Classification and concentration estimation of CO and NO2 mixtures under humidity using neural network-assisted pattern recognition analysis

Jin-Young Kim; Somalapura Prakasha Bharath; Ali Mirzaei; Hyoun Woo Kim; Sang Sub Kim

doi:10.1016/j.jhazmat.2023.132153

Classification and concentration estimation of CO and NO₂ mixtures under humidity using neural network-assisted pattern recognition analysis

J Hazard Mater. 2023 Oct 5:459:132153. doi: 10.1016/j.jhazmat.2023.132153. Epub 2023 Jul 25.

Authors

Jin-Young Kim¹, Somalapura Prakasha Bharath², Ali Mirzaei³, Hyoun Woo Kim⁴, Sang Sub Kim⁵

Affiliations

¹ Department of Materials Science and Engineering, Inha University, Incheon 22212, Republic of Korea.
² Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Republic of Korea.
³ Department of Materials Science and Engineering, Shiraz University of Technology, Shiraz, Iran.
⁴ Division of Materials Science and Engineering, Hanyang University, Seoul 04763, Republic of Korea. Electronic address: hyounwoo@hanyang.ac.kr.
⁵ Department of Materials Science and Engineering, Inha University, Incheon 22212, Republic of Korea. Electronic address: sangsub@inha.ac.kr.

PMID: 37506649
DOI: 10.1016/j.jhazmat.2023.132153

Abstract

This study addresses the concerns regarding the cross-sensitivity of metal oxide sensors by building an array of sensors and subsequently utilizing machine earning techniques to analyze the data from the sensor arrays. Sensors were built using In₂O_3, Au-ZnO, Au-SnO₂, and Pt-SnO₂ and they were operated simultaneously in the presence of 25 different concentrations of nitrogen dioxide (NO₂), carbon monoxide (CO), and their mixtures. To investigate the effects of humidity, experiments were conducted to detect 13 distinct CO and NO₂ gas combinations in atmospheres with 40% and 90% relative humidity. Principal component analysis was performed for the normalized resistance variation collected for a particular gas atmosphere over a certain period, and the results were used to train deep neural network-based models. The dynamic curves produced by the sensor array were treated as pixelated images and a convolutional neural network was adopted for classification. An accuracy of 100% was achieved using both models during cross-validation and testing. The results indicate that this novel approach can eliminate the time-consuming feature extraction process.

Keywords: CO; Gas sensors; NO(2); Neural networks; Noble metal; Pattern recognition.