近红外光谱技术快速检测紫薯半干面中菌落总数及新鲜度

Rapid Determination of Total Viable Count (TVC) and Freshness in Semi-dried Purple Sweet Potato Noodles by NIR Spectroscopy

利用近红外光谱分析技术和化学计量学分析方法,对紫薯半干面菌落总数(TVC)及其新鲜度鉴别进行研究。结合偏最小二乘法对比10种光谱预处理方法,建立紫薯半干面的TVC预测模型,最终选定SNV+2D为最优预处理方法。经优化TVC最优模型的校正集和预测集的决定系数R分别为0.99213和0.97537,交互验证均方根误差及预测集均方根误差分别为0.250和0.445。采用主成分分析结合马氏距离的定性判别分析法,定性鉴别紫薯半干面新鲜程度。当选用标准归一化为光谱预处理方法时,正确识别率达到100%。利用近红外光谱分析技术可以快速、无损地检测紫薯半干面中TVC及新鲜度。

This paper studied the total viable bacteria count（TVC） and freshness in semi-dried purple sweet potato noodles by using near infrared spectrum analysis technology and chemometrics analysis method. The optimal models for TVC were established by comparing 10 different preprocessing methods and partial least squares（PLS） factor number.Experimental results showed that the performance of standard normal variate and the second derivative（SNV＋2D） were selected as the optimal pretreatment method. The correlation coefficients of calibration（Rc）, correlation coefficient of prediction（Rp）, root mean square error of cross validation（RMSECV） and root mean square error of prediction（RMSEP）were achieved 0.99213, 0.97537, 0.250, 0.445 respectively. The discriminant analysis（DA） method of principal component analysis combined with mahalanobis distance was used to identify the freshness of semi-dried purple sweet potato noodles. The recognition correct rate reached 100% by the tandard normal variate（SNV） combined with Mahalanobis distance（MD）. This paper succeeded to show the feasibility of measuring the TVC and freshness of Semi-dried purple sweet potato noodle by NIR spectroscopy technique.