Construction and analysis of the near-infrared prediction model of Figs' brix
Download as PDF
DOI: 10.23977/amce.2019.037
Author(s)
Yunhu Wang, Chengzhong Wang, Rui Sun, Lei Sun
Corresponding Author
Rui Sun
ABSTRACT
In order to solve the problem that the quality of hand-picked figs is difficult to be classified and graded, near-infrared spectral diffuse reflection detection technology is adopted. The prediction model of brix was established respectively, and PLS and PCR modeling effects were compared and analyzed. The results showed that PLS modeling was better than PCR. Among them, the brix model was the best. The technical method of this study provides a reasonable reference for the quality classification and grading of hand-picked mature figs.
KEYWORDS
Figs, brix, modeling, genetic algorithm