基于连续小波变换的信号检测技术与故障诊断  

FEATURE DETECTION AND FAULT DIAGNOSIS BASED ON CONTINUOUS WAVELET TRANSFORM

作　　者：林京[1] 屈梁生[2] 

机构地区：[1]中国科学院声学研究所声场声信息国家重点实验室,北京100080 [2]西安交通大学

出　　处：《机械工程学报》2000年第12期95-100,共6页Journal of Mechanical Engineering

摘　　要：通过分析指出，连续小波变换具有很强的弱信号检测能力，非常适合故障诊断领域。从参数离散到参数优化系统研究了连续小波变换的工程应用方法，建立了“小波熵”的概念，并以此作为基小波参数的择优标准。论文最后把连续小波技术应用在滚动轴承滚道缺陷和齿轮裂纹的识别中，诊断效果十分理想。it is pointed out that continuous wavelet transform (CWT) has powerful ability for weak signal detection which help itself to be used for fault diagnosis. The method for parameter discretization and optimization of CWT is estabished. The concept of wavelet entropy is introduced and it is used as a rule for parameter optimization. In the end, CWT is ed for fault diagnosis of rolling bearing and sear-for. Very good results are obtained using this method.

关 键 词：故障诊断 滚动轴承 齿轮 连续小波变换 信号检测 

分 类 号：TH133.33[机械工程—机械制造及自动化] TH132.41