广域照射拉曼光谱技术结合簇类独立软模式法快速鉴别原料肉及掺假肉

应用广域照射（wide area illumination，WAI）拉曼光谱技术与簇类独立软模式（soft independent modeling of class analogy，SIMCA）法，结合多元散射校正（multiplicative scatter correction，MSC）和光谱仪降噪和波长标定（spectrometer noise reduction and wavelength calibration，SNRWC）降噪技术，建立鸭、羊、猪3 种原料肉及掺假羊肉的定性识别模型。结果表明：经MSC与SNRWC处理后，鸭、羊、猪3 种原料肉之间及羊肉、掺假羊肉之间的主成分分析结果具有明显的聚类趋势，在此基础上建立SIMCA定性分类模型，对不同产地的37 个原料肉样品种属进行定性鉴别，识别正确率达100%；对4 个掺假羊肉和5 个未掺假羊肉样品识别正确率也为100%。因此，拉曼光谱分析技术结合有效的数据前处理方法及化学计量学方法可对鸭、羊、猪原料肉种属及掺假羊肉进行鉴别。与常规方法相比，该检测过程快速、方便，并且无需样品前处理。

Rapid Identification of Pure and Adulterated Meat by Wide Area Illumination Raman Scheme Coupled with Soft Independent Modeling of Class Analogy

Qualitative recognition models were established by using wide area illumination (WAI) Raman scheme and soft independent modeling of class analogy (SIMCA) for rapid identification of duck, lamb, pork and adulterated meat. The spectra of all samples were pre-processed by multiplicative scatter correction (MSC) and spectrometer noise reduction and wavelength calibration (SNRWC) method and then principal component analysis was implemented to observe the clustering trend. It turned out that most of the duck, lamb and pork samples as well as most of the lamb samples and adulterations were well separated. Finally, the qualitative classification models were established by using SIMCA. All validation samples were identified by the SIMCA model with an accuracy of 100%, including 37 meat samples from different species and geographical origins, as well as four adulterated and five unadulterated lamb samples. Therefore, the WAI Raman scheme coupled with chemometrics could distinguish among lamp, duck, and pork and adulterated lamb, and it proved to be more fast, convenient, without the need for any sample pretreatment compared with the routine method.