拉曼光谱法快速检测猪肉脯中的掺伪鸡肉

Rapid detection of adulterated chicken meat in dried pork slice by Raman spectroscopy

中文关键词:  拉曼光谱  猪肉脯  肉脯掺假  变量筛选  定量检测  定性鉴别

英文关键词:Raman spectroscopy  dried pork slice  adulteration  variables screening  quantitative detection  qualitative identification

中文摘要:

      目的 建立拉曼光谱法快速、准确、无损地检测猪肉脯样品中掺假鸡肉的方法。方法 制备33份猪肉中掺入不同比例鸡肉的肉脯样品, 采集拉曼光谱数据, 分别采用标准正态变换、多元散射校正、卷积平滑、归一化、一阶导数等5种不同预处理方法, 对原始光谱数据进行预处理, 采用连续投影算法、竞争性自适应重加权算法及随机蛙跳算法对光谱数据进行特征波长筛选, 建立偏最小二乘法(partial least squares, PLS)模型对猪肉脯进行定性定量判别。结果 拉曼光谱数据经过多元散射校正处理的效果最佳, 竞争性自适应重加权算法竞筛选效果更佳, 构建猪肉脯中猪肉含量的PLS定量模型, 其预测集决定系数和预测均方根误差分别为0.9762、7.2998。建立的PLS判别模型的校正集和预测集总判别正确率分别为100.00%和98.33%。结论 拉曼光谱分析技术可有效用于定性鉴别猪肉脯是否掺伪及定量分析猪肉肉脯中掺入鸡肉的比例, 为肉脯掺假的快速无破坏性检测的应用提供支持。

英文摘要:

      Objective To establish a rapid, accurate and non-destructive method for the detection of adulterated chicken in dried pork slice samples by Raman spectroscopy. Methods The 33 samples of dried pork slice mixed with different proportions of chicken meat were prepared, and the Raman spectral data were collected. Five kinds of different pretreatment methods, including standard normal transformation, multivariate scattering correction, convolutional smoothing, normalization and first derivative, were used to preprocess the original spectral data. The successive projections algorithm, competitive adaptive reweighted sampling algorithm and random frog algorithm were used to screen the characteristic wavelength of spectral data. A partial least squares (PLS) model was established for the qualitative and quantitative discrimination of dried pork slice. Results The results of Raman spectral data were best after multivariate scattering correction, and the competitive adaptive reweighted sampling algorithm was more effective in screening. The PLS quantitative model of pork content in dried pork slice was constructed, and the predictive set determination coefficient and predictive root-mean-square error were 0.9762 and 7.2998, respectively. The correct rate of calibration set and prediction lumped of the established PLS model was 100.00% and 98.33%, respectively. Conclusion Raman spectral analysis technology can be used to qualitatively identify the adulteration of dried pork slice and quantitatively analyze the proportion of dried pork slice mixed with chicken, which provides support for the application of rapid and non-destructive detection of adulteration of dried pork slice.