Study On The Authenticity And Quality Evaluation Of Edible Vegetable Oil Based On Chemometrics

Edible oil provides nutritional properties to people,and its quality is closely related to human beings.However,it is a commonplace that reports of quality problems with edible oil in recent years,so it is necessary to establish a fast,accurate,and highly sensitive method for evaluating the authenticity and quality of vegetable oil.In this paper,three spectroscopic techniques combined with chemometrics methods were used to qualitatively and quantitatively study edible vegetable oil from the aspects of authenticity,adulteration,and quality.The specific contents are as follows:In the second chapter,the visible-near-infrared spectrum(Vis-NIR),middle infrared spectrum(MIR),and three-dimensional fluorescence spectrum(EEMs)data of five kinds of vegetable oils were collected.Two two-dimensional pattern recognition algorithms,principal component analysis(PCA)and partial least squares discriminant analysis(PLS-DA),were used to classify vegetable oils based on the MIR and Vis-NIR data.While for the EEMs data,two three-dimensional pattern recognition algorithms,multidimensional principal component analysis(M-PCA)and N-dimensional partial least squares discriminant analysis(N-PLS-DA),were used to process the data to establish a vegetable oil classification model.The results showed that spectral detection techniques combined with chemometrics could be applied to the identification of different vegetable oils,and Vis-NIR and MIR were better than EEMs for the classification of five vegetable oils.In the third chapter,soybean oil adulterated olive oil was taken as the research object to collect the data on Vis-NIR,MIR,and EEMs.For MIR and Vis-NIR data,PCA and PLS-DA were used to calculate the two-dimensional data to build an olive oil adulteration identification model,and M-PCA,N-PLS-DA,and self-weighted alternating trilinear decomposition(SWATLD)combined with back-propagation artificial neural network(BP-ANN)were used to construct a model to detect of adulterated olive oil in the EEMs data.The results showed that all three spectroscopic techniques provided reliable results for the determination of olive oil adulteration,and the adulteration models established by MIR and NIR spectroscopy combined with PLS-DA had a better recognition rate than EEM spectroscopy.MIR and VisNIR could be used as reliable,fast and environmentally friendly methods for identifying adulterated olive oil.In the fourth chapter,a linear partial least squares(PLS)method based on MIR and VisNIR data was used to develop predictive models to determine the content of soybean oil quality parameters(peroxide value,acid value,five fatty acids,phytosterols,and tocopherols),and the quality of soybean oil was evaluated.The correlation coefficients of the proposed models ranged from 0.8663 to 0.9902,and the results indicated that Vis-NIR and MIR spectroscopy techniques combined with chemometrics can be used for rapid quality evaluation of edible vegetable oils.The above content shows that spectroscopy combined with chemometrics is a feasible method for authenticity evaluation and quality assessment of edible vegetable oil,which is simple,rapid,reliable and environmentally friendly.