反向传播人工神经网络结合熵权法多指标优化当归身的提取工艺  

Multi-index optimization of extraction process of the body of Angelica sinensis(Oliv.)Diels by BP neural network combined with entropy weight method

作　　者：徐志伟[1] 王宝才[1] 旦孝三 毕映燕[1] 李季文[1] 边娜 杜伟锋[3] XU Zhi-wei;WANG Bao-cai;DAN Xiao-san;BI Ying-yan;LI Ji-wen;BIAN Na;DU Wei-feng(Gansu Provincial Hospital of TCM,Lanzhou 730050,China;Lanzhou City Chengguan District,Baiyin Road Street Community Sanitary Service Center,Lanzhou 730030,China;Research Center of Processing Technology for Chinese Materia Medica,Zhejiang Chinese Medical University,Hangzhou 311401,China)

机构地区：[1]甘肃省中医院,兰州730050 [2]兰州市城关区白银路街道社区卫生服务中心,兰州730030 [3]浙江中医药大学中药炮制技术研究中心,杭州311401

出　　处：《药物分析杂志》2023年第2期341-347,共7页Chinese Journal of Pharmaceutical Analysis

基　　金：国家重点研发计划-中药饮片质量识别关键技术研究(2018YFC1707001);国家中药标准化项目(ZYBZH-H-ZY-45);国家中医药行业科研专项(201507002);中华中医药学会青年人才托举工程项目(QNRC2-C12)。

摘　　要：目的:采用反向传播人工神经网络(BP-ANN)结合熵权法多指标优化当归身的提取工艺。方法:以提取温度、加液量、提取时间为考察因素,每个因素3个水平,开展L9(34),正交设计;运用熵权法计算阿魏酸、绿原酸、欧前胡素、藁本内酯百分含量及浸出物含量的综合评分作为评价指标;再以上述评价指标为数据基础建立BP-ANN模型,通过网络训练,预测当归身的最优提取工艺。结果:优化得到的当归身最优提取工艺为加入药材重量的12倍量体积提取液,在87℃条件下提取80 min,检测样本的网络预测值和实际测量值的相对误差为0.7764%。结论:建立的数学模型可对当归身提取工艺进行分析和预测,所得工艺稳定可行,可高效提取当归身的有效成分。Objective:To optimize the extraction process of the body of Angelica sinensis(Oliv.)Diels BP neural network combined with orthogonal experiment.Methods:The extraction temperature,the liquid amount and the extraction time were taken as factors.Entropy weight method was used to calculate the comprehensive scores of the multi-indicators of the content and four active components of chlorogenic acid,ferulic acid,imperatorin and butenylphthalide.using comprehensive score as an evaluation indicator.The BP neural network model was established by orthogonal experiment design,and the optimal extraction process of the body of Angelica sinensis(Oliv.)Diels was predicted through network training.Results:The optimized extraction process of the body of Angelicasinensis(Oliv.)Diels was carried out by adding 12 times of 70%methanol,extracting 80 minute at 87℃.The relative error between the network predicted value and the actual measured value of the test sample was 0.7764%.Conclusion:The established mathematical model can analyze and predict the extraction process of the body of Angelica sinensis(Oliv.)Diels.The obtained process is stable and feasible,and can effectively extract the active ingredients in the body of Angelica sinensis(Oliv.)Diels.

关 键 词：BP-ANN 熵权法 当归身 多指标 阿魏酸 绿原酸 欧前胡素 藁本内酯 浸出物含量 

分 类 号：R917[医药卫生—药物分析学]