摘要
提出一种基于小波变换理论的小波神经网络模式识别方法，并将该方法应用于旋转机械轴心轨迹的识别。在本文提出的方法中，模式识别分两步来完成。第一步通过小波网络完成对数据的压缩，第二步应用前馈神经网络对第一步中的压缩数据进行学习与分类，从而达到模式识别的目的。
The wavelet neural network pattern recognition method based on the wavelet transform theory is presented and applied into the identification of shaft center orbit of rotating machineryin this paper. The proposed method complete the identification process through two steps. In the first step, the data are compressed by wavelet network. In the second step, the compressed data are learned and classified by feedforward neural network. Thus, the goal of pattern identification is arrived.
出处
《系统工程理论与实践》
EI
CSCD
北大核心
1998年第10期98-103,共6页
Systems Engineering-Theory & Practice
