基于参数优化的MCKD的滚动轴承早期故障诊断  

Early Fault Diagnosis of Rolling Bearings Based on Parametric Optimized MCKD

作　　者：潘昕怡 岳建海[1] PAN Xinyi;YUE Jianhai(School of Mechanical,Electronic and Control Engineering,Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]北京交通大学机械与电子控制工程学院,北京100044

出　　处：《噪声与振动控制》2021年第5期109-113,218,共6页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(W17A300080)。

摘　　要：针对滚动轴承早期故障被淹没在噪声信号下特征信号微弱、故障特征难以提取的问题,将最大相关峭度解卷积(maximum correlated kurtosis deconvolution,MCKD)应用于滚动轴承早期故障诊断。并针对MCKD参数滤波器长度及移位数需人为选择的问题,提出一种基于参数优化的最大相关峭度解卷积的滚动轴承早期故障诊断方法。首先,针对轴承工作的实际工况讨论了最优移位数;然后,以经解卷积后信号的形态能量熵作为评价函数,利用网格搜索法对滤波器长度进行寻优;最后,利用参数优化后的MCKD算法增强信号中的冲击成分,通过包络谱判断轴承故障类型。实验表明,该方法可有效地增强轴承信号中微弱的故障特征成分,实现滚动轴承早期故障的诊断。In early fault diagnosis of rolling bearings,the fault signal is weak and covered by the environmental noise,and it is difficult to extract the fault features.To solve the problem,the maximum correlated kurtosis deconvolution(MCKD)was introduced to the early fault diagnosis for rolling bearings.In view of the problem that the filter length and the shift number of MCKD need to be selected manually,parametric optimized MCKD was proposed for rolling bearing early fault diagnosis.Firstly,the optimal shift number was analyzed according to the actual working condition of the bearing.Then,the morphological energy entropy of the signal was used as the evaluation function and the filter length was optimized by grid search method.Finally,the impulse components in signal were enhanced with the optimized MCKD,and the fault type of the bearing was judged by the envelope spectrum.The results show that this method can effectively enhance fault characteristics of the bearing signal and realize the early fault diagnosis of rolling bearings.

关 键 词：故障诊断 滚动轴承 最大相关峭度解卷积 网格搜索法 形态能量熵 

分 类 号：TH133.3[机械工程—机械制造及自动化] TH165.3