摘要
对滚动轴承的故障类型识别已有很多研究,但很少涉及到其故障等级,即损伤程度的检测与识别。文中采用小波包多层分解的方法,提取滚动轴承的振动信号的能量谱,经归一化后,结合RBF为核函数的支持向量机,对美国Case Western Reserve大学的轴承数据中心的滚动轴承规范数据集进行研究测试,取得很好的实验效果。
Though the faults of rolling bears were studied much more, the fault levels for rolling bears was rarely researched. This paper proposed a method of fault levels identification for rolling bearing based on wavelet power spectrum and support vector machine （SVM）. And the method applied to the data of Case Western Reserve University Bearing Data Center, results showed that this method performed well.
出处
《机械工程师》
2009年第6期70-72,共3页
Mechanical Engineer
关键词
滚动轴承
损伤等级
小波能量谱
支持向量机
rolling bears
fault levels
wavelet power spectrum
support vector machine
