峭度滤波器用于电机轴承早期故障特征提取  

Kurtosis filter in feature extraction for incipient motor bearing fault diagnosis

作　　者：安国庆[1] 秦程[1] 郭立炜[2] 梁永春[1] 

机构地区：[1]河北科技大学电气工程学院,河北石家庄050018 [2]河北科技大学信息科学与工程学院,河北石家庄050018

出　　处：《电机与控制学报》2014年第6期55-60,共6页Electric Machines and Control

基　　金：河北省自然科学基金(f2006000101);河北省高等学校科学技术研究青年基金项目(QN2014205);河北科技大学五大平台开放基金(TY34)

摘　　要：针对电机轴承早期故障特征十分微弱,容易被机器运转时的背景噪声淹没而难以识别的问题,在分析峭度特性的基础上,确定了信号峭度值和噪信比的数学关系。设计了受控于信噪比的峭度滤波器,并将其应用于包络分析的滤波环节,以提取轴承故障的高频调制信息,可避开选取带通滤波器时人为因素的影响。实验结果表明,对于轴承早期故障特征比较微弱的情况,基于峭度滤波器的诊断方法是有效的,将滤波后的信号进行包络分析可凸显故障特征。The incipient motor bearing fault characteristic is so weak that it is very difficult to be discriminated because of the running noises in mechanical devices.On the basis of kurtosis characteristics analysis,the mathematical relationship between the kurtosis value and noise-to-signal ratio was determined.A filter controlled by signal-to-noise ratio was designed to be used in filter link of the envelope analysis.The filter avoiding the influence of personal factor in choosing the band-pass filter,could extract the high frequency modulation information of bearing fault.Experiment results indicate that the method based on kurtosis filter is feasible in the case of the weak incipient bearing fault information,and the fault characteristic is distinct in the envelope spectrum of the filtered signals.

关 键 词：电机轴承 故障诊断 峭度 滤波器 包络分析 

分 类 号：TM343[电气工程—电机]