PLS-Discriminant Analysis of Adulterated Yak Milk Powder based on Two-Dimensional Correlation Near-Infrared Spectroscopy
To develop a rapid discrimination analysis method for adulterated yak milk powder using two-dimensional near-infrared(2D-NIR)correlation spectroscopy combined with partial least squares discriminant analysis(PLS-DA).Objective:This study.Methods:56 adulterated samples were prepared by blending yak milk powder with regular milk powder in specific proportions.Using the Kennard-Stone(KS)method,the samples were divided into calibration and prediction sets at a ratio of 3∶1.We gathered near-infrared(NIR)spectra from all samples and generated 2D correlation spectra using concentration as the external perturbation.Following this,employing 2D-NIR synchronous and auto-correlation spectra analysis,we identified 26 characteristic variables to develop the PLS-DA discrimination model for detecting adulterated yak milk powder.Results:The 26 characteristic wavelengths identified through 2D-NIR analysis were primarily concentrated in spectral regions showing notable differences and were closely associated with key nutritional components like proteins and fats in the milk powder.The developed PLS-DA discrimination model for detecting adulterated yak milk powder achieved a coefficient of determination(R2)exceeding 0.97,withclassification accuracy rates of 100％for both the calibration and prediction sets.Conclusion:The results demonstrate that 2D-NIR combined with PLS-DA method can identify subtle differences in adulterated yak milk powder and enable rapid discrimination.This method offers a new approach for detecting adulteration in dairy products.
NETL
NSTL
万方数据
