摘要
基于数值仿真技术分析干式空心电抗器的结构参数对其涡流损耗的影响,并结合工程实际,建立考虑绕组截面填充结构、导线截面形状、气道宽度和每个包封内层数等因素的通用模型。为提高干式空心电抗器涡流损耗的计算精度,根据输出响应和输入参数之间的函数关系,建立基于指数型径向基函数(RBF)神经网络的干式空心电抗器涡流损耗计算模型,并采用粒子群算法和梯度下降法对网络参数进行优化。算例结果分析表明,基于RBF神经网络的干式空心电抗器涡流损耗模型具有计算精度高和速度快的优点,适用于干式空心电抗器的优化设计。
Based on numerical simulations, the structural parameters of dry-type air-core reactor were analyzed for the effect on eddy current losses. A unified model in engineering practice was then proposed to consider the fit scheme of winding cross section, the shape of conductor cross section, the airway width, and the number of layers per package. In order to improve the computational accuracy of reactor eddy current losses, a radial basis function（RBF） neural network model was established, in which the exponential function was determined as the activation function according to the relationship between the input and output variables. Moreover, an improved particle swarm algorithm for optimizing network parameters was presented. Numerical results indicate that the proposed model exhibits the highest precision and best computational performance. As a result, this model applies especially to the optimum design of dry-type air-core reactors.
作者
陈锋
王嘉玮
吴梦晗
马西奎
Chen Feng;Wang Jiawei;Wu Menghan;Ma Xikui(State Key Laboratory of Electrical Insulation and Power Equipment Xi’an Jiaotong University Xi’an 710049 China;Shanghai Sieyuan Electric Corporation Limited Shanghai 201100 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2018年第11期2545-2553,共9页
Transactions of China Electrotechnical Society
基金
国家自然科学基金资助项目(51407139)
关键词
交流电阻
干式空心电抗器
径向基函数
神经网络
粒子群算法
AC resistance
dry-type air-core reactor
radial basis function
neural network
particle swarm algorithm
作者简介
陈锋,男,1979年生,博士,讲师,研究方向为电力设备的优化设计和特性分析.E-mail：chenf@mail.xjtu.edu.cn（通信作者）;王嘉玮,男,1990年生,博士研究生,研究方向为工程电磁场数值计算及其软件技术.E-mail：wangjiawwei@outlook.com
