Discrimination of volatile organic compounds emitted by building materials using an electronic nose
Résumé
This work concerns the development of a real time air quality monitoring tool using an electronic nose based on conducting polymer sensors and artificial neural network pattern recognition technique. Eight aromatic Volatile Organic Compounds (VOCs) frequently emitted by the building materials were chosen to assess electronic nose discrimination capability. This discrimination was based on three criteria: carbon chain length (toluene, ethylbenzene and propylbenzene), substituent position on the cycle (o-xylene, m-xylene and p-xylene), and insaturation level of the substituent (ethylbenzene, styrene, phenylacetylene). An acquisition protocol was defined and a data base constituted. All compounds were evaluated below their saturation vapor pressure limit and with similar response amplitudes for comparison purposes. An optimized neural network performed a good discrimination of the set of samples with a classification rate, for the three criteria, reaching 93 %.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...