Application of independent component analysis for gear diagnosis

In this paper, an independent component analysis (ICA)-based vibration signal processing is proposed for the detection and identification of machinery malfunctions/faults, especially for the gear fault diagnosis. Synchronous Signal Averaging (SSA) technique is used to enhance the measured vibration signal. The response of gear due to some internal and external excitations can be modelled as the superposition of gear meshing components and the non-harmonic additive noise. The ICA-based signal processing can accurately estimate the gear meshing components and simultaneously reduce the effect of noise. The time-frequency analysis is applied to the estimated gear meshing components. The health condition of the gear is diagnosed by characterizing the time-frequency transforms. The proposed technique is supported by the numerical simulation.