论文部分内容阅读
目的:运用R语言结合BP神经网络和遗传算法优化复方养阴通脑颗粒中总黄酮提取工艺。方法:在单因素试验的基础上采用4因素3水平的正交试验设计方法,提取并测得养阴通脑颗粒中总黄酮的含量。通过R语言,建立BP神经网络模型,再利用遗传算法对网络进行目标寻优,从而得到养阴通脑颗粒总黄酮的最佳提取工艺。结果:养阴通脑颗粒中总黄酮的最优工艺条件为乙醇浓度55%、提取时间1.5h、提取温度80℃、液料比15∶1,模型预测值为119.04mg,而实验平均值为120.35mg,相对误差为1.09%,因此具有良好的网络预测性。结论:此数学模型可用来对养阴通脑颗粒中总黄酮的提取工艺进行分析和预测,为实现中药有效部位和成分目标寻优提供了新的参考。
OBJECTIVE: To optimize the extraction of total flavonoids from Compound Yangyin Tongnao Granules by using R language combined with BP neural network and genetic algorithm. Methods: Based on the single factor test, orthogonal design of 4 factors and 3 levels was used to extract and measure the content of total flavonoids in Yangyintongnao granules. Through R language, the BP neural network model is established, and then the target of the network is optimized by using genetic algorithm to obtain the optimal extraction process of the total flavonoids of Yangyintongnao granules. Results: The optimum conditions of total flavonoids in Yangyintongnao Granules were ethanol concentration 55%, extraction time 1.5h, extraction temperature 80 ℃, liquid ratio 15:1, model predictive value 119.04mg, while the experimental average was 120.35mg, the relative error is 1.09%, so it has a good network predictability. Conclusion: This mathematical model can be used to analyze and predict the extraction technology of total flavonoids in Yangyintongnao granules, and provide a new reference for the optimization of the effective parts and components of traditional Chinese medicine.