Qualitative Identification of Adulterated Huajiao Powder Using Near Infrared Spectroscopy Based on DPLS and SVM

Wu Xi-yu,Zhu Shi-ping,Wang Qian,Long Ying-kai,Xu Dan,Tang Chao

DOI: https://doi.org/10.3964/j.issn.1000-0593(2018)08-2369-05

2018-01-01

Spectroscopy and spectral analysis

Abstract:Huajiao is one of the "eight famous condiments" in China. Some cheaper adulterants were found to be added into Huajiao powder and in order to identify adulterated Huajiao powder qualitatively and quickly, a direct detecting method using near infrared (NIR) spectroscopy coupled with discriminant partial least squares (DPLS) and support vector machine (SVM) had been developed in this study. Wheat bran, rice bran, corn flour and rosin powder with 1 Wt/Wt. % incremental concentration gradient were mixed with red Huajiao powder and green Huajiao powder separately and the adulterated Huajiao powder with range of 1 similar to 54 Wt/Wt. % were prepared. Diffuse NIR spectra (800 similar to 2 500 nm) of pure and adulterated Huajiao powder were acquired. Principal component analysis (PCA) on the spectral data of all 462 samples was used and the first three principal components accounted for 98. 72% of total variation. It was effective for clustering different adulterated Huajiao powder from the main composition PC1, PC2 and PC3 score plot. 347 samples as a calibration set and with the characteristic band spectrum 2 000 similar to 2 200 nm as input, kinds of qualitative models with different spectra pretreatment were established using DPLS and SVM analysis, which were for predicting the rest 115 samples. Results showed that, using different pretreatment methods, and the qualitative identification accuracy of the validation set were between 97. 39%similar to 100%, in which adulterated Huajiao powder could be identified totally. NIRS based on DPLS and SVM is a rapid and nondestructive tool for the qualitative analysis of adulterated Huajiao powder.