Fault Diagnosis of Gearbox Based on Principal Component Analysis and Sequential Probability Ratio Test

This paper1 presents a novel method for diagnosing gear crack in gearbox based on principal component analysis(PCA) and sequential probability ratio test(SPRT). The vibration signal collected in the gearbox experimental system is denoised. The wavelet packet transform is suitable for the noise reduction to extract the features for the identification of the faulty gears. PCA is used to extract useful parameters of vibration signals. The parameter with the largest contribution rate after dimensionality reduction is chosen as the testing parameter of SPRT. The results indicate that this method is feasible and practical to classify different conditions of the gearbox.