电器多目标优化方法综述

Overview on Multi-objective Optimization Methods of Electric Apparatus

基于电器的优化设计是典型的多目标优化问题,梳理了国内外电器及相关领域的多目标优化方法研究概况,并从多目标优化算法、多目标最优解的决策、近似模型快速计算3个方面对电器的多目标优化问题进行介绍。在多目标优化算法方面,主要使用正交试验、多目标智能算法等;在多目标最优解的决策方面,主要使用加权求和、模糊隶属度函数、熵权理想点法等;在近似建模快速计算方面,主要使用磁路法、响应面、神经网络近似模型等。最后,结合实际工程应用预测了未来电器领域多目标优化发展的方向。

The optimal design of electric apparatus is a typical multi-objective optinfization problem. Based on it, this paper summarizes the research situation of multi-objeetive optinfization methods in electric apparatus and related fields at home and abroad. Then multi-objeetive optinfization problem is introduced from three aspeets：multi- objective optinfization algorithm, decision making method and approximation models method. In the aspect of multi- objective optimization algorithm, the optimization of electric apparatus mainly uses orthogonal experiments and multi- objective intelligent algorithms. In the aspect of decision making method, weighted summation, fuzzy membership function and entropy weight ideal point method are mainly used. In the aspect of approximation models, magnetic path method, response surface and neural network approximation models are the main methods. Finally, this paper combines the practical engineering application to predict the research direction of multi-objective optimization in the future electric apparatus field.