作者: M Penza , G Cassano
DOI: 10.1016/J.ACA.2003.12.026
关键词:
摘要: The paper reports the use of a chemoresistive multisensor array for recognition some adulterated Italian wines (two white, four red, two rose) added with methanol, ethanol or other same-colour wine. A constituted by thin-film semiconducting metal oxide sensors, surface-activated Pt, Au, Pd, Bi catalysts, has been used to generate chemical pattern volatile compounds present in wine samples. responses towards tested headspace sampling have evaluated. Multivariate analysis including principal component (PCA) as well back-propagation method trained artificial neural networks (ANNs) applied analytical data generated from identify both adulteration and determine content adulterant agent methanol up 10 vol.%. cross-validated ANNs provide highest achieved percentage correct classification 93% correlation coefficient 0.997 0.921 predicted-versus-true concentration agent, respectively.