EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Low-Cost CO2 NDIR Sensors: Performance Evaluation and Calibration Using Machine Learning Techniques

Ravish Dubey, Arina Telles iD, James Nikkel iD, Chang Cao iD, Jonathan Gewirtzman iD, Peter A Raymond, Xuhui Lee

Sensors · 31.08.2024 · 24 (17) · 5675

DOI: 10.3390/s24175675 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 31.08.2024. Verfasst von Ravish Dubey und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of the Environment, Yale University, New Haven, CT 06511, USA. Die Arbeit wurde bisher 47-mal zitiert (Crossref) und nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: This research was funded by the National Natural Science Foundation of China, XL by Robert Wood Johnson Foundation, XL, PR and JN by Yale Planetary Solutions a…, National Natural Science Foundation of China, Robert Wood Johnson Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The study comprehensively evaluates low-cost CO2 sensors from different price tiers, assessing their performance against a reference-grade instrument and exploring the possibility of calibration using different machine learning techniques. Three sensors (Sunrise AB by Senseair, K30 CO2 by Senseair, and GMP 343 by Vaisala) were tested alongside a reference instrument (Los Gatos precision greenhouse gas analyzer). The results revealed differences in sensor performance, with the higher cost Vaisala sensors exhibiting superior accuracy. Despite its lower price, the Sunrise sensors still demonstrated reasonable accuracy. Meanwhile, the K30 sensor measurements displayed higher variability and noise. Machine learning models, including linear regression, gradient boosting regression, and random forest regression, were employed for sensor calibration. In general, linear regression models performed best for extrapolating data, whereas decision tree-based models were generally more useful in handling non-linear datasets. Notably, a stack ensemble model combining these techniques outperformed the individual models and significantly improved sensor accuracy by approximately 65%. Overall, this study contributes to filling the gap in intercomparing CO2 sensors across different price categories and underscores the potential of machine learning for enhancing sensor accuracy, particularly in low-cost sensor applications.

Schlagwörter

Stichwörter der Autor:innen

collocated measurementslow-cost CO2 sensorsmachine learning calibrationperformance evaluation

Autor:innen und Institutionen

  • School of the Environment, Yale University, New Haven, CT 06511, USA.
  • Department of Physics, Yale University, New Haven, CT 06511, USA.
  • School of Applied Meteorology, Nanjing University of Information Science & Technology (NUIST), Nanjing 210044, Jiangsu, China.

ORCID-Profile: Arina Telles, James Nikkel, Chang Cao, Jonathan Gewirtzman

Förderung

  • This research was funded by the National Natural Science Foundation of China, XL by Robert Wood Johnson Foundation, XL, PR and JN by Yale Planetary Solutions a…
  • National Natural Science Foundation of China
  • Robert Wood Johnson Foundation
  • Yale Planetary Solutions
  • Three Cairns Climate Impact Innovation Fund

Zitieren

APA

Dubey, R., Telles, A., Nikkel, J., Cao, C., Gewirtzman, J., Raymond, P. A., & Lee, X. (2024). Low-Cost CO2 NDIR Sensors: Performance Evaluation and Calibration Using Machine Learning Techniques. Sensors, 24(17), 5675. https://doi.org/10.3390/s24175675

Vancouver

Dubey R, Telles A, Nikkel J, Cao C, Gewirtzman J, Raymond PA, et al. Low-Cost CO2 NDIR Sensors: Performance Evaluation and Calibration Using Machine Learning Techniques. Sensors. 2024;24(17):5675. doi:10.3390/s24175675

BibTeX
@article{Dubey2024,
  author = {Dubey, Ravish and Telles, Arina and Nikkel, James and Cao, Chang and Gewirtzman, Jonathan and Raymond, Peter A and Lee, Xuhui},
  title = {Low-Cost CO2 NDIR Sensors: Performance Evaluation and Calibration Using Machine Learning Techniques},
  journal = {Sensors},
  year = {2024},
  volume = {24},
  number = {17},
  pages = {5675},
  doi = {10.3390/s24175675}
}
Kostenlos starten →