Open AccessSynchronous reluctance machine geometry optimisation through a genetic algorithm based technique
Author(s) -
Ruba Mircea,
Jurca Florin,
Czumbil Levente,
Micu Dan D.,
Martis Claudia,
Polycarpou Alexis,
Rizzo Renato
Publication year - 2018
Publication title -
iet electric power applications
Language(s) - English
Resource type - Journals
ISSN - 1751-8679
DOI - 10.1049/iet-epa.2017.0455
Subject(s) - rotor (electric) , torque , stator , matlab , control theory (sociology) , genetic algorithm , inductance , magnetic reluctance , torque density , finite element method , computer science , switched reluctance motor , process (computing) , control engineering , engineering , mechanical engineering , magnet , physics , voltage , artificial intelligence , structural engineering , electrical engineering , machine learning , control (management) , thermodynamics , operating system
In this study, the design optimisation of a synchronous reluctance machine for light electric vehicles is proposed, to increase efficiency and reduce torque ripples. The existing machine was structurally optimised, using dedicated genetic algorithms, replacing only the rotor and keeping the stator and it's winding untouched. Starting from the original design of the rotor implemented in Flux2D, a finite element analysis software, and the genetic algorithm optimisation implemented in Matlab, a complex co‐simulation was accomplished to obtain a rotor architecture that increases the machine's performances and decreases the torque ripples. By this, performing rotor skewing is not needed any more, hence the torque loss due to it was cancelled. The optimised rotor design increases the machine performances by higher mean torque, no skewing, <8% torque ripples, higher efficiency and better inductance characteristics. Comparative results obtained both in simulations and experimental measurements prove positive outcomes of the optimisation process.