposed algorithms with the PS and NLC modulation methods is comprehensively investigated and compared based on theoretical analysis and experimental results. The study results demonstrate the capability of the proposed methods to balance capacitor voltages while reducing the low-frequency voltage ripple and avoiding additional switching actions.

Original languageEnglish (US)
Article number9110715
Pages (from-to)1004-1017
Number of pages14
JournalIEEE Transactions on Power Electronics
Volume36
Issue number1
DOIs
StatePublished - Jan 2021

Keywords

  • Capacitor voltage balancing
  • capacitor voltage ripple
  • modular multilevel converter (MMC)
  • solid-state transformer (SST)

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Analysis of Capacitor Charging Characteristics and Low-Frequency Ripple Mitigation by Two New Voltage-Balancing Strategies for MMC-Based Solid-State Transformers'. Together they form a unique fingerprint.

  • Cite this