基于不同贮藏温度下油茶籽油氧化模型的建立

Establishment of Oxidation Model of Camellia oleifera Seed Oil Based on Different Storage Temperatures

研究贮藏过程中油茶籽油品质指标变化及预测模型,为油茶籽油的安全贮藏及品质预测提供参考,探讨不同温度下(20、40、60℃)油茶籽油酸价和过氧化值随储藏时间的变化趋势,采用动力学和人工神经网络模型进行拟合,并采用25℃贮藏条件下油茶籽油的酸价和过氧化值对模型进行校验。结果表明:酸价和过氧化值的变化随储藏时间的增加而增加,而且温度越高变化越快;酸价的动力学和神经网络模型的相关系数分别为0.9959和0.9980;过氧化值动力学和神经网络模型的相关系数分别为0.9339和0.9913;模型拟合效果极好;模型校正过程,酸价和过氧化值动力学模型预测值和实测值之间的平均相对误差分别为:6.66%、42.17%;而酸价和过氧化值对应的神经网络模型预测效果更好,预测值和实测值之间的平均相对误差分别为:2.34%、6.38%;表明神经网络模型较之动力学模型,泛化能力更好,误差更小,能更加准确预测油茶籽油在储藏过程中酸值和过氧化值的变化趋势。

The change of quality index and prediction model of Camellia oleifera seed oil during storage were studied to provide a theoretical basis for the safe storage and quality prediction of Camellia oleifera seed oil.The acid value and peroxidation value of Camellia oleifera seed oil at different temperatures(20,40 and 60℃)were investigated with the storage time,and the kinetics and artificial neural network models were used for matching.The acid value and peroxide value of Camellia oleifera seed oil were used to verify the model under 25℃storage conditions.The results indicated that the changes of acid value and peroxide value increased with the increase of storage time,and the higher the temperature was,the faster the change was;the correlation coefficients of the kinetics and the neural network models of acid value were 0.9959 and 0.9980,respectively;the correlation coefficients of the kinetics and the neural network models of peroxidation value were 0.9339 and 0.9913,respectively.The model fitting effect was excellent.The average relative errors between the predicted value and the measured value of the kinetic models corresponding to acid value and peroxide value were 6.66%and 42.17%,respectively.The neural network models corresponding to acid value and peroxide value predict better,the average relative error between predicted and measured values were 2.34%and 6.38%,respectively.It indicated that the neural network model was more better than the kinetic model and the generalization ability of the neural network model was better and the error was smaller,so the neural network model could more accurately predict the change trend of acid value and peroxide value of Camellia seed oil during storage.