the 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 KICA, which is based on nonlinear function space, can solve this problem effectively. Compared to current ICA algorithms, the KICA is notable for its flexibility and robustness. We presented the principle of KICA, and illustrated it with simulations. Finally, through the analysis of transient acoustic signal on gearbox combined with order envelope spectrum analysis, we found the fault characters, thus showed its effectiveness.