基于正交设计与人工神经网络模型的醋酸菌A3菌株醋酸发酵条件优化

Optimization of acetic acid fermentation condition of Acetobacter A3 based on orthogonal design coupled with artificial neural networks model

以甘肃传统食醋酿造过程中分离得到的1株优势产酸醋酸菌A3为出发菌株,采用正交设计与人工神经网络模型相结合的方法优化其发酵条件以提高产酸量。用正交实验中的因素水平组合作为ANN的输入变量,用产酸量作为输出变量,正交实验数据作为建立BP网络模型的训练样本,另外两组实验(17号、18号)用于检验模型的泛化能力,并运用建立好的人工神经网络模型在正交实验分析基础上寻找最佳发酵条件。实验结果表明,该菌株A3醋酸发酵最佳条件为:初始乙醇浓度4.2%(v/v)、发酵温度30℃、起始pH6.4、发酵周期8d,在此条件下的产酸量明显高于正交实验中的最高产酸量,达到4.3086g/100mL。

A bacterium strain A3 producing acetic acid was isolated from the traditional brewing process of Gansu vinegar.The optimization method based on orthogonal design coupled with artificial neural networks model（ANN） was employed to select fermentation conditions for increasing the yield of acetic acid.The combinations of factor and level in the orthogonal design were used as the input variables, and the content of acid as the output for building the ANN model.The sixteen tests data were used as BP architecture training samples,and two other test results（ No.17 and No.18）were used to examine the generalization capability of the model.Finally,the established ANN was applied to the optimization of fermentation conditions.The results showed that the optimized fermentation conditions for acetic acid production of this Acetobacter A3 were found to be 4.2% （v/v）alcohol,30℃, pH6.4 and 8d fermentation time.Under the conditions,the content of acetic acid was superior to orthogonal test,increasing to 4.3086g/100mL.