Research On Detecting Method And Experimental Study For Edible Blend Oil Based On Parafac

The security of edible blend oil has an influence on the healthy of people and thestable development of market economy. The edible blend oil has comprehensivenutrition and improved the single nutrient composition shortcoming of the ordinaryvegetable oil effectively. Detecting the types and content of vegetable oil in edibleblend oil is an important component of food safety. After the analysis and research onthe methods for monitoring the quality of edible blend oil currently, the paperproposed a new detection method of edible blend oil based on PARAFAC ofthree-dimensional fluorescence spectrum according to the present situation of theimperfect detection technology in detecting major components’ type and content ofedible blend oil.Based on the fluorescence analysis detection theory, the paper do a research onthe feasibility of the fluorescence detection for edible blend oil combining with themultifarious vegetable oil of edible blend oil, the molecular structure characteristics ofthe main components and the complexity of the process.Three-dimensional fluorescence spectrometer for the experiment was instructedin detail about the working principle and structure, the spectrum error principle andcorrection methods. According to the three dimensional fluorescence spectra,excitation/emission spectra and contour map, the paper analysis the outline and shapeof the spectra、 the number and wavelength of apparent peak and the fluorescenceintensity value of peak. The paper mainly focused on the PARAFAC, and thedevelopment of chemo metrics, the analysis of Multivariate calibration and theprinciple of the trilinear decomposition are also studied.Three-dimensional fluorescence spectra of the configured samples of the blend oiland pure vegetable oil was canned by steady-state spectroscopy FS920, selectingsunflower oil, soybean oil and peanut oil simulation edible blend oil Calculating thesample factor number by CORCONDIA.Selecting the correct number of parallel factor analysis, combined with the physical truth, the predict fluorescence excitation/emission spectrums and predictioncalculation of sunflower oil, soybean oil and peanut oil concentration and the averagerecovery were gotten. According to the experimental results the paper concluded thatthe method of combining with three-dimensional fluorescence spectroscopy and usingparallel factor analysis algorithm to detect qualitative and quantitative of edible blendoil is effect. The detection of qualitative and quantitative of edible blend oil based onPARAFAC has practical significance.