基于Matlab的原料乳中金黄色葡萄球菌生长模型的分析

摘要

研究了在12、25、37℃三个温度下的金黄色葡萄球菌的生长状况，以 Matlab为平台，结合Gompertz、Logistic、与神经网络三种模型进行深入分析，本文列举出预测微生物模型的10种常用评价方法及对应的参数，以其中的修正决定系数R2Adj，均方误差MSE，预测标准误差SEP三种评价函数为主，分别对模型的拟合程度、精确度、准确度进行了比较。综合考虑模型的性质，最终确定Gompertz为金黄色葡萄球菌的最佳生长模型；同时，建立以温度为自变量的平方根模型，以此来预测原料乳中金黄色葡萄球菌不同温度下的最大生长速率，为金黄色葡萄球菌的检测提供理论依据。%In this paper, the growth of Staphylococcus aureus in raw milk at 12, 25 and 37℃was studied. Base on Matlab, the Gompertz, logistic and ANN (Artificial Neural Network) was used to analysis the data thoroughly, common evaluations methods for predictive microbial growth models with emphasis on the concept and role of 10 related evaluation parameters were enumerated in this paper. From them,R2Adj, MSE, SEP was chosen to evaluate the models' fitting degree, precision and veracity, consider all the features of the models, ultimately, Gompertz regression model can predict the growth rate of Staphylococcus aureus best of the three. Meanwhile, a square root model had been developed predicting the maximum growth rate of Staphylococcus aureus in raw under the condition of different temperature as a function of temperature, which provided a theoretical basis for the detection of Staphylococcus aureus.