基于小波分析和神经网络的渗碳层深度分类  

Classification of Carburized Layer Depth Based on Wavelet Analysis and Neutral Network

作　　者：陈祯[1] 游凤荷[1] 

机构地区：[1]武汉理工大学机电工程学院,湖北武汉430070

出　　处：《武汉理工大学学报（信息与管理工程版）》2008年第5期678-680,684,共4页Journal of Wuhan University of Technology：Information & Management Engineering

摘　　要：提出了一种基于小波理论的新的特征值提取方法,较全面地反映了信号的时频特征,并将小波包提取的特征值输入到BP网络,对7种不同渗碳层深度的试件进行了分类。实验结果表明,小波包特征值提取和BP神经网络分类器相结合,可以实现对不同渗碳层深度的分类,其效果良好、精度较高,有一定的实用价值。A new way to feature extraction originated from wavelet theory was proposed, which is a very good method to characterize the feature both in the time field and in the frequency field. Input features extracted by the way described above were put into a BP neutral network by classifying seven types of different earburized layer depth specimens. The experiment result indicates that the combination of wavelet packet method to extract features with the BP neutral network can implement the classification of different metal earburized layer depths, h is effective and precise with some practical value.

关 键 词：渗碳层深度 小波分析 特征提取 神经网络 

分 类 号：TN911.72[电子电信—通信与信息系统]