Analysis and Design Considerations of a Dual-Rotor Multiple-Winding Machine

Hang Zhao, Chunhua Liu, Zaixin Song, Senyi Liu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

9 Citations (Scopus)

Abstract

In this article, a novel dual-rotor multiple-winding machine (DRMWM) is proposed for the continuous variable transmission (CVT) in hybrid electric vehicles (HEVs). First, a mathematical model of the proposed machine is constructed by using the harmonic modeling method. Then, on the basis of the magnetic field analysis, the operating principle of the DRMWM is deduced. By introducing multiple windings with two different working harmonics, namely the permanent magnet (PM) harmonics, and the magnetic gear harmonics, the proposed DRMWM can cope with various operating conditions of HEVs. To further obtain optimal performance, a finite-element analysis and multiobjective genetic algorithm II combined sequential optimization is carried out for the proposed DRMWM. The proposed optimization method is proven to be more computationally efficient compared to the global optimization method. Finally, a prototype is manufactured and tested in various operating modes. The results validate the simulation and analytical method predictions. Also, it proves that the proposed DRMWM is suitable to work as the CVT in HEVs.

Original languageEnglish
Pages (from-to)8727-8738
Number of pages12
JournalIEEE Transactions on Industrial Electronics
Volume69
Issue number9
DOIs
Publication statusPublished - 1 Sept 2022
Externally publishedYes

Keywords

  • Continuous variable transmission (CVT)
  • dual rotor
  • harmonic modeling method (HMM)
  • hybrid electric vehicle
  • operating mode
  • permanent magnet (PM) machine
  • sequential optimization

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Analysis and Design Considerations of a Dual-Rotor Multiple-Winding Machine'. Together they form a unique fingerprint.

Cite this