Fault diagnosis of high-speed rolling element bearings using wavelet packet transform

The time-frequency analysis techniques like Continuous Wavelet Transform (CWT), Discrete Wavelet Transform (DWT) and wavelet packet analysis have been compared to detect and diagnose faults in rotor bearing system. Discrete Wavelet Transform (DWT) provides flexible time frequency resolution which suffers from a relatively low resolution in the high-frequency region. This deficiency leads to difficulty in differentiating high-frequency transient components. WPT based signal decomposition process up to n-level produces a total of 2(n) sub-bands, with each sub-band covering 1/2(n) of the signal frequency spectrum. WPT based global threshold criterion is applying before denoising of detail information. This denoised signal is then auto correlate with original signal and energy spectrum is generated for diagnosis of bearing fault. The enhanced signal decomposition capability makes WPT an attractive tool for detecting and differentiating transient elements with high-frequency characteristics and helping in the minimisation of interventions by the end user.