基于改进熵权法结合TOPSIS模型和BPNN建模优化甘草多糖醇沉工艺  

Optimization of alcohol precipitation process for Glycyrrhiza uralensis polysaccharides based on improved entropy weight method combined with TOPSIS model and BPNN model

作　　者：何宝峰 马新换 马忠祥[3] 王宝才[3] 徐志伟[3] 马芳兄 杨天艳 李荣焜 HE Bao-feng;MA Xin-huan;MA Zhong-xiang;WANG Bao-cai;XU Zhi-wei;MA Fang-xiong;YANG Tian-yan;LI Rong-kun(Gansu University of Chinese Medicine;Second Provincial People′s Hospital of Gansu,Lanzhou 730000,China;Gansu Provincial Hospital of Traditional Chinese Medicine,Lanzhou 730050,China)

机构地区：[1]甘肃中医药大学 [2]甘肃省第二人民医院,兰州730000 [3]甘肃省中医院,兰州730050

出　　处：《天然产物研究与开发》2023年第2期271-280,230,共11页Natural Product Research and Development

基　　金：兰州市科技计划(2020-XG-17)。

摘　　要：基于改进熵权法结合TOPSIS模型和BPNN建模寻求甘草多糖最佳醇沉工艺。在单因素试验基础上采用正交设计,以浓缩比、乙醇体积分数、醇沉时间为影响因素,甘草总糖、单糖、多糖含量及提取量为评价指标,利用改进熵权法确定评价指标权重,分别采用TOPSIS法和综合评分法处理正交结果,结果显示,TOPSIS法所得数据误差更小,最佳醇沉工艺为浓缩比为2.5 mL/g,乙醇体积分数为70%,醇沉时间为20 h。选取正交设计所得综合评分利用BPNN建模进行仿真寻优,得到最佳醇沉工艺为水提液浓缩至2.0 mL/g,调节乙醇体积分数为67%,醇沉24 h。所得最优工艺验证结果表明,BPNN建模所得综合评分误差更小,工艺更加稳定。该方法优选的最佳醇沉工艺稳定可行,可为甘草多糖进一步开发利用提供客观依据和新思路。Based on the improved entropy weight method combined with TOPSIS model and BPNN modeling,to seek the optimum alcohol precipitation process of Glycyrrhiza uralensis polysaccharides.On the basis of single factor experiment,with concentration ratio,ethanol volume fraction and alcohol precipitation time as influencing factors,saccharides,total polysaccharide,polysaccharide content and extracted amount of G.uralensis polysaccharides as evaluation indexes,using orthogonal design to determine the weight of evaluation index by improved entropy weight method.TOPSIS method and comprehensive scoring method were used to deal with the orthogonal results.The results showed that the data error obtained by TOPSIS method was smaller.The optimum alcohol precipitation process was as follows:concentration ratio was 2.5 mL/g,ethanol volume fraction was 70%,and alcohol precipitation time was 20 h.The comprehensive score obtained by orthogonal design was simulated and optimized by BPNN modeling,the optimal alcohol precipitation process was as follows:the water extract was concentrated to 2 mL/g,the volume fraction of ethanol was adjusted to 67%,and the alcohol precipitation was 24 h.The optimal process verification results show that the comprehensive score error obtained by BPNN modeling was smaller and the process was more stable.The optimal alcohol precipitation process optimized by this method was stable and feasible,which could provide objective basis and new ideas for the further development and utilization of G.uralensis polysaccharides.

关 键 词：甘草多糖 正交设计 改进熵权法 TOPSIS模型 BP神经网络 

分 类 号：R284.2[医药卫生—中药学] R932[医药卫生—中医学]