BROWSE
Detail
Conference Paper
SpatioTemporal Transformer-based Regressive Domain Adaptation for Remaining Useful Life Prediction
Cited 0 time in 
Share 
Share
- Authors
- HyunYong Lee, Nac-Woo Kim, Jungi Lee, Seok-Kap Ko
- Issue Date
- 2023-10
- Citation
- International Conference on Information and Communication Technology Convergence (ICTC) 2023, pp.1662-1664
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- In realizing the accurate remaining useful life (RUL) that is quite important in many industrial areas, the data-driven domain adaptation (i.e., the regressive domain adaptation) has been widely used. In designing the effective regressive domain adaptation model, there are two main issues: model architecture and loss functions. In this paper, we first propose the spatiotemporal transformer-based model to effectively extract features. The proposed transformer considers the spatial relationships across multiple sensors and the temporal relationships of each sensor. We also discuss the usefulness of three loss functions in terms of domain adaptation.
- KSP Keywords
- Data-Driven, Model architecture, Spatial relationships, adaptation model, domain adaptation, extract features, loss function, multiple sensors, remaining useful life prediction, transformer-based