Wavelet domain principal feature analysis for spindle health diagnosis

This article introduces a hybrid signal processing technique for spindle health monitoring and diagnosis, through the integration of wavelet packet transform and principal feature analysis. Vibration signals measured from a spindle test system with different defect conditions are first decomposed into multiple sub-frequency bands by means of the wavelet packet transform. Statistical parameters such as energy and Kurtosis of these sub-frequency bands are then calculated. Subsequently, Principal Feature Analysis, which is an extension of the Principal Component Analysis, is performed on the statistical parameters to aid in the selection of the most representative features, which can be distinctively separated from each other, as inputs to a diagnostic classifier. Experimental analysis of sensor data measured from the spindle test system has verified the effectiveness of the developed technique.