Prognostics of Remaining Useful Life for Aviation Structures Considering Imperfect Repairs

Abstract

Maintenance plays an important role in fulfilling the goals of
the Prognostics and Health Management (PHM) field. As of
now, no publication has addressed the impact of imperfect
repair actions from the prognostics perspective. Imperfect
repairs introduce complexities, altering system degradation
processes and increasing prediction uncertainties, thereby impacting
the accuracy of Remaining Useful Life (RUL) predictions.
To fill this gap in the literature, the study proposes developing
a robust prognostic model adaptable to post-repair
operations. The prognostic model that will be developed is
stochastic since stochastic models have already proven their
adaptability to unseen test data. However, further development
of such models is needed to deal with data on repaired
systems. In addition to that, the implementation of a Bayesian
Extension allows uncertainty interpretability to be considered
to account for the uncertainty coming from the repair action
itself but also from the different sources of uncertainties that
have not been studied in the field of prognostics.