拉曼光谱结合距离匹配法快速鉴别掺伪食用油

以农贸市场购买的散装问题油为掺伪物,采用大豆油和玉米油为简单背景制备掺伪样本65份,采用4类食用调和油为复杂背景制备掺伪样本40份,收集市售合格食用植物油样本27份。按样本数3∶1划分建模集和校验集,采用拉曼光谱和距离匹配法分别建立简单背景和复杂背景的食用油掺伪快速定性识别模型:在简单背景掺伪下采用全谱建模预测可得真样本识别率为85.7%,伪样本识别率为94.1%,总识别率为91.7%;在复杂背景掺伪下经谱区挑选优化建模预测可得真样本识别率为87.5%,伪样本识别率为100%,总识别率为94.4%。试验结果表明拉曼光谱结合距离匹配法能简单、有效、快速地检测食用植物油是否掺伪。

Rapid Detection of Adulterated Edible Oil Using Raman Spectroscopy and Distance Match Method

Experimental subject were edible oil samples adulterated with problematic bulk oil, which were purchased from the farmers market. There were 65 adulterated samples prepared in the single background (soybean oil and corn oil), 40 adulterated samples prepared in the complex background respectively. 27 true edible vegetable oil samples were collected from the supermarket. The calibration samples and the validation samples were divided according to the ratio 3:1.The rapid qualitative identification model was built based on Raman and distance match method to identify the adulteration quickly. In the case of adulterations made in single background, the model was built based on the whole spectrums. The recognition rate of true sample (pure edible vegetable oil samples) was 85.7%. The recognition rate of false sample (adulteration samples) was 94.1%. The overall recognition rate reached to 91.7%. In the case of adulterations made in complex background, the model was built based on the selected spectrums. The recognition rate of true sample was 87.5%. The recognition rate of false sample was 100%.The overall recognition rate reached to 94.4%.The experimental results showed that Raman spectroscopy combined with distance match method could identify adulterated edible oil simply, quickly and effectively.