Detection of Bearing Faults Early with Zero-Frequency Resonators and Wavelet Transform

Rotating equipment, including generators, motors, pumps, and turbines, frequently employ rolling element bearings. Primary reason for failure in rotating machinery is bearing failures. Early bearing fault diagnosis is essential for avoiding machinery breakdown. In general, condition-based maintenance is used to prevent bearing failure. The vibration signal of a rolling element bearing with a localized fault shows periodic impulses. The bearing fault vibration signal is feeble at the early stages of the bearing fault. Present work uses zero-frequency filter and wavelet transform-based algorithm for early detection of the bearing faults. The present work aims to check the noise robustness and reduction in aliasing using various mother wavelets vibrations caused by bearing faults. Using a simulated noisy signal containing periodic impulses, the algorithm is explained. Results from various mother wavelets, such as the Symlet, Coiflet, and Daubechies, are compared using the de-noising algorithm with two levels.