Neumann L. G., Konzen P. H. A., Ferrão M. F., Furtado J. C., Morgano M. A., Bragagnolo N., Ferreira M. M. C.,  “Desenvolvimento e Otimização de Rotinas Não Destrutivas de Análise de Tanino em Café Empregando Algoritmo Genético” [“Development and Optimization of Non-Destructive Routines for Analysis of Tannin in Coffe Using Genetic Algorithm”]. Curitiba, PR, Brazil, 23-25/10/2002: XXII Encontro Nacional de Engenharia de Produção [The XXII National Meeting on Production Engeneering], Anais [Annals], 1-8 (2002). Poster.
[Article. More information about the meeting see at Notes.]
 

Abstract.
In this work a new methodology of analysis for the quantification of tannin in green coffee is proposed, using genetic algorithm optimization with diffuse infrared spectroscopy and partial least squares regression method (PLS). After the best wavelength selection, which results in models with better coefficient regression (R²), these were improved using routines based on genetic algorithm heuristic method. For the evaluation of the optimization Standar Error Validation (SEV) and Standar Error Calibration (SEP) were used. The results reflect the importance to choose the set variable selection, as well as the kind of pre-processing applied. These results let us conclude that good models aiming the prediction of tannin levels can be obtained, and that reflection techniques are adequate to allow a fast obtain of spectra of green coffee milling and not to create wastes which are harmful to the environment.

Keywords. Genetic Algorithm; Multivariat Regression; Green Coffee.

Keywords Plus.