Open AccessMulti-objective Optimization Design of PMASynRM Based on RBF Neural Network
Author(s) -
Shuai Kang,
Zequan Li,
Libing Zhou,
Jin Wang
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2183/1/012013
Subject(s) - latin hypercube sampling , artificial neural network , sobol sequence , computer science , surrogate model , sensitivity (control systems) , hypercube , finite element method , mathematical optimization , artificial intelligence , mathematics , machine learning , engineering , statistics , monte carlo method , structural engineering , electronic engineering , parallel computing
This paper presents an optimized design method of PMASynRM based on the RBF neural network. Firstly, Sobol sensitivity analysis method is used to analyze the mutual influence of the parameter variables of the motor. Then, in order to establish the surrogate model of the finite element model, the samples are obtained by the Latin hypercube sampling method to train the RBF neural network, and the NSGA-? algorithm is used for multi-objective optimization based on the trained RBF neural network. Finally, the optimization scheme is verified by the results of finite element analysis that the proposed method can provide an optimal design.