Rapid qualitative and quantitative identification of adulterated wheat flour in edible milk powder by a portable near-infrared spectroscopy.

Due to its high nutritional value and increasing global demand, milk powder is susceptible to adulteration for economic reasons. This study aimed to use a portable near-infrared (NIR) spectrometer to rapidly detect and analyze wheat flour adulteration in milk powder, covering both qualitative and quantitative tasks. The qualitative analysis involved identifying the presence and extent of adulterants in the milk powder samples using spectral data. The quantitative analysis aimed to measure the amount of wheat flour accurately added. For the qualitative analysis, we processed the NIR spectral data using principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and support vector classification (SVC). The results showed that these methods effectively discriminated adulterated milk powder samples and identified the degree of adulteration, achieving an accuracy rate of 100%. For quantitative analysis, this research used partial least squares regression (PLSR) and support vector machine regression (SVR) to model and predict the proportion of wheat flour adulteration. The experimental results showed that both PLSR and SVR models could accurately predict the adulteration level. In this research, more robust detection models were established by using 14 spectral bands selected by CARS. In particular, the SVR model achieved a correlation coefficient (R) of 0.9921, a root mean square error (RMSE) of 2.1360%, and a residual prediction deviation (RPD) of 8.0626 between predicted and actual values. These results indicate that portable near-infrared spectrometers are highly efficient and reliable for detecting milk powder adulteration. This study provides a convenient, rapid, and accurate tool for monitoring food safety, protecting consumer rights, and maintaining market order.