Inspection Of Camellia Oil Adulteration Based On Near Infrared Spectroscopy

Fatty acid content, proportion of camellia oil is to be roughly the same as the West olive oil, known as the "Oriental olive oil", has a high nutritional value, some indexes are even better than olive oil, so by the enthusiasm of the vast number of consumers. The existing market camellia oil variety, various species of various qualities, but also because of its nutritional value and the price is expensive, often appears the phenomenon of adulterated. And judging adulterated standards can only qualitative GB/T5539- 2008 "qualitative test of oils" inspection of oil quality and no up detection of adulterated camellia oil from type and content, and the detection time is longer, the steps of redundancy. This requires on camellia seed oil content and characteristic to study, to seek a rapid and accurate detection method.This paper takes the camellia oil as the research object, in the 350nm~1800nm spectral range of two element doped with soybean oil, rapeseed oil, peanut oil, camellia oil,camellia oil of essence and; three element system of soybean oil and rapeseed oil doped,soybean oil and peanut oil, rapeseed oil and peanut oil; special spectrum in Figure four element system of soybean oil, peanut oil, rapeseed oil doped. Using visible / near-infrared spectroscopy preliminary establishment of rapid identification of camellia oil model,design of portable devices and can provide theoretical data for rapid detection. The results showed that the obtained:1、Different system and the same system doped with different oil samples have better pretreatment range of different.2、The use of discriminant analysis method, near infrared technology combined with Linear Discriminant Analysis(LDA) based on linear discriminant analysis, support vector machine classification(C-SVM) support vector machine classification analysis for qualitative analysis of the experimental samples, found that each system to distinguish the effect is very good, correct prediction rates are above 97%. The amount of adulteration in5%~50% within the scope of the camellia oil discrimination rate reached 100%, but in the range below 5%, to distinguish the effect is not ideal.3、In order to realize the detection of adulteration in camellia oil, will be the subject of the different oil samples of each system, select the optimal spectral range, pretreatment methods and modeling different analysis methods, and the model is validated, the results are as follows:The two element system:For the incorporation of different gradient(0% ~ 50%) two yuan system adulterated camellia oil model of soybean oil, correlation coefficient of calibration of the model(rc),set the root mean square error of calibration(RMSEC) and cross validation correlationcoefficient(rcv) and standard error of cross validation(RMSECV) were 0.9748, 0.02414,0.9655 and 0.02812. The prediction set the root mean square error of prediction(RMSEP)was 0.02649, the correlation coefficient of prediction set(rp) 0.96599.For the incorporation of different gradient(0% ~ 50%) two yuan adulterated camellia oil rapeseed oil system model, the rc, RMSEC, rcv and RMSECV were 0.9706, 0.02587,0.9603 and 0.0309. The prediction set RMSEP to 0.02241, rp is 0.9765,R2 is 0.97682.For the incorporation of different gradient(0% ~ 50%) two yuan system adulterated peanut oil camellia oil model, the rc, RMSEC, rcv and RMSECV were 0.9899, 0.01553,0.9749 and 0.02494. RMSEP, the prediction set is 0.03330, rp is 0.9472, R2 is 0.96041.The three element system:For the incorporation of different gradient(0% ~ 50%) three yuan system adulterated camellia oil model of soybean oil and peanut oil, its rc, RMSEC, rcv and RMSECV were0.9705, 0.02613, 0.9503 and 0.03538.The prediction set RMSEP is 0.04459, rp is 0.9070,R2 is 0.9167.For the incorporation of different gradient(0% ~ 50%) three yuan system adulterated camellia oil of soybean oil and Rapeseed Oil model, the rc, RMSEC, rcv and RMSECV were 0.9756, 0.02360, 0.9686 and 0.02744.The prediction set RMSEP is 0.02924, rp is0.9594, R2 is 0.9839.For the incorporation of different gradient(0% ~ 50%) three yuan system adulterated camellia oil and peanut oil Rapeseed Oil model, the rc, RMSEC, rcv and RMSECV were0.9813, 0.02067, 0.9381 and 0.03794.The prediction set RMSEP to 0.03295, rp is0.9484,R20.9728.The four element system:For the incorporation of different gradient(0% ~ 50%) four yuan system adulterated camellia oil model of soybean oil, peanut oil and rapeseed oil, rc, RMSEC, rcv and RMSECV were 0.9851, 0.01844, 0.9742 and 0.02529. RMSEP is 0.0455,rp is 0.9016, R2 is0.9618.