Investigation of frying oil quality using VIS/NIR hyperspectral analysis

Abstract

Traditional chemical methods of analyzing frying oil quality are time-consuming and not amenable to on-line measurement. The main objective of this study was to evaluate quality changes of heated oils based on visible/near infrared spectral analysis using a hyperspectroradiometer. The reflectance spectra of the heated oils were analyzed within the range 400-1,750 nm. Acid value, total polar component, and viscosity of oil samples were used as indicators of different quality levels of oil. Partial least squares calibration models were developed for quantitative evaluations of these parameters. The R 2 and root mean square error for each prediction were calculated to assess the prediction capability of calibration models. The study demonstrated that using the established calibration models, quality parameters could be predicted with R 2 values over 0.92. © 2009 AOCS.