Distinguishing methane from other hydrocarbons using machine learning and atmospheric pressure plasma optical emission spectroscopy

Mansouri, Tahereh Shah and Wang, Hui and Mariotti, Davide and Maguire, Paul (2024) Distinguishing methane from other hydrocarbons using machine learning and atmospheric pressure plasma optical emission spectroscopy. Journal of Physics D: Applied Physics, 57 (34). 345202. ISSN 1361-6463 (https://doi.org/10.1088/1361-6463/ad4f97)

[thumbnail of Mansouri-etal-JPD-2024-Distinguishing-methane-from-other-hydrocarbons-using-machine-learning-and]
Preview
 Text. Filename: Mansouri-etal-JPD-2024-Distinguishing-methane-from-other-hydrocarbons-using-machine-learning-and.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (2MB)| Preview

Abstract

The ability to detect gas molecule and assign a concentration offers an inventive solution in the field of plasma integrated with machine learning. The most important finding of this work is firstly, to develop an algorithm for gas-molecule identification using three different hydrocarbons (CH4, C2H2, C2H6) and secondly, organize a model for detecting gas concentration (classification). For this reason, initially eight different gases evaluated. The study confirms the present of the unique emission lines as a gas indicator, i.e., a wavelength peak related to hydrocarbons identified via increasing in Cx Hy concentration. By means of unique variable important in projection, hydrocarbons can be distinguished. Our proposed Chemometric analysis strategy examined on >1000 samples and results development of suitable techniques that are sufficiently rapid, accurate and innovative. This demonstrates the potential for real-time, portable, and continuous monitoring of trace gases with potential applications in medical, environmental, and industrial gas sensing.

ORCID iDs

Mansouri, Tahereh Shah, Wang, Hui, Mariotti, Davide ORCID logoORCID: https://orcid.org/0000-0003-1504-4383 and Maguire, Paul;
CORE (COnnecting REpositories)