Improving signal-to-noise ratio (SNR) for inchoate fault detection based on principal component analysis (PCA)

Detection of inchoate fault demands high level of fault classification accuracy under poor signal-to-noise ratio (SNR) which appears in most industrial environment. Vibration signal analysis methods are widely used for bearing fault detection. In order to guarantee improved performance under poor SNR, feature extraction based on statistical parameters which are free from Gaussian noise become inevitable. This paper proposes a feature extraction framework based on principal component analysis (PCA) for improving SNR. Features extracted based on PCA have the tendency to alleviate the impact of non-Gaussian noise. PCA algorithm provides useful time domains analysis for no-stationary signals such as vibration in which spectral contents vary with respect to time. Experimental studies on vibration caused by ball bearing faults show that the proposed algorithm demonstrates the improvements in term of classification accuracy under poor signal-to-noise ratio (SNR).