摘要
齿轮箱故障诊断实质上是一个特征提取和状态识别的过程,在旋转机械研究中,其特征频率往往是已知或者可以计算的,而实际测量的机械振动信号往往含有故障特征以外的频率成分,为了避免多余信息的干扰,提出了结合EMD和分形几何算法的齿轮箱有效频率分量的特征提取方法。实验表明,该方法可以有效地计算出不同齿轮故障的特征值,实现了定量表征。
Diagnosis of gearbox fault is essentially a process of feature extraction and state identification,in the study of rotating machinery , characteristic frequencies are often known or can be calculated, the actual measurement of mechanical vibration signals often contain frequency components other than the fault signatures. In order to avoid the interference from redundant information, this article presents an approach for feature extraction of working frequency component in gearbox that combines EMD and fractal geometry algorithm. Experimental resuhs show that this approach can effectively calculate the characteristic values of different gear failures and achieve a quantitative characterization.
出处
《机械制造》
2014年第3期67-70,共4页
Machinery
关键词
故障诊断技术
齿轮箱
Fault Diagnosis Technology Gearbox EMD
