Deep Learning‐Based Prediction of Remaining Useful Life of Aircraft Engines Using Non‐Linear Labels
Electronics Letters · 2026-01 · 62 (1)
DOI: 10.1049/ell2.70613 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics Letters, online erschienen am 02.06.2026. Verfasst von Min‐Jung Kim und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 12 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
Accurate prediction of the remaining useful life (RUL) of aircraft turbofan engines is critical for prognostics and health management. Recent deep learning‐based research studies have largely focused on improving prediction accuracy by designing more sophisticated model architectures. In this work, we take a complementary approach that targets the RUL label representation so that it can reflect the non‐linear characteristics of engine degradation. We maintain the existing pre‐training architecture but replace the conventional linear or piece‐wise linear RUL label with a concave function that better captures the actual degradation pattern. Evaluations on the NASA C‐MAPSS dataset show that the proposed non‐linear label improves the root mean square error and the NASA S‐Score compared to the state‐of‐the‐art piecewise‐linear baseline.
Zitieren
APA
Kim, M., & Lee, K. (2026). Deep Learning‐Based Prediction of Remaining Useful Life of Aircraft Engines Using Non‐Linear Labels. Electronics Letters, 62(1). https://doi.org/10.1049/ell2.70613
Vancouver
Kim M, Lee K. Deep Learning‐Based Prediction of Remaining Useful Life of Aircraft Engines Using Non‐Linear Labels. Electronics Letters. 2026;62(1). doi:10.1049/ell2.70613
BibTeX
@article{Kim2026,
author = {Kim, Min‐Jung and Lee, Kang‐Won},
title = {Deep Learning‐Based Prediction of Remaining Useful Life of Aircraft Engines Using Non‐Linear Labels},
journal = {Electronics Letters},
year = {2026},
volume = {62},
number = {1},
doi = {10.1049/ell2.70613}
}
EUVIMED