Abstract: Performance degradation is one of the main reasons leading to the aeroengineremoval. Aiming atthe huge difference between individual performance degradation modes of the same type of aeroengine, the engine residual life prediction method based on similarity matching is proposed. Firstly, EGTM is selected to characterize the current performance state of the engine, similar correction and Dixon outliers are used to eliminate the influence of external factors. Secondly, K-means is conducted to the EGTM data. Finally,Euclidean distance method is used to calculate the similarity degree between predicted and historical samples,giving corresponding weight of the reference samples to predict the residual life of the engine. The actual data verification of airlines, shows that the proposed method has high prediction accuracy and can provide reliable basis for maintenance plan.

Key words: aeroengine, similar correction, EGTM, time series, cluster analysis, similarity-based residual life prediction