补阳还五汤提取工艺的响应面法和人工神经网络模型优化

Extraction Process Optimization for Buyanghuanwu Decoction by Response Surface Methodology and Back Propagation Artificial Neural Network

目的利用响应面法和人工神经网络法优选补阳还五汤提取工艺,并对这两种方法的优化结果进行比较。方法以黄芪甲苷含量、芍药苷含量、总固体得率为指标,单因素试验考察加水量、提取时间、提取次数对补阳还五汤提取效果的影响,运用响应面法优化补阳优化还五汤提取工艺,在响应面法数据的基础上通过构建BP神经网络模型,对提取工艺进行进一步优化,最终确定最优提取工艺。结果响应面法优选补阳还五汤最佳提取工艺为:药材加8.38倍量52%浓度的乙醇提取2次,每次1.7h;BP神经网络模型仿真优化的提取工艺为:药材加8.45倍量54%浓度的乙醇提取2次,每次1.5h。BP神经网络优选工艺的综合评分较高,且相对误差较小。结论 BP神经网络模型法仿真优化的提取工艺具有提取效率高、能耗低的优点,适合工业化大生产。

Objective To optimize the extraction process of Buyanghuanwu Decoction by Response Surface Methodology(RSM)and Back Propagation Artificial Neural Network(BP ANN).Methods With content of Astragaloside Paeoniflorin and extractum yield as indexes,to investigate the effect of ethanol concentration,the volume of ethanol,extraction time,extraction times by RSM and BP ANN.Results The extraction process was optimized by RSM as following:Buyanghuanwu Decoction was added with8.38 times 52%ethanol,decocted for 1.7 h for 2 times.Optimum extraction pross of BP ANN model was as following:Buyanghuanwu Decoction was added with8.45 times 52%ethanol,decocted for 1.5 h for 2 times.Conclusion BP ANN modle with RSM can be used to optimize the extraction process of Buyanghuanwu Decoction with short time,low cost and low energy consuming.