摘要
针对摩擦焊接头超声检测信号,介绍了一种用于信号分类的小波神经网络结构及其学习算法,用小波包分析工具提取信号特征作为网络输入,实现了宏观焊接缺陷和微观焊接缺陷的分类识别.实验结果表明,较BP网络小波神经网络获得良好的识别结果。
Considering the ultrasonic inspection signal in friction welded joints, a kind of structure of wavelet neural network used for the signal classification and its learning algorithm is introduced. We use wavelet package analytical tools to extract characteristic of the signal and action characteristic number to the importation of wavelet neural network. The classification of macro welded defects and tiny welded defects are carried out. The result showed that comparing BP neural network a wavelet neural network acquired to a good identifying result.
出处
《机械设计与制造》
北大核心
2007年第1期87-89,共3页
Machinery Design & Manufacture
关键词
摩擦焊
小波神经网络
小波包
识别
Friction welding
Wavelet neural network
Wavelet package
Identification
