基于核主成分分析和朴素贝叶斯的滚动轴承故障诊断

Fault Diagnosis of Rolling Bearing Based on Kernel Principal Component Analysis and Naive Bayes

针对滚动轴承故障难以准确诊断的问题,提出一种基于主成分分析和朴素贝叶斯算法的滚动轴承故障诊断方法。首先采用经验模态分解方法对滚动轴承的振动信号进行分解,得到多个不同特征时间尺度的本征模态函数分量,并对各个本征模态函数分量计算样本熵,构造成特征向量集;然后利用核主成分分析进行降维;最后利用朴素贝叶斯算法进行故障诊断。实验结果表明,所提出的方法准确率达到95%,基本满足一般的轴承故障诊断要求。

Aiming at the problem that the fault of rolling bearing is difficult to diagnose accurately,proposes a fault diagnosis method of rolling bear.ing based on principal component analysis and naive Bayesian algorithm.Firstly,the vibration signal of rolling bearing is decomposed by empirical mode decomposition method,and the intrinsic mode function components with different characteristic time scales are obtained,and the sample entropy of each intrinsic mode function component is calculated,which is constructed into eigenvector set.Secondly,the di.mensionality is reduced by using the kernel principal component analysis.Finally,the fault diagnosis is carried out by using the naive Bayesian algorithm.The experimental results show that the accuracy of the proposed method reaches 95%,which basically meets the gener.al requirements of bearing fault diagnosis.