基于Kriging代理模型的无刷双馈发电机转子优化设计  

Rotor Optimization Design of Brushless Doubly-fed Generators Based on Kriging Surrogate Model

作　　者：于思洋[1] 田培均 王皓 张凤阁[1] YU Siyang;TIAN Peijun;WANG Hao;ZHANG Fengge(School of Electrical Engineering,Shenyang University of Technology,Shenyang 110870,China)

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870

出　　处：《微电机》2024年第3期47-51,64,共6页Micromotors

基　　金：国家自然科学基金项目(52007124);沈阳市中青年科技创新人才支持计划项目(RC220248)。

摘　　要：无刷双馈发电机由于其自身没有电刷、滑环等结构特点,因此被认为是最有希望成为风力发电系统的主流发电机之一。无刷双馈发电机是依靠转子的磁场调制机理实现机电能量转换,其转子的耦合能力是影响电机性能的重要因素。因此本文提出了一种串联笼条辅助磁障转子的新型转子结构,分析了该种转子不同结构参数对电机性能的影响,并提出一种基于Kriging代理模型的多目标差分进化算法的无刷双馈发电机转子优化设计方法。首先建立该种新型转子的参数化模型,通过Kriging代理模型建立优化变量与优化目标之间的响应关系,再利用多目标差分进化算法找到最优的变量组合,最后通过有限元仿真对电机的性能进行分析。The brushless doubly fed generator,seen as a promising mainstream candidate for wind power systems due to the absence of brushes and slip rings,relies on rotor magnetic field modulation for electromechanical energy conversion.A novel series cage bar-assisted magnetic barrier rotor structure was proposed in this paper.It investigated the impact of different rotor structural parameters on motor performance and introduces a rotor optimization design method for the brushless doubly fed generator using a multi-objective differential evolution algorithm based on the Kriging surrogate model.The method started with establishing a parametric model of the innovative rotor,then used the Kriging model to map the relationship between optimization variables and objectives,followed by the multi-objective differential evolution algorithm to identify the optimal variable combination,and finally,finite element simulation to analyze the motor s performance.

关 键 词：无刷双馈发电机 笼障转子 KRIGING模型 多目标优化 差分进化算法 

分 类 号：TM34[电气工程—电机]