基于人工神经网络的黑米花色苷提取的研究

目的:建立黑米花色苷提取的人工神经网络模型,得到最佳提取工艺参数。方法:正交实验与人工神经网络相结合,利用正交实验获得的数据作为神经网络的训练样本,建立输入为实验因素参数,输出为花色苷提取率的神经网络模型;采用人工神经网络模拟和预测黑米花色苷提取的最佳条件和提取率。结果:黑米花色苷最佳提取条件,提取液乙醇/水/盐酸体积比为55:45:0.5,温度50℃,固液比为1:10(g/mL),提取时间为1h,提取次数为4次。结论:人工神经网络模型准确预测花色苷提取率,且得到最佳提取条件下花色苷提取率为3.944%,高于正交实验的3.740%,将神经网络与正交实验结合用于实验条件优化可以缩短优化实验参数的时间,获得比正交实验更优化的实验条件。

Artificial neural network-based optimization of extraction of anthocyanins in black rice

Objective:To get the optimal parameters of anthocyanins of black rice,artificial neural network(ANN) model of extracting anthocyanins was founded.Method:Based on the training of artificial neural networks using orthogonal arrays,a model for the productivity of anthocyanins as the output of the input consisting of five technological parameters for extraction was developed and validated for reliability using selected specimens.The further optimization of optimal values of these parameters obtained using orthogonal array design was conducted based on the AAN model by means of small-step search.Result:The optimal parameters for extraction were ethanol/water/hydrochloric acid 55/45/0.5,temperature 50 ℃,solid-liquid ratio of 1:10,the extraction time is 1 h,times of extraction is 4.Conclusion:AAN-based optimization gave a productivity of anthocyanins of 3.942%,higher than the value of 3.740% from orthogonal array optimization.More optimized technological parameters and higher optimization efficiency can be obtained using combined ANNs and orthogonal array design than using orthogonal array design alone.