基于熵权法结合层次分析法和反向传播神经网络优选酒萸肉蒸制工艺

Optimization of processing technology of wine-steamed Corni Fructus based on entropy method combined with analytic hierarchy process and BP neural network

目的优选酒萸肉蒸制最佳工艺。方法基于L_(9)(3^(4))正交试验,以没食子酸、5-羟甲基糠醛、莫诺苷、马钱苷、齐墩果酸、熊果酸、多糖、水溶性浸出物含量为指标,采用熵权法结合层次分析法(analytic hierarchy process,AHP)计算复合评分,考察黄酒量、闷润时间、蒸制时间对酒萸肉质量的影响,采用反向传播(back propagation,BP)神经网络寻求最佳工艺参数,并进行工艺验证。结果正交试验筛选的酒萸肉最佳蒸制工艺为黄酒量20%,即每100 kg山萸肉,用黄酒20 kg,闷润1 h,蒸制8 h,正交试验最高复合评分工艺为黄酒量20%,闷润4 h,蒸制8 h;BP神经网络模型优化的最佳炮制工艺为黄酒量20%,闷润4 h,蒸制7.5 h。BP神经网络优化工艺实际复合评分最高,为97.0078,与预测值96.3586较接近,正交试验筛选的最佳工艺实际复合评分排其次,BP神经网络优化工艺优于正交筛选工艺,且预测结果稳定可靠。酒萸肉最佳蒸制工艺为每100 kg山萸肉,用黄酒20 kg,闷润4 h,蒸制7.5 h。结论建立的酒萸肉炮制工艺稳定可行,得到的炮制品质量较佳,为酒萸肉工艺研究提供了依据。

Objective To optimize the best wine-steaming way of Corni Fructus.Methods Based on the L_(9)(3^(4)) orthogonal experiment,the contents of gallic acid,5-hydroxymethyl-2-furaldehyde,morroniside,loganin,oleanolic acid,ursolic acid,polysaccharides,and water-soluble extracts were used as indicators.Entropy method combined with analytic hierarchy process(AHP)was employed to calculate the comprehensive score and investigate the influence of the amount of rice wine,moistening time,and steaming time on the quality of Corni Fructus,the best wine-steaming process of Corni Fructus was optimized by BP neural network and finally,the processing verification was carried out.Results The best processing method of wine-steamed Corni Fructus selected by orthogonal experiment was adding 20%rice wine(using 20 kg of rice wine for every 100 kg of Corni Fructus),moistening for 1 h until the wine was exhausted,steaming for 8 h,taking it out and drying it.The highest composite score of orthogonal experiment was adding 20%rice wine,moistening for 4 h,and steaming for 8 h.The best processing technology optimized by BP neural network model was adding 20%rice wine,moistening for 4 h,and steaming for 7.5 h.The actual composite score(97.0078)of the BP neural network was the highest,which was close to the predicted value of 96.3586.The actual composite score of the best process screened by the orthogonal experiment ranked second.The BP neural network optimization process was better than the orthogonal screening process and the forecast result was stable and reliable.The best processing method of wine-steamed Corni Fructus was using 20 kg of rice wine for every 100 kg of Corni Fructus,moisturizing for 4 h,and steaming for 7.5 h.Conclusion The established processing method of wine-steamed Corni Fructus is stable and feasible,and the quality of processed products is good,which provides a basis for the processing research of wine-steamed Corni Fructus.