Bearing Fault Detection Using Adaptive Noise Cancelling

Abstract

The ability to diagnose a mechanical fault is enhanced if the monitoring signal can be preprocessed to reduce the effect of unwanted noise. To this end, the adaptive noise cancelling technique (ANC) can substantially improve the signal to noise ratio where the required signal is contaminated by noise. ANC makes use of two inputs—a primary input which contains the corrupted signal, and a reference input containing noise correlated in some unknown way with the primary noise. A variation of ANC is also proposed and it is shown that this can be applied effectively in those situations where inputs contain correlated signals but uncorrelated or weakly correlated noises. Using vibrational data derived from a reasonably complex bearing rig and preprocessing the data by the ANC technique, this paper shows that the statistical and spectral analysis techniques can be made more effective in their diagnostic roles after the application of ANC.