Rapid Indentification Model And Optimized Of Waste Cooking Oil Based On Near Infrared Spectra

As a kind of daily necessaries, the safety of edible vegetable oil isextremely important. However, due to the lack of an effective technicalmethod and the instrument to identify waste cooking oil and adulterationoil, the production and sales of waste cooking oil and adulteration oilhaven’t be effectively prohibited yet. It is serious influencing the safetyof edible oil. Hence for the safety control for edible vegetable oil and toensure safety in food hygiene and health of people, it is essentiallynecessary to research and develop a convenient method which canidentify waste cooking oil and adulteration oil professionally, rapidly andaccurately. And then it is possible to develop a rapid detection portableinstrument of waste cooking oil and adulteration oil, and to makescientifically the national standard for rapid detection on the basis of thetechnical method.The study adopts the near-spectroscopy analysistechnology to rapidly identify edible vegetable oil and waste cooking oil.A qualitative-quantitative method for detection techniques of ediblevegetable oil and waste cooking oil is developed in this thesis. The maincontents and conclusions are as follows:A certain typical samples of group library and NIR database of mainedible vegetable oil and waste cooking oil has been established andstudied. The model for discriminating the waste cooking oil and fourkinds of normal edible vegetable oil is developed using near infraredspectroscopy combined with hierarchical cluster analysis (HCA), learningvector quantization neural network (LVQ-NN), and support vectormachine (SVM).A quantitative modeling method of waste cooking oil andedible vegetable oil is developed. A quantitative modeling method for lowcarbon fatty acids by combining near infrared spectroscopy with supportvector regression is studied. The result shows that it is feasible to use the near infrared method to identify the waste cooking oil rapidly andaccurately and for quantitative analysis of waste cooking oil andadulteration oil. it is feasible to use near infrared spectral technology todetect the low carbon fatty acid content.