Fault Diagnosis of Gearbox Gear Wearing Based on Kernel ICA and Transient Acoustic Signal

When we use the acoustic signal to diagnose gearbox faults, it will be affected by the nonlinear factors and other noise, and thus reduce the accuracy of diagnosis. ICA is a kind of signal processing method based on high order statistic, it can recover the source signals from the linearly mixed signals, but there are disadvantages in processing non-linearity signals, and the Kernel ICA, which is based on nonlinear function space, can solve this problem effectively. Compared to current ICA algorithms, the Kernel ICA is notable for its flexibility and robustness. The paper presented the principle and algorithm steps of Kernel ICA, through the analysis of transient acoustic signal on gearbox combined with order cepstrum analysis, we found the fault characters, and distinguished the gear wearing fault of gearbox successfully, thus showed its feasibility and validity.