A study of the fault prediction problem of permanent magnet faulttolerant motor equipped in the aircraft actuation system is advantageous to monitoring aircraft health status accurately, and providing decision support for aircraft maintenance. For a permanent magnet faulttolerant motor system in case of a hit from the outside, the traditional grey model is not high prediction accuracy. A new improved forecast method based on grey theory is proposed in the paper. The permanent magnet faulttolerant motor is a complex system belonging to high order, nonlinear and strong coupling. The paper adopts a strengthening buffer operator to deal with data sequence, restore data, eliminate the interference of shock, and forecast the failure energy data sequence through the establishment of the basic model. The results show that the original sequence is processed by strengthening buffer operator, the prediction accuracy is increased to  96.3% above, the relative prediction error is reduced by 43.06% on average compared to the primitive sequence basic model, and the fault prediction of the behavior is improved effectively.