Detection of the Rolling Element Bearing Faults using Optimized--Wavelet De-Noising Technique.

Rolling bearing fault detection approach using wavelet de-noising technique based on optimized wavelet-base function derived from the impulse response of the bearing system is proposed. The wavelet parameters are optimized based on maximization kurtosis criteria to ensure large similarity with the impulse generated by incipient bearing fault. The results show the effectiveness of the proposed wavelet de-noising to reveal the impulses produced by faulty bearing and estimate its period.