基于加权峭度的滚动轴承故障特征提取  

Extracting Fault Features of Rolling Bearings Based on Weighted Kurtosis

作　　者：陈祥龙[1] 张兵志[2] 冯辅周[1] 江鹏程[1] 

机构地区：[1]装甲兵工程学院机械工程系,北京100072 [2]北京特种车辆研究所,北京100072

出　　处：《装甲兵工程学院学报》2017年第4期46-51,共6页Journal of Academy of Armored Force Engineering

基　　金：军队科研计划项目

摘　　要：针对峭度谱(Kurtogram)无法有效区别振动信号中的瞬态故障冲击和脉冲噪声,难以准确提取微弱的滚动轴承故障特征的问题,提出一种基于加权峭度(Weighted Kurtosis,WK)的滚动轴承故障特征提取方法,通过固定设置滤波带宽,利用加权峭度识别共振中心频率,确定带通滤波器的滤波中心频率和带宽,结合包络分析提取滚动轴承故障特征,并通过采集变速箱滚动轴承振动数据对该方法进行了验证。结果表明:该方法能够有效克服峭度谱的缺陷,稳健识别滚动轴承共振中心频率,准确提取滚动轴承故障特征,验证了该方法的有效性。As the Kurtogram cannot effectively differentiate noise impulse and transient fault shock in the vibration signal, it is difficult to extract weak fault features of rolling bearings, a novel method based on Weighted Kurtosis（WK） is proposed to extract fault features of rolling bearings in this paper. The proposed method utilizes a WK to identify the resonant central frequency and confirm the band-pass filter central frequency by setting triple fault frequency as filtered bandwidth, coupling with envelope analysis. Finally, rolling bearing vibration data sampled from gearbox are ultilized to verify the efficiency of the proposed method. The results show that, the proposed method can effectively overcome the defects of Kurtogram, steadily identify resonant central frequency band, and accurately extract fault features of rolling bearings. The validity of the method is verified.

关 键 词：峭度谱 加权峭度(WK) 滚动轴承 特征提取 

分 类 号：TH133.3[机械工程—机械制造及自动化] TP206[自动化与计算机技术—检测技术与自动化装置]