论文部分内容阅读
采用傅里叶变换红外光谱(FTIR)结合簇类独立软模式识别技术(SIMCA)建立了真伪食用油的快速鉴别方法.该方法依据FTIR的指纹特性,收集并分析了53个合格食用油和13个伪造食用油的FTIR谱图;通过对谱图取二阶导数和标准化处理,主成分分析(PCA)提取特征变量;采用SIMCA方法分别随机选取43个合格食用油和9个伪食用油样品的FTIR谱图组成训练集,构建得到真伪食用油的SIMCA分类模型.该模型经过剩余10个合格食用油和4个伪食用油的验证,正确识别率达到了100%.说明FTIR结合SIMCA可能成为快速鉴别食用油真伪的一种新方法.
A rapid identification method of authentic edible oil was established by using Fourier transform infrared spectroscopy (FTIR) combined with cluster independent soft pattern recognition (SIMCA) method. According to the fingerprint characteristics of FTIR, 53 qualified edible oils and FTIR spectra of 13 fake edible oils were obtained. Principal component analysis (PCA) was used to extract eigenvalues by taking the second derivative and normalization of the spectra. SIMCA was used to select 43 edible oils and 9 pseudo-edible oils FTIR spectra to form a training set to build the SIMCA classification model of authentic and edible oil.It has been verified by the remaining 10 qualified edible oils and 4 pseudo-edible oils that the correct recognition rate is 100% Become a new method to quickly identify the authenticity of edible oil.