Blind Source Separation and Blind Equalization for Vibration Based Diagnostics of Rotorcraft Bearings

The vibration signal collected from a defective bearing is an aggregate of vibrations from the bearing and numerous nearby sources in the power train. Such mixing of signals coupled with noise makes incipient damage detection nearly impossible. This paper presents a blind source separation (BSS) and blind equalization (BE) techniques to demix the acquired vibration signal into correctly identified source signals. Some widely used blind separation algorithms have been compared to evaluate their performance in the separation of the faulty bearing signal from other signals. Experimental data from a seeded fault test of AH-64 hanger bearings were used to verify the effectiveness of the proposed algorithms.