Abstract

DURAN-ACEVEDO, Cristhian Manuel; GUALDRON-GUERRERO, Oscar Eduardo  and  HERNANDEZ-ORDONEZ, Mariela. Electronic Nose to Determine the Maturity Index of the Tree Tomato (Cyphomandra Betacea Sendt). Ing. invest. y tecnol. [online]. 2014, vol.15, n.3, pp.351-362. ISSN 2594-0732.

This paper presents the development of an Electronic Nose for nondestructive monitoring of tree tomato ripening process (Cyphomandra Betacea Sendt). An array of 16 chemical gas sensors was arranged for the detection of three ripeness levels of tree types of tomato (green, ripe and overripe). A Probabilistic Neural Network (PNN) as variable selection technique (Simulated Annealing) was coupled to improve the result and the PCA (Principal Component Analysis) technique was applied to discriminate each one of volatile compounds. A number of measures for physicochemical tests were analyzed with the goal of evaluating the physical, chemical and sensory properties (i.e, pH, acidity and Brix) of the product, and the results of the Electronic Nose were compared. The olfactory system was able to classify the samples of tree tomato in three different stages with very high accuracy, to reach a success rate 99.886% in classification.

Keywords : multisensor system; PCA; PNN; variable selection; tree tomato.