An expert system based on 1H NMR spectroscopy for quality evaluation and adulteration identification of edible oils
یک سیستم خبره مبتنی بر طیف سنجی 1H NMR برای ارزیابی کیفیت و شناسایی زون روغن های خوراکی-2019
The advantages of nuclear magnetic resonance (NMR) such as nondestructive and simultaneous detection, high reproducibility and rapidity make it easily develop the objective and credible methods for food analysis and identification. In this study, we developed a computer-aided, MATLAB-scripted expert system which enables NMR data to distinguish different edible oils and evaluate the quality of edible oils. The NMR spectral data of seven species of most popular vegetable edible oils in China were used to establish the assessment criterions including the content percentage of fatty acids and the quality parameters of edible oils. In our case, the identification accuracy of vegetable origin for the pure edible oils is 95.83% and that for the mixed edible oils is 89.58%, and all the recycled waste cooking oils and fried oils were correctly screened out and identified by the expert system. Further, the quality information of the edible oils was also provided. Our results show that the current expert system is a fast, easy-operated and convenient tool for the adulteration identification and quality control of edible oils.
Keywords: Edible oil | 1H nuclear magnetic resonance | Food analysis | Food composition | Quality parameters | Expert system | Classification | Identification | Adulteration | Quantitative analysis