A novel remaining useful life prediction combining IWOA-BiLSTM and PCA for aircraft engine

Remaining useful life (RUL) Prediction is crucial for the predictive maintenance of industrial manufacturing systems. The challenges encountered in the prediction of RUL include identifying a pertinent health indicator (HI), effectively modeling t he degradation pattern across diverse features, and selecting optimal methodologies to enhance prediction accuracy. This paper establishes the construction of the HI using PCA, utilizing data from aircraft engines. Bi-LSTM with optimization algorithm is investigated for RUL evaluation and prediction of aircraft engines. Subsequent experiments are performed on simulation data from a commercial modular aero propulsion system using the proposed method, demonstrating its superiority over traditional approaches.