遗传神经网络优化牛蒡固体发酵灵芝提取多糖工艺  

Optimization of Polysaccharide Extraction From Ganoderma lucidum of Solid-state Fermentation in Burdock by Genetic Neural Network

作　　者：朱会霞[1] 董玉玮[2] ZHU Huixia;DONG Yuwei(College of Management,Liaoning University of Technology,Jinzhou,Liaoning 121001;College of Food(Biology)Enginccring,Xuzhou Institute of Technology,Xuzhou,Jiangsu 221018)

机构地区：[1]辽宁工业大学管理学院,辽宁锦州121001 [2]徐州工程学院食品(生物)工程学院,江苏徐州221018

出　　处：《北方园艺》2020年第22期103-108,共6页Northern Horticulture

基　　金：辽宁省自然科学基金资助项目(2019-ZD-0804);辽宁省教育厅高校基本科研资助项目(JQW201715407)。

摘　　要：以牛蒡和灵芝为试材,采用一种新的自适应遗传神经网络方法,研究了牛蒡固体发酵灵芝提取多糖最优工艺问题。用自适应遗传神经网络算法对多糖含量的试验值进行了拟合,并与回归分析方法进行了比较。结果表明:自适应遗传神经网络算法比回归分析方法具有更高的预测和优化能力。该方法得到牛蒡固体发酵灵芝提取多糖最佳工艺为液固比0.32 mL·g^-1,装瓶量0.18 g·mL^-1,粉碎程度6.19目筛。该条件下,多糖含量最大值为25.13 mg·g^-1,优于回归方法得到的多糖含量25.07 mg·g^-1。Burdock and Ganoderma lucidum were used as test materials,a new adaptive genetic neural network algorithm was used to study the optimal process of extracting polysaccharides from Ganoderma lucidum of solid-state fermentation in burdock.The test data of polysaccharide content were used to fit the predictive data of the adaptive genetic neural network algorithm,and the predictive data of the adaptive genetic neural network algorithm were compared with that of the regression analysis method.The results showed that prediction and optimization of the adaptive genetic neural network algorithm was higher than that of the regression analysis method.The optimal theoretical parameters for the polysaccharides extracting fromGanoderma lucidum of solid-state fermentation in burdock by this method was follow,liquid-solid ratio 0.32 mL·g^-1,the filling volume 0.18 g·mL^-1,the degree of crushing 6.19 mesh sieve,and the maximum of polysaccharide content was25.13 mg·g^-1,which was better than the polysaccharide content of 25.07 mg·g^-1 obtained by the regression method.

关 键 词：自适应遗传神经网络 牛蒡 灵芝 多糖 回归分析 

分 类 号：S646.9[农业科学—蔬菜学]