Skip to main navigation Skip to search Skip to main content

An Improved Model Predictive Torque Control for Switched Reluctance Motors With Candidate Voltage Vectors Optimization

  • Xi'an Jiaotong University

Research output: Contribution to journalArticlepeer-review

89 Scopus citations

Abstract

This article proposes an improved model predictive torque control (MPTC) strategy for switched reluctance motor (SRM) drives with candidate voltage vectors (CVVs) optimization, which can suppress the torque ripple effectively and enhance the system efficiency. This MPTC method is improved by three aspects. First, the flux linkage calculation is removed compared to conventional MPTC. Second, the electric cycle of SRM is divided into six sectors based on the ideal torque contribution profile and the commutation region of the motor is redefined. Finally, CVVs of each sector are adopted based on phase torque characteristics and the total number of CVVs is reduced to 2 or 3 at each control period. The cost function is designed for the torque ripple suppression and copper loss minimization by selecting the optimal voltage vector from CVVs. This improved MPTC method avoids mass useless computation compared to conventional MPTC and no longer relies on hysteresis control loops compared to direct torque control (DTC) method. In the simulation and experimental verification on a three-phase 12/8 SRM, the proposed MPTC method effectively reduces the torque ripple, improves the torque-ampere ratio and system efficiency, and improves dynamic response compared to DTC method.

Original languageEnglish
Pages (from-to)4595-4607
Number of pages13
JournalIEEE Transactions on Industrial Electronics
Volume70
Issue number5
DOIs
StatePublished - 1 May 2023

Keywords

  • Model predictive torque control (MPTC)
  • switched reluctance motor (SRM)
  • torque ripple

Fingerprint

Dive into the research topics of 'An Improved Model Predictive Torque Control for Switched Reluctance Motors With Candidate Voltage Vectors Optimization'. Together they form a unique fingerprint.

Cite this