摘要
本文利用基于小波变换的单隐层前馈型神经网络，模拟高度复杂的非线性映射，针对电磁机构的优化问题，小波神经网络使用来自有限元分析的训练信息，通过选取一簇适当的小波基函数合成具有一定拓朴结构的小波网络，且网络的训练过程是对一个凸函数的优化过程，从而能得到全局最优解，学习的收敛速度很快。我们将之应用于交流真空接触器直流激磁系统的优化设计中，取得了较好的效果。
A single hidden layer feed forward neural network based on wavelet transform is discussed in this paper, which can be applied to the approximating of complex nonlinear functions. As to a optimization problem in electromagnetics, the training samples are generated using finite element analysis,the representation of network topology can be definitely developed.Since the training problem can be transformed into an convex optimization process,the global minimum can be obtained and the learning speed is increased. We've applied wavelet neural network to design optimization of A. C. vacuum contactor with D. C. exciting electric circuit and obtained satisfactory scheme.
出处
《中国电机工程学报》
EI
CSCD
北大核心
1996年第2期83-86,共4页
Proceedings of the CSEE
基金
国家自然科学基金
关键词
小波变换
神经网络
电磁场
优化设计
接触器
wavelet transform
localization characteristic
feedforward meural metwork
optimization design in electromagnetics.
