论文部分内容阅读
目的结合正交试验与BP神经网络,筛选木瓜蛋白酶提取鱼腥草多糖的最佳工艺。方法以提取温度、提取时间及酶添加量为输入量,综合评分为输出量,用训练好的BP神经网络模型进行预测,寻找最优参数组合。结果由BP神经网络得到的最佳提取工艺为:pH=7.0,提取温度60℃,酶添加量2.0%,提取时间2.5 h,经试验验证其综合评分高于正交试验的最优组合的得分。结论 BP神经网络与正交试验相结合,可用于木瓜蛋白酶法提取鱼腥草多糖最优提取工艺的筛选。
OBJECTIVE To screen the optimal extraction process of polysaccharide from Houttuynia cordata with papain by combining orthogonal experiment with BP neural network. Methods Taking the extraction temperature, the extraction time and the amount of enzyme as the input, the comprehensive score was output, and the trained BP neural network model was used to predict the best parameter combination. Results The optimum extraction conditions obtained by BP neural network were as follows: pH = 7.0, extraction temperature 60 ℃, enzyme addition 2.0%, extraction time 2.5 h, the result of experiment showed that the comprehensive score was higher than the optimal combination of orthogonal test . Conclusion The combination of BP neural network and orthogonal test can be used to extract the optimal extraction process of polysaccharide from Houttuynia cordata by papain.