燕麦生鲜湿面贮藏品质变化及货架期预测模型研究

目的 构建一种新的径向基函数神经网络货架期预测模型。方法 研究不同贮藏温度下燕麦生鲜湿面的微生物、理化等指标的变化情况，通过Pearson相关性分析，筛选影响燕麦生鲜湿面货架期的主要因素，利用微生物生长动力学模型和径向基函数神经网络模型对燕麦生鲜湿面的剩余货架期进行预测。结果 微生物生长动力学模型不能很好地拟合燕麦生鲜湿面菌落总数的变化情况，预测精度较差，径向基函数神经网络预测模型的预测值与实际值的相对误差为2.66%。结论 径向基函数神经网络预测模型的效果较好，为以后食品货架期的预测提供了一定的参考依据。

Storage Quality Change and Shelf Life Prediction Model of Oat Fresh Wet Noodles

The work aims to construct a new radial basis function (RBF) neural network shelf life prediction model. The microbial, physical and chemical indexes of oat fresh wet noodles at different storage temperatures were studied, and the main factors affecting the shelf life of oat fresh wet noodles were screened out through Pearson correlation analysis. The remaining shelf life of oat fresh wet noodles was predicted by models of microbial growth kinetics and the RBF neural network, respectively. The microbial growth kinetics model could not fit the change of the total bacterial count of oat fresh wet noodles very well, and the prediction accuracy was poor. On the contrary, the relative error between the predicted value and the actual value of the RBF neural network prediction model was 2.66%, which was very little. The RBF neural network prediction model is effective, which provides a certain reference for the future prediction of food shelf life.