Parameter Prediction for Aeroengine Lubrication System Based on a BiLSTM-LSTM Model with Attention Mechanism

The monitoring and analysis of oil status is an important means of condition monitoring and fault diagnosis for aeroengines. With the development of artificial intelligence technology, deep learning methods have been gradually applied in various fields, including the field of fault diagnosis. To achieve the prediction of fault diagnosis of aeroengine lubrication system, based on the actual airborne flight data, this paper describes a BiLSTM and LSTM network to establish an engine lubrication system model of the flight process, including oil pressure difference, pressure altitude, Mach number, oil level, oil return temperature and other parameters. The attention mechanism is also introduced to improve the prediction accuracy of the model. The calculation results of the model are in good agreement with the flight test results, indicating the feasibility and effectiveness of the model prediction. Compared with other deep learning models, the prediction results of BiLSTM-BiLSTM-LSTM-Attention model are more accurate.