Establishment of a Quantitative Prediction Model for Quality Indices of Flaxseed Oil in Kazakhstan Based on Near-infrared Spectroscopy
Article
Figures
Metrics
Preview PDF
Reference
Related
Cited by
Materials
Abstract:
In order to establish a rapid, accurate, and efficient method for the determination of acid value, peroxide value and fatty acid composition of flaxseed Oil in Kazakhstan, and characterize flaxseed oil’s quality, safety and characteristics using scientific means,116 representative samples were selected in this study. The content of peroxide value, acid value, and the contents of palmitic acid, stearic acid, oleic acid, linoleic acid and linolenic acid of the flaxseed oil samples were determined by iodometry, the hot ethanol metho and GC-MS, while the characteristic spectra were established by using a near-infrared component analyzer. The characteristic spectra after different pretreatments were fitted according to the partial least squares. After the regression analysis, the near-infrared quantitative prediction models for the acid value, peroxide value, and the contents of palmitic acid, stearic acid, oleic acid, linoleic acid and linolenic acid of the flaxseed oil were established. After the model was verified and the prediction accuracy was evaluated, it was concluded that the correlation coefficient (R) was 0.960 1~0.985 7, the absolute deviation was 0.039 14~1.246 7, and the root mean square error (RMSEP) was 0.059 21~1.430 2, indicating that the model has a good prediction effect and can effectively predict the quality indices of flaxseed oil in Kazakhstan. The establishment of this model provides a fast approach for simultaneous determination of the acid value, peroxide value and multiple fatty acid components of the Kazakhstan linseed oil, which greatly shortens the time duration of the customs clearance.
