Hybrid Deep Learning Model-based For Remaining Useful Life Estimation.

Abstract

The precise estimation of Remaining Useful Life (RUL) holds crucial importance in mitigating the risks of machine failures, specifically in prognosis and condition-based maintenance (CBM). Through continuous monitoring of machine conditions and accurate RUL prediction, preemptive measures can be implemented to avert unforeseen breakdowns. Over the past years, deep learning models have demonstrated remarkable effectiveness in accurately predicting RUL across diverse applications. These techniques are able to effectively extract meaningful patterns and features from complex data. The primary objective of this Master’s thesis is to propose a hybrid deeplearning model that leverages advanced deep learning techniques such as multi-head architecture and attention mechanism, to offer an accurate prognosis of the Remaining Useful Life (RUL) for airplane turbofan engines.