Open AccessDesign and multi‐objective optimisation of switched reluctance machine with iron loss
Author(s) -
Yan Wenju,
Chen Hao,
Liu Xuekun,
Ma Xiaoping,
Lv Zhongwei,
Wang Xing,
Palka Ryszard,
Chen Lei,
Wang Kai
Publication year - 2019
Publication title -
iet electric power applications
Language(s) - English
Resource type - Journals
ISSN - 1751-8679
DOI - 10.1049/iet-epa.2018.5699
Subject(s) - torque , torque ripple , switched reluctance motor , sensitivity (control systems) , control theory (sociology) , taguchi methods , sorting , flux linkage , engineering , computer science , voltage , mechanical engineering , algorithm , electronic engineering , direct torque control , induction motor , rotor (electric) , physics , control (management) , artificial intelligence , machine learning , electrical engineering , thermodynamics
In this study, the design optimisation of a switched reluctance machine (SRM) with the layered method has been studied. Firstly, a multi‐physical analytical model for the SRM is established. The proposed model consists of an electromagnetic model, an electrical model, a loss model, and a thermal model. Then a layered optimisation design method suitable for the SRM is proposed in combination with the multi‐physical field simulation model. Taguchi method is used to analyse the influence degree of the main geometric dimensions on the dynamic performance, and the parameters are divided into high‐sensitivity parameters (HSPs) and low‐sensitivity parameters. The non‐dominated sorting genetic algorithm‐II algorithm is used to optimise the HSP by the proposed model. The iron loss, average torque, average torque per unit mass, torque ripple and efficiency are taken as the objective functions. Finally, the prototype is manufactured and a testing platform is built. The flux linkage of several special positions of the machine is verified experimentally, and the accuracy of the electromagnetic model in the multi‐physical field model is verified. The temperature experiments are carried out on the machine, and the accuracy of the equivalent heat circuit model in the multi‐physical field model is verified.