基于小波包分解与支持向量机的金刚笔钝化识别研究  

Research of the Identification of Diamond Dresser Blunt Using Wavelet Packet Decomposition and Support Vector Machine

作　　者：昝涛[1] 王民[1] 费仁元[1] 

机构地区：[1]北京工业大学北京市先进制造技术重点实验室,北京100022

出　　处：《机床与液压》2007年第6期34-37,共4页Machine Tool & Hydraulics

基　　金：北京市自然科学基金重点项目(4031001-1)资助;北京市先进制造技术重点实验室项目

摘　　要：金刚笔修整器的钝化导致所修整砂轮廓形恶化,对磨削质量产生严重影响。本文研究了基于声发射的金刚笔钝化识别,利用小波包分解频带能量提取特征量,通过支持向量机(SVM)进行钝化的分类识别。试验结果表明,与常用的人工神经网络相比,该方法在小样本的情况下也具有良好的识别效果,适合于实际应用。In grinding process the diamond dresser blunt leads to bad topography of the wheel and has direct affection on the gain of good grinding precision. AE signals were used to identify the diamond blunt. Characteristic quantities were picked up through wavelet packet decomposition and then support vector machine （SVM） was applied to identify the diamond blunt. The result of the experiment shows that compared with traditional ANN methods it has better recognition performance in small sample.

关 键 词：金刚笔钝化 支持向量机 小波包分解 声发射 

分 类 号：TH164[机械工程—机械制造及自动化]