The daily maintenance work to the aircraft engine takes a lot of human, financial and material resources, once the engine failure will cause a major flight accidents. Therefore, quick and effective prediction to the aircraft engine fault can guarantee the engine reliable work and the plane flight safety. In this paper, the Fast ICA blind source separation algorithm was applied to some type of double rotor aircraft engine’s vibration signal, and the simulation results shown that the proposed Fast ICA algorithm has got a good separation effect, effectively isolated aircraft engine vibration signal containing noises. According to the fault type characters of the aircraft engine, Fast ICA blind source separation algorithm identified the fault type of the aircraft engine, which has fewer iteration times, low computing complexity, the separation effect and stability is good.