摘要
针对滚动轴承振动信号的非平稳特性和在现实条件下难以获取大量故障样本的实际情况,提出一种经验模态分解、复杂度测量分析和支持向量机相结合的故障诊断方法。运用经验模态分解方法对其去噪信号进行分析,利用互相关系数准则对固有模式分量进行筛选,再计算所选分量的复杂度以组成故障特征向量,并将其作为支持向量机的输入以识别滚动轴承的故障类型。最后,利用实际滚动轴承试验数据的诊断与对比试验验证了该方法的有效性和泛化能力。
According to the non-stationarity characteristic of the vibration signals from rolling bearing and the situation is hard to obtain enough fault samples,a comprehensive fault diagnosis method based on Empirical Mode Decomposition（EMD）,complexity measure analysis and Support Vector Machine（SVM） is proposed.The denoised vibration signal is analyzed by using the method of EMD decomposing,and the Intrinsic Mode Functions（IMF） components are chose by using the criteria of mutual correlation coefficient between IMF components and denoised signal.The complexity of each IMF component is calculated as faulty eigenvector and served as input of SVM to recognize the fault type of rolling bearing.Practical rolling bearing experimental data diagnosis and contrast test are used to verify the effectiveness and generalization ability of this method.
出处
《机械传动》
CSCD
北大核心
2011年第2期20-23,31,共5页
Journal of Mechanical Transmission
关键词
滚动轴承
故障诊断
经验模态分解
复杂度
支持向量机
Rolling bearing Fault diagnosis Empirical mode decomposition（EMD） Complexity Support vector machine（SVM）
