A novel method for feature extraction of rotating machinery vibration signals

A novel method named correlation of time series and squaring for suppressing the noise in the feature extraction was developed. It eliminates the impact of zero-average noise and non-stationary variance intrusion colored noise in sampling series, gets over the difficulty of frequency identification because of strong background noise in the subsequence analysis, protrudes the feature components of original signals, and provides a wonderful former data for spectrum analysis using the multi-correlation of time series. The results of simulation analysis to extract the feature of some engine´s rotating shaft vibration signals validate that this method can be used to extract the features of vibration signals of rotating machinery. They also prove this method has bright future in the engineering applications.