Improved stator current spectral analysis technique for bearing fault diagnosis

Although detection of bearing faults from vibration signal has been extensively used, detection using electrical signal such as the stator current has also proven to be interesting and relevant, but improvements of the stator current spectrum are necessary so as to assist in a more readable signal. In this paper, two improvements have been proposed and discussed. Firstly, the proposition of a designed and realized anti-aliasing filter to be inserted in the acquisition chain so as to improve the processed signals quality. Secondly, the proposition of a developed algorithm to enable the maxima localization in a given frequency band for a more readable stator current spectrum. The experimental results obtained, show the merit and effectiveness of the proposed contribution.