BP人工神经网络结合正交试验优化解毒止血颗粒的提取工艺

目的:优选解毒止血颗粒的提取工艺。方法:以黄芩苷、汉黄芩苷、丹皮酚含量以及干膏收率为评价指标,采用正交试验联合BP人工神经网络考察加水量、提取时间、提取次数等因素的影响,利用层次分析法(AHP)综合加权评分筛选解毒止血颗粒提取工艺。结果:最佳提取工艺参数为11倍量水,加热回流提取3次,105 min/次。结论:优选的提取工艺科学合理,稳定可行。正交试验联合BP人工神经网络优化方法实用高效,适宜于医疗机构中药制剂研发需求。

Optimization of Extraction Process of Jiedu Zhixue Granule by BP Artificial Neural Network and Orthogonal Test

To optimize the extraction process of Jiedu Zhixue Granules. Methods:Taking the baicaline, wogonoside, paeonol content and dry extract yield as the evaluation index, the extraction process of Jiedu Zhixue Granule was researched by using the combination methods between orthogonal test and BP artificial neural network to study the influence of water addition, extraction time, extraction times etc., and using the analytic hierarchy process (AHP) to weighted mark comprehensively for screening extraction process of Jiedu Zhixue Granule. Results:The optimum extraction process was as following: adding 11 times the amount of water, heating reflux extraction 3 times, and each time for 105 minutes. Conclusion:The optimal extraction process is scientific, rational, stable and feasible. The optimization method of orthogonal experiment combined with BP artificial neural network is suitable for R&D needs of TCM preparation in medical institution because of its practical and efficient characteristics.