Abstract
Vegetable blend oils play a vital role in the daily human diet, but they have become the target of forgeries. This work proposes a method for the rapid on-site quantification of high-value pure vegetable oil (e.g., extra virgin olive oil (EVOO)) content in vegetable blend oils by combining Raman spectroscopy and metaheuristics-based variable selection models. First, the corn oil-EVOO and peanut oil-EVOO vegetable blend oil samples were prepared for Raman spectra measurement. Then, partial least squares regression and four metaheuristics-based variable selection models, including genetic algorithm combined moving window (GA-MW), particle swarm optimization combined moving window (PSO-MW), grey wolf optimizer combined moving window (GWO-MW), and whale optimization algorithm combined moving window (WOA-MW), were applied to the collected Raman spectra for a comparative study. The method developed in this work could accurately quantify the EVOO content in vegetable blend oil samples. In addition, the stability tests and limit of detection values verified the stability and sensitivity of the developed method. Finally, two-tailed paired t-tests identified that there was no significant difference between the gas chromatography–mass spectrometry method and the developed method. Therefore, the developed method provides a rapid and cost-effective strategy for quantitative authentication of vegetable blend oils.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call