文献类型：期刊文献

中文题名：基于AHP和BPNN结合模型优化当归烟熏加工工艺

英文题名：Optimization of smoked processing technology of Angelicae Sinensis Radix based on an AHP combined with BPNN model

作者：顾志荣[1];潘新波[1];王亚丽[1,2];孙宇靖[1]

第一作者：顾志荣

机构：[1]甘肃中医学院当归研究所,兰州730000;[2]甘肃中医学院科研实验中心,兰州730000

第一机构：甘肃中医药大学药学院（西北中藏药协同创新中心办公室）

年份：2014

卷号：23

期号：22

起止页码：2664

中文期刊名：中国新药杂志

外文期刊名：Chinese Journal of New Drugs

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD_E2013_2014】；

基金：国家自然科学基金(30960037);甘肃省发改委战略新兴产业和产业技术研究与开发专项项目(2011)

语种：中文

中文关键词：当归;层次分析;反向传播人工神经网络;烟熏加工;高效液相色谱

外文关键词：Angelicae Sinensis Radix; analytic hierarchy process; back-propagation neural network; smoked processing; HPLC

摘要：目的:建立不同烟熏条件加工当归的层次分析(AHP)和反向传播人工神经网络(BPNN)结合模型,优化当归烟熏加工工艺。方法:测定烟熏当归药材中5种指标性化学成分(阿魏酸、Z-藁本内酯、正丁基苯酞、正丁烯基苯酞、亚油酸)以及醇溶性浸出物含量,采用建立的AHP与BPNN结合模型优化当归烟熏加工工艺。结果:AHP模型优化工艺为50℃烟熏8 h,AHP与BPNN结合模型优化工艺为46.5℃烟熏8.8 h,后者工艺优于前者。结论:AHP与BPNN结合模型优化的当归烟熏工艺有利于减少能耗及保存有效成分,有一定实用价值。
Objective： To establish an analytic hierarchy process （AHP） combined with back-propagation neural network （BPNN） model for preparing smoked Angelicae Sinensis Radix by different processing conditions, and to optimize the smoked processing technology. Methods： Five index chemical compositions （ferulic acid, Z-li- gustilide, 3-butylphthalide, n-butylidenephthalide and linolic acid） and ethanol-soluble extracts from Angelicae Sinensis Radix were determined, and the established AHP combined with BPNN model was used for optimization of smoked processing technology. Results： Optimum smoked processing technology by the AHP model was smoking for 8 hours at 50 ℃ , and that by the AHP combined with BPNN model was smoking for 8.8 hours at 46.5℃. The latter was superior to the former. Conclusion： The optimum smoked processing technology by the AHP combined with BPNN model is practically valuable, which is helpful to reduce energy consumption and save effective compo- nents.