摘要
目的:建立花生红衣色素综合评价值的BP人工神经网络模型,得出最佳提取工艺参数。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,输出为花生红衣色素的综合评价值的神经网络模型;采用人工神经网络模拟和预测花生红衣色素提取的最佳条件和综合评价值。结果:花生红衣色素最佳提取条件为温度55℃,提取剂为乙醇,乙醇的浓度为60%,提取时间为3 h,料液比为1∶35(g/m L)。结论:人工神经网络模型准确预测花生红衣色素提取的最佳条件和综合评价值,且得到最佳提取条件下花生红衣色素的综合评价值为33.0824,优于正交实验33.08的综合评价值。
Objective. To get the optimal parameters of red pigment of peanut- skin, artificial neural network （ANN） model of extracting conditions based on the comprehensive values of peanut red pigment was founded. Method： Based on the training of artificial neuralnetworks using orthogonal arrays,a model for the comprehensive value of peanut red pigment as the output and the input consisting of technological peanut red pigment for extraction was developed and validated for reliability using selected specimens. The further optimization of optimal values of peanut red pigment obtained using orthogonal array design was conducted based on the AAN model by means of small-step search.Result：The optima} peanut red pigment for extraction were temperature 55 ℃, extracting agent was ethanol,60% ethanol concentration,extraction time was 3 h, material liquid ratio was 1：35（ g/mL）.Conclusion： ANN-based optimization gave an exact predicted value on the extracting conditions and the comprehensive values of peanut red pigment. The comprehensive values was 33.0824, higher than the value of 33.08 from orthogonal array optimization.
出处
《食品工业科技》
CAS
CSCD
北大核心
2016年第17期162-166,共5页
Science and Technology of Food Industry
基金
湖北省教育厅科研项目(B2015205)
关键词
花生红衣色素
正交实验
BP人工神经网络
peanut red pigment
orthogonal experiment
BP artificial neural network
