论文部分内容阅读
加热终点温度(EPT)是在肉制品生产中控制食源性疾病的关键因素。现有的EPT检测方法虽然很多,但普遍存在耗时长,样品处理繁杂等不足。本文验证了近红外光谱法检测肉类EPT的可行性,为其在肉制品生产流通中的应用提供理论依据。采用近红外光谱(NIR)结合偏最小二乘法(PLS),由校正集建立模型,用已知EPT的检验集检验模型的判别效果。采用校正集的内部交叉验证均方差(RMSECV)确定猪肉、牛肉、羊肉的主成分数分别为12,10,13,此时校正集的RMSECV值分别为2.57%,1.98%和0.89%;所得校正模型的预测加热终点温度与实际加热终点温度之间的相关系数分别为0.9808,0.9872,0.9892;由所建模型对检验集样品的检验结果看,实际加热终点温度与近红外模型预测的加热终点温度具有很高的相关性,预测值的相关系数r分别为0.9940,0.9968,0.9915,预测均方差RMSEP分别为4.03%,3.86%,3.57%,预测标准差σ为1.06,0.92,1.49。由此得出结论:近红外光谱法能很好地用于肉制品EPT检测。
Heated End Point Temperature (EPT) is a key factor in controlling foodborne illness in the manufacture of meat products. Although there are many existing methods for detecting EPT, there are many shortcomings, such as long time consuming and complex sample processing. This paper verifies the feasibility of detecting meat EPT by near infrared spectroscopy and provides a theoretical basis for its application in the meat product circulation. Near infrared spectroscopy (NIR) combined with partial least squares (PLS) was used to establish the model from the calibration set and the discriminant effect of the model was tested using a known EPT test set. The root mean square error (RMSECV) of the calibration set was used to determine the main components of pork, beef and mutton, respectively. The RMSECV of the calibration set was 2.57%, 1.98% and 0.89%, respectively. The correlation coefficients between the predicted heating end point temperature and the actual heating end point temperature are 0.9808, 0.9872 and 0.9892, respectively. According to the test results of the test sample set by the model, the actual heating end point temperature and the predicted end point temperature of the near infrared model The correlation coefficients r of the predicted values were 0.9940, 0.9968 and 0.9915 respectively. The predicted RMSEPs were 4.03%, 3.86% and 3.57%, respectively. The prediction standard deviation σ was 1.06, 0.92 and 1.49 respectively. This concludes that NIR spectroscopy is well suited for EPT testing of meat products.