Optimal variable switching frequency scheme for reducing switching loss in single-phase inverters based on time-domain ripple analysis

Xiaolin Mao, Raja Ayyanar, Harish K. Krishnamurthy

Research output: Contribution to journalArticlepeer-review

161 Scopus citations

Abstract

The choice of switching frequency for pulsewidth modulation single-phase inverters, such as those used in grid-connected photovoltaic application, is usually a tradeoff between reducing the total harmonic distortion (THD) and reducing the switching loss. This paper discusses an approach to minimize the switching loss while meeting a given THD requirement using variable switching frequency schemes (switching schemes with the switching frequency varying within a fundamental period). An optimal switching scheme is proposed based on time-domain current ripple analysis and the calculus of variations. The analysis shows that, to meet the same THD requirement, the optimal scheme has a significant saving on switching loss, compared to the fixed switching frequency scheme and the hysteresis control scheme, in addition to other benefits such as reduced peak switching loss and a spread spectrum of the current harmonics. The optimal scheme has been implemented in a prototype and the experimental results have verified the theoretical analysis. Also, a straightforward design method for designing filter inductors for single-phase converters is provided based on the time-domain current ripple analysis.

Original languageEnglish (US)
Pages (from-to)991-1001
Number of pages11
JournalIEEE Transactions on Power Electronics
Volume24
Issue number4
DOIs
StatePublished - 2009

Keywords

  • Calculus of variations
  • Photovoltaic (PV)
  • Pulsewidth modulation (PWM)
  • Single-phase converter
  • Switching loss
  • Variable switching frequency

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Optimal variable switching frequency scheme for reducing switching loss in single-phase inverters based on time-domain ripple analysis'. Together they form a unique fingerprint.

Cite this