MSCNN-BLSTM based Prediction of the Remaining Useful Life of Aeroengine

Abstract

Accurate prediction of the aeroengine remaining useful life (RUL) is essential to improve engine availability and reliability. Aiming at the reliable prediction of residual life of aeroengine system, an engine residual life prediction model based on the fusion of multiscale fusion two-dimensional convolutional neural network and bidirectional long and short term memory (MSCNN-BLSTM) is proposed. Based on the fusion of two-dimensional convolutional neural network and bidirectional long and short time memory (BLSTM) network, the engine medium and advanced features extracted by the convolutional neural network are integrated to make residual life prediction. Finally, C-MAPSS dataset provided by NASA was used for validation. It is shown that the proposed multiscale hybrid model, compared with other model predictions, reduces the performance index score and root mean square error by 32.2% and 14.7% respectively. It can be seen that the data-driven model can effectively extract the information from the degradation data, which improves the prediction performance of aeroengine remaining life.