Novel SABO-optimized LSTM and BiLSTM models for enhancing aircraft turbine engine emissions prediction

This paper proposes a turbine engine emission prediction method based on improved long short-term memory (LSTM) algorithms using subtraction average-based optimizer (SABO), specifically aiming at high-precision modeling of the cooling bleed air flow from the high pressure turbine (HPT) and low pressure turbine (LPT). Firstly, from the 21 sensor detection data in the C-MAPSS dataset, variables with high correlation to cooling bleed air are selected to construct the objective parameters of the model. Secondly, to enhance the prediction accuracy of the LSTM and bidirectional LSTM (BiLSTM) networks, the particle swarm optimization (PSO) and SABO are employed to optimize the parameters of the two memory networks. This leads to the proposal of cooling bleed air emission prediction models based on the frameworks of PSO-LSTM, PSO-BiLSTM, SABO-LSTM and SABO-BiLSTM. Through the training and testing of 33,991 data sets of cooling bleed air emission data measured from HPT and LPT, the results demonstrate that the SABO-BiLSTM model exhibits the best emission prediction accuracy and fitting performance, which provides technical support for assessing the environmental performance of aircraft emissions and formulating effective emission reduction strategies.