ABSTRACT
Vibration-based condition monitoring is useful for fault detection in rotating machines. The time and frequency domain feature unification with signals from multiple sensors for fault diagnosis to achieve a single analysis have been well established in earlier studies. An initial study developed data fusion of acceleration and velocity features (dFAVF) with a useful machine condition diagnosis. However, these studies focused on rotor faults only. This study tends to incorporate bearing data in the developed dFAVF model. The aim is to diagnose an extensive range of machines’ faults in a single analysis. Signals observed were from a test rig operating with multiple speeds below and above its first critical speed. The preliminary result showed useful identification of the machines’ faults. The usefulness of this approach in industrial-scale fault diagnosis may improve the simplicity of understanding the machine’s overall behaviour with regularly measured vibration data.
