High-Performance Rul Prediction for Aeroengine Based on Combining Health States and Trajectory Similarity

In this paper, we have proposed a high-precision method on the RUL prediction of aeroengines based on health states and trajectory similarity. We use K-means clustering method to distinguish different health states of aero-engines, and constructs the aero-engine life database under different health states. Effectively reduce the false prediction caused by curve overlap in the single-life model library. With the help of segmented similarity measurement method, the trajectory similarity between the HI curve of the data to be tested and the HI curve of the life library can be better matched. The RUL of the example can be obtained by weighted combination of multiple predictions. Finally, the validity of this method has been verified by the simulation data set of turbofan aeroengines . Compared with other two similar algorithms, the accuracy has been increased by 4% and 6% respectively, in which the lowest penalty score is obtained.