EUVIMEDEUVIMED entdecken
Crossref

Design, implementation, and validation of a portable electronic nose based on embedded system and deep learning model

Feiyang Jiang, Rongjie Lai, Haigen Wu, Yongli Zhao

Measurement Science and Technology · 17.04.2026 · 37 (16) · 165101

DOI: 10.1088/1361-6501/ae5df4 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 20.04.2026. Verfasst von Feiyang Jiang und 3 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This work presents a measurement-oriented portable electronic nose (E-nose) system framework, addressing the key challenges brought by measurement uncertainties in high-similarity gas sensing scenarios under small sample size conditions. It does not simply treat the gas identification as a classification problem, but rather as the construction of a structured measurement space, and modeling the measurement process as an end-to-end chain that includes gas sampling, signal acquisition, representation mapping, and gas identification. Within this framework, we propose a gas feature extraction method based on metric learning to explicitly construct the geometric structure of the gas feature space. By introducing the triplet loss algorithm, the learned gas features are constrained to form a structured representation with intra-class compactness and inter-class separability. This structured representation provides a strong guarantee for achieving consistency and identification accuracy of gas measurements under different conditions. To verify the effectiveness of this method, this paper designs an E-nose with a heterogeneous dual-core embedded architecture for real-time gas identification measurement and on-device inference. Experimental results on gas samples with highly similar volatile components showed that the proposed method can effectively improve the spatial structure quality of gas features and achieve high identification accuracy. The results of the long-term stability evaluation experiments further indicated that it can effectively overcome the drift of gas sensors and the anti-interference capability, enhancing the robustness of gas measurements.

Zitieren

APA

Jiang, F., Lai, R., Wu, H., & Zhao, Y. (2026). Design, implementation, and validation of a portable electronic nose based on embedded system and deep learning model. Measurement Science and Technology, 37(16), 165101. https://doi.org/10.1088/1361-6501/ae5df4

Vancouver

Jiang F, Lai R, Wu H, Zhao Y. Design, implementation, and validation of a portable electronic nose based on embedded system and deep learning model. Measurement Science and Technology. 2026;37(16):165101. doi:10.1088/1361-6501/ae5df4

BibTeX
@article{Jiang2026,
  author = {Jiang, Feiyang and Lai, Rongjie and Wu, Haigen and Zhao, Yongli},
  title = {Design, implementation, and validation of a portable electronic nose based on embedded system and deep learning model},
  journal = {Measurement Science and Technology},
  year = {2026},
  volume = {37},
  number = {16},
  pages = {165101},
  doi = {10.1088/1361-6501/ae5df4}
}
Kostenlos starten →