摘要
该研究基于人工神经网络和遗传算法,对黑木耳红枣发酵乳发酵工艺进行优化,在此基础上分析其抗氧化特性。在单因素试验的基础上,通过Box-Behnken试验设计不同因素条件下黑木耳红枣发酵乳的酸度,以此构建人工神经网络优化模型,其模型拟合度达到0.991 12。结合遗传算法对黑木耳发酵乳进行127次迭代后,求得其最佳工艺条件为黑木耳红枣汁与鲜牛乳体积比3∶2.1、发酵温度38.8℃、发酵时间36.6 h、接种量6.70%,此条件下酸度最佳为75.12°T,试验预测值与实际值无显著差异,表明网络模型结果准确;此条件下,黑木耳红枣发酵乳的DPPH自由基、羟基自由基清除率比发酵前分别提高12.04%、13.52%。
In this study,based on artificial neural network and genetic algorithm,the fermentation process of the fermented milk of black fungus and red jujube was optimized,and its antioxidant properties were analyzed on this basis. On the basis of single factor,the acidity of the fermented milk of black fungus and red jujube under different conditions was designed by Box-Behnken experiment to construct an artificial neural network optimization model,and the model fitting degree reached 0.991 12. After 127 iterations of the black fungus fermented milk combined with the genetic algorithm,the optimal process conditions were obtained as follows.The ratio of the addition of black fungus and red jujube juice to the milk addition was 3 ∶ 2.1,and the fermentation temperature and time were 38.8 ℃ and 36.6 h,the inoculum amount was 6.70%,and the acidity was optimal at 75.12 ° T. Under these conditions,the predicted value of the test and the actual value had no significant difference,proving that the results of the network model were accurate. Under these conditions,1,1-diphenyl-2-picrylhydrazyl(DPPH)and hydroxyl radical scavenging rates of the fermented milk of black fungus and red jujube were 12.04% and 13.52%,respectively,higher than those before fermentation.
作者
张剑林
张亮亮
姜露熙
裴龙英
王寒博
孙博
梁睿武
房丹丹
ZHANG Jian-lin;ZHANG Liang-liang;JIANG Lu-xi;PEI Long-ying;WANG Han-bo;SUN Bo;LIANG Rui-wu;FANG Dan-dan(College of Food Science and Engineering,Xinjiang University of Technology,Aksu 843100,Xinjiang,China)
出处
《食品研究与开发》
CAS
北大核心
2023年第5期141-147,共7页
Food Research and Development
基金
校级科研项目(ZQ202207)
自治区教育厅高校科研计划项目(XJEDU2020Y047)
国家级大学生创新创业项目(202113558006)
校级平台项目(PT202101)。
