A Bidirectional Long Short-Term Memory Autoencoder Transformer for Remaining Useful Life Estimation

Estimating the remaining useful life (RUL) of aircraft engines holds a pivotal role in enhancing safety, optimizing operations, and promoting sustainability, thus being a crucial component of modern aviation management. Precise RUL predictions offer valuable insights into an engine’s condition, enab...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Mathematics (Basel) Ročník 11; číslo 24; s. 4972
Hlavní autori: Fan, Zhengyang, Li, Wanru, Chang, Kuo-Chu
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Basel MDPI AG 01.12.2023
Predmet:
Adaptation
Aircraft
Aircraft engines
Analysis
Artificial intelligence
autoencoder
bidirectional LSTM
Coders
Computational linguistics
Deep learning
Electric transformers
Engines
Estimation
Feature extraction
Language processing
Measuring instruments
Methods
Natural language interfaces
Neural networks
Preventive maintenance
remaining useful life prediction
self-supervised learning
Sensors
Support vector machines
Time series
Training
Transformer
turbofan engine
Useful life
ISSN:2227-7390, 2227-7390
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Estimating the remaining useful life (RUL) of aircraft engines holds a pivotal role in enhancing safety, optimizing operations, and promoting sustainability, thus being a crucial component of modern aviation management. Precise RUL predictions offer valuable insights into an engine’s condition, enabling informed decisions regarding maintenance and crew scheduling. In this context, we propose a novel RUL prediction approach in this paper, harnessing the power of bi-directional LSTM and Transformer architectures, known for their success in sequence modeling, such as natural languages. We adopt the encoder part of the full Transformer as the backbone of our framework, integrating it with a self-supervised denoising autoencoder that utilizes bidirectional LSTM for improved feature extraction. Within our framework, a sequence of multivariate time-series sensor measurements serves as the input, initially processed by the bidirectional LSTM autoencoder to extract essential features. Subsequently, these feature values are fed into our Transformer encoder backbone for RUL prediction. Notably, our approach simultaneously trains the autoencoder and Transformer encoder, different from the naive sequential training method. Through a series of numerical experiments carried out on the C-MAPSS datasets, we demonstrate that the efficacy of our proposed models either surpasses or stands on par with that of other existing methods.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2227-7390
2227-7390
DOI:10.3390/math11244972