振动故障信号奇异性指数的统计特征研究  

Studies on Statistical Features of Singularity of Faulty Vibration Signals

作　　者：申弢[1] 黄树红[1] 张燕平[1] 侯敬宏[1] 

机构地区：[1]华中科技大学能源与动力工程学院,武汉430074

出　　处：《振动工程学报》2003年第4期476-479,共4页Journal of Vibration Engineering

基　　金：国家自然科学基金资助项目(编号 :50 10 50 0 4)

摘　　要：基于信号奇异性检测的小波变换理论 ,利用时间分辨率优化的 Gabor小波变换工具 ,对几种典型故障的波形和轨迹信号的奇异性特征进行了研究。研究表明 ,不同类型的故障信号 ,不仅奇异性指数的数值大小有明显的不同 ,而且奇异点在时间上的分布特征也有显著差异 ,这种差异对轴心轨迹信号尤为明显。因此 。Based on the theory of singularity detection by wavelet transformatio n, the statistical features of singularity of several faulty vibration signals a nd orbits are studied with Gabor wavelet transformation, which parameters are op timized for best frequency resolution. It is shown that not only the Lipschitz E xponent, but also the singularity distribution along time, differ between the fa ulty signals, and it is more evident for orbits. So the statistical features of singularity can be used to differentiate vibration faults effectively.

关 键 词：故障诊断 奇异性指数 旋转机械 振动信号 特征提取 GABOR小波 信号处理 

分 类 号：TH17[机械工程—机械制造及自动化] TH113.1