基于高阶统计量的机械故障特征提取方法研究  

A Method for Extracting Mechanical Faults Features Based on Higher Order Statistics

作　　者：张桂才[1] 史铁林[1] 杨叔子[1] 

机构地区：[1]华中理工大学机械科学与工程学院

出　　处：《华中理工大学学报》1999年第3期6-8,共3页Journal of Huazhong University of Science and Technology

基　　金：国家"九.五"攀登计划预选资助

摘　　要：对高阶统计量用于机械故障特征提取进行了研究．首先利用Ｈｉｌｂｅｒｔ变换构造原始信号的解析信号，求取信号的包络，然后计算包络信号的高阶统计量．研究表明，用高阶统计量提取信号特征，可以容易地将正常齿轮信号和齿轮裂纹、断齿的信号分离．The application of higher order statistics extracting of mechanical falut features is studied. The analytical signals are acquired by means of Hilbert transformation of the original singals. The enveloped signals are calculated from the analytical signals. The higher order cumulants and moments of the enveloped signals are estimated. The results of the research show that normal gear signals, cracked gear signals and broken gear signals can be easily separated by using higher order statistics as the signal features.

关 键 词：高阶统计量 特征提取 故障诊断 机械设备 统计量 

分 类 号：TH17[机械工程—机械制造及自动化]