Mobile electronic nose based on carbon nanotube-SnO2 gas sensors: Feature extraction techniques and its application

Abstract

In this paper, a mobile electronic nose (E-nose) based on novel hybrid carbon nanotube-SnO2 gas sensors is described. The instrument combines new feature extraction techniques including integral and primary derivative, which leads to higher classification performance comparing with the classical features (Δ R and Δ R/R0). The results show that doping of carbon nanotube (CNT) improves the sensitivity of hybrid gas sensors while quantity of CNT has a direct effect on selectivity to volatile organic compounds, i.e. MeOH and EtOH. The realworld applications of this E-nose were also presented. Based on the proposed methods, this instrument can monitor and classify 1% vol. of MeOH contamination in whiskey. ©2009 IEEE.