基于小波函数的核主元特征约简方法研究

Feature reduction method based on wavelet kernel-PCA

提出一种基于小波核函数的核主元特征约简方法。核函数是核主元分析的关键,将Mexican hat小波函数引入核主元分析中,以增强核主元分析的非线性映射能力。用转子在正常、油膜涡动、不平衡和径向碰摩状态下的实验数据对该方法进行了检验,比较了主元分析、核主元分析与小波核主元分析的效果。结果表明,小波核主元分析方法能有效地区分转子故障模式,更适合于故障诊断中的非线性特征约简。

A feature reduction method based on wavelet kernel-PCA （WKPCA） is presented. Mexican hat wavelet kernel is introduced to enhance nonlinear mapping capability of kernel-PCA. The experimental data sets of rotor operating under four conditions： normal, oil whirling, rub and unbalance are used to test the WKPCA method. The feature reduction results of WKPCA are compared with that of PCA and KPCA method. The results indicate that WKPCA can classify the rotor fault modes efficiently. The WKPCA is more suitable for nonlinear feature reduction in the field of fault diagnosis.