脂肪酸分析结合基团红外光谱吸收识别稻米油

RECOGNITION OF RICE BRAN OIL BY FATTY ACID ANALYSIS IN COMBINATION WITH FTIR

采用气相色谱法和红外光谱法结合簇类的独立软模式(SIMCA)识别方法对稻米油进行识别分析方法的研究.计算4种植物油(稻米油、大豆油、花生油和菜籽油)中8种脂肪酸含量;油脂样品和糠蜡样品采用傅里变换红外吸收光谱仪分别扫描其光谱,以1 080～1 125 cm^(-1)和1 680～1 780 cm^(-1)波段处的吸收值为不同种类油脂的红外特征信息.用SIMCA识别法分别对4种植物油建立种类识别模型.然后,以各种植物油脂的脂肪酸和红外吸收光谱的吸收值信息为变量,建立各种油脂的主成分分析模型,随机抽取2/3的样本为训练集,1/3为验证集,对所建立模型进行识别并验证,两种变量得到的模型的Q值均为0.9.两种方法的识别结果分别为97%和100%.

In this paper, we studied the recognition analysis method of rice bran oil by gas chromatography, FTIR and SIMCA. The contents of 8 kinds of fatty acids in 4 kinds of vegetable oils （rice bran oil, soybean oil, peanut oil and rapeseed oil） were calculated; the spectra of an oil sample and a bran wax sample were collected by FTIR; and the absorption values at the wavebands of 1 080 to 1 125 cm-1 and 1 680 to 1 780 cm1 were taken as the IR characteristic information of different oils. A type recognition model for the 4 kinds of vegetable oils were constructed by SIMCA method; a main component analysis model for each oil was constructed by taking the fatty acids of each vegetable oil and the IR absorption information as variables; 2/3 samples were extracted randomly as the training set, and 1/3 samples were the validation set to recognize and check the constructed model; and the Q values of the models taking the two variables were 0.9. The recognition rates of the two methods were 97% and 100%, respectively.