Author(s):
Peres, António M. 
; Dias, L.G. ; Barcelos, Tânia P. ; Morais, Jorge Sá ; Machado, A.A.S.C.
Date: 2009
Persistent ID: http://hdl.handle.net/10198/3028
Origin: Biblioteca Digital do IPB
Document details
An electronic tongue for juice level evaluation in non-alcoholic beverages
Description
An electronic tongue with 36 cross-sensibility polymeric membranes was built and applied for qualitative beverage analysis. An electronic tongue with 36 cross-sensibility polymeric membranes was built and applied for qualitative beverage analysis. The objective was to distinguish non-a1coholic beverages with different leveis of added fruit juice. The different signal profiles recorded by the electronic tongue devi ce, together with a supervised , multivariate statistical method for pattem recognition, were used to differenliate 4 beverage groups with different added fruit juice contents: higher than 30%, between 14% and 30%, between 5% and 10% and between 0.1% and 2%. A set of 16 portuguese beverages (4 for each group), purchased in commercia[ superrnarkets, were analyzed. A stepwise linear discriminant analysis was used allowing a very satisfactory overall correct classification of 100% for the original grouped cases and of 93 .8% for the "[eaving one-out" cross-validation procedure
An electronic tongue with 36 cross-sensibility polymeric membranes was built and applied for qualitative beverage analysis. An electronic tongue with 36 cross-sensibility polymeric membranes was built and applied for qualitative beverage analysis. The objective was to distinguish non-a1coholic beverages with different leveis of added fruit juice. The different signal profiles recorded by the electronic tongue devi ce, together with a supervised , multivariate statistical method for pattem recognition, were used to differenliate 4 beverage groups with different added fruit juice contents: higher than 30%, between 14% and 30%, between 5% and 10% and between 0.1% and 2%. A set of 16 portuguese beverages (4 for each group), purchased in commercia[ superrnarkets, were analyzed. A stepwise linear discriminant analysis was used allowing a very satisfactory overall correct classification of 100% for the original grouped cases and of 93 .8% for the "[eaving one-out" cross-validation procedure
Document Type
Conference Object
Language English
Language English
| Financiadores do RCAAP | |||||||
|
|
|
|
|
|
||