摘要
论述了小波神经网络用于信号分类识别的模型结构，在此基础上，充分利用小波变换时频分析的局部化特性，提出了一种改进的网络结构，建立了非显式小波的网络的学习算法。计算机模拟表明，该结构提高了信号分类识别的精度和灵敏度。
This paper introduced the general model structure of wavelet neural networks (WNN) for signal recognition and classification. By making full use of the advantages of wavelet transform time frequency localization, the paper proposed an improved network structure and presented a learning algorithm for the hidden function wavelet neural network. A simulation on computer showed that using the structure, the precision and sensitivity of signal recognition and classification were improved.
出处
《西南交通大学学报》
EI
CSCD
北大核心
1999年第4期436-440,共5页
Journal of Southwest Jiaotong University
关键词
信号分类
小波神经网络
小波变换
学习算法
signals
classification
model building
recognition
wavelet neural network
