基于SVM和KPCA的滚动轴承故障诊断

Fault Diagnosis of Rolling Bearing Based on SVM and KPCA

从时域角度分析了滚动轴承的振动信号,综合利用SVM和KPCA方法来实现对滚动轴承的故障诊断研究。首先对滚动轴承的原始信号从时域角度分析提取典型特征,再利用KPCA方法对输入的典型特征降维,最后采用SVM算法对降维后的数据进行故障诊断。实验证明:该方法在保证较高的故障识别能力的前提下,不仅能够有效地提取损伤特征、降低数据维数,而且实现了数据可视化。

The vibration signals of rolling bearings were analyzed from the perspective of time domain,and having SVM and KPCA methods comprehensively applied to discuss the fault diagnosis of rolling bearings was implemented,including firstly having the original signal of the rolling bearings extracted from the time domain to extract typical features,then having the KPCA method adopted to reduce dimensions of the input and finally having the SVM algorithm employed to diagnose the mapped data.The experimental results demonstrate that,under the premise of ensuring a high fault recognition ability,the proposed method can effectively extract fault features and reduce data dimension and realize data visualization.