BP神经网络结合正交试验法优选锦锈杜鹃黄酮的提取工艺

Optimization of Extraction Process of Flavonoids in Rhododendron Pulchrum Sweet. Based on Artificial

摘要: 将BP人工神经网络技术与传统的正交试验方法相结合,提出一种新的试验分析和处理方法,利用神经网络特有的自学能力,对主要影响因素进行仿真优化,获得锦锈杜鹃黄酮物质提取的最佳工艺:料液比1∶20(g/mL),超声波提取时间为40 m in,提取温度为60℃,乙醇质量分数为40%.应用优化得到的工艺条件进行验证试验,取得了较好的效果

Abstract: Artificial neural network combined with traditional orthogonal design constitute the basis on which a new method of test data analyzing and processing was presented.With the help of artificial neural network,through emulation and optimization,the optimal conditions for extraction of total flavonoids in Rhododendron pulchrum Sweet.was obtained as follows: the solid-liquid ratio of 1∶20(g/mL),40 min of extraction time,60 ℃ of extraction temperature and 40% of ethanol concentration.The optimal tech..