Remaining Useful Life Estimation of Aircraft Engines Using Siamese Attention-Augmented Quantum Convolutional Neural Networks

Remaining Useful Life (RUL) prediction is vital in the prognostics and health management (PHM) systems for maintaining complex machinery. However, extracting meaningful features from multivariate time-series machinery data is highly challenging due to the data correlations and temporal information. Therefore, this paper develops a competent RUL prediction method using Siamese Attention-Augmented Quantum Convolutional Neural Networks (SAAQCNN). The SAAQCNN-based model is a hybrid classifier developed by integrating the Siamese Neural Networks (SNN) and Convolutional Neural Networks (CNN) with attention-augmented convolution and hyper-parameter optimization by Quantum theory enhanced Equilibrium Optimizer (QEO). It can learn the spatial and temporal features from the multivariate time-series monitoring dataset and provides the optimal features for the prediction process with reduced model complexity. The NASA Turbofan Jet Engine Dataset is utilized for the evaluation experiments. The results showed that the proposed SAAQCNN-based prediction model achieved better RUL forecasting of the aircraft engines with high accuracy and reduced processing time and RMSE values.