中文摘要：

捅要：为提高丁酸梭菌（Clostridium but yricum）VPI 1718产1，3-丙二醇的能力，采用均匀设计试验和人工神经网络结合遗传算法对培养基进行优化，得到最优配方（g／L）：废甘油70，酵母粉2．0，KH2PO4 0．6，K：HPO41．2，（NH4）2SO4 2.6，MgSO4 0.02；使用该培养基在5L发酵罐中分批培养14h丁酸梭菌后，1，3-丙二醇质量浓度为33.0g／L，转化率为55．3％，生产强度为2．36g／（L·h），较优化前分别提高了32．5％，7．17％和33．3％．关键词：丁酸梭菌；1，3-丙二醇；均匀设计；人工神经网络；遗传算法中图分类号：TQ92文献标志码：A文章编号：0438—0479（2012）05一0893一05

英文摘要：

In the present work, uniform design and artificial neural network linked genetic algorithm were applied to optimize 1 , 3- propanediol fermentation medium to enhance 1 , 3- propanediol production by Clostridium butyricum VPI 1 71 8. The composi- tions of the optimized fermentation medium were as follows （g/L）： raw glycerol 70, yeast extract 2. 0, K2HPO4 1.2, KH2 PO4 0. 6 , （ NH4 ）2 SO4 2. 6 , Mg SO4 0. 2. When the C. butyricum was cultivated in the optimized medium for 1 4 h on a 5 L fermentor,the 1,3 propanediol concentration, yield and productivity were 33. 0 g/L,55. 3% and 2. 36 g/（L·h）, which were 32. 5%, 7. 17% and 35. 6% highter than that in the initial medium, respectively.