响应面法和神经网络优化Acinetobacter sp.DNS32发酵基质  

Optimization of fermentation medium for Acinetobacter sp.DNS32 by response surface methodology and artificial neural network

作　　者：王洋[1] 王志刚[1] 王溪[1] 郭火生[1] 孟冬芳[1] 张颖[1] 

机构地区：[1]东北农业大学资源与环境学院,哈尔滨150030

出　　处：《环境工程学报》2013年第2期791-795,共5页Chinese Journal of Environmental Engineering

基　　金：国家教育部新世纪优秀人才项目(NCET-10-0145);黑龙江省高校长江学者后备支持计划(2012CJHB001);"十二五"农村领域国家科技计划(2011BAD04B02-1);哈尔滨市科技创新人才研究专项基金(2012RFXXN013)

摘　　要：为了提高阿特拉津降解菌Acinetobacter sp.DNS32的产量,分别采用响应曲面法和基于人工神经网络的遗传算法对阿特拉津降解菌DNS32发酵培养基中3个重要基质成分(玉米粉、豆饼粉、K2HPO4)进行优化研究。响应曲面法确定3种成分的含量为玉米粉39.494 g/L,豆饼粉25.638 g/L和K2HPO43.265 g/L时,预测发酵活菌最大生物量为7.079×108CFU/mL,实测量为7.194×108CFU/mL;人工神经网络结合遗传算法优化确定3种主要成分含量为玉米粉为39.650 g/L,豆饼粉为25.500 g/L,K2HPO4为2.624 g/L时,预测最大值为7.199×108CFU/mL,实测量为7.244×108CFU/mL;最终确定培养基配方:玉米粉为39.650 g/L,豆饼粉为25.500 g/L,K2HPO4为2.624 g/L,CaCO3为3.000 g/L,MgSO4.7H2O和NaCl均为0.200 g/L;优化后阿特拉津降解菌DNS32发酵生物量比优化前提高了36.6%。结果表明,在阿特拉津降解菌DNS32发酵培养基组分优化方面,响应面法和基于人工神经网络的遗传算法都是可行的,基于人工神经网络的遗传算法具有更好的拟合度和预测准确度。The aim of this research was to increase the biomass production of atrazine-degrading Acinetobacter sp. DNS32 by adopting response surface methodology （RSM） and genetic algorithm based on artificial neural network （ANN-GA） to optimize the three important fermentation medium compositions, respectively. According to RSM, these three optimized compositions were composed as follows ： corn flour 39. 494 g/L, soybean flour 25. 638 g/L and KzHPQ 3. 265 g/L. The predicted and verifiable values by RSM were 7. 079× 10^8CFU/mL and 71 194 × 10^8CFU/mL, respectively. The maximum biomass concentration predicted by hybrid ANN-GA was 7. 199 × 10^8 CFU/mL at the optimum level of medium variables as follows： corn flour 39. 650 g/L, soybean flour 25. 500 g/ L and K2HPQ 2. 624 g/L, while the experimentally measured value was 7. 244 × 10^8CFU/mL. Finally, according to the above results, the optimized medium composition was ： corn flour 39. 650 g/L, soybean flour 25.50 g/L, CaCO3 3. 000 g/L, K2HPO4 2. 624 g/L, MgSO4 · 7H2O 0. 200 g/L and NaCl 0. 200 g/L. After medium optimization, the biomass yeild of atrazine-degrading strain DNS32 increased by 36.6% than that using non-optimized medium. The results showed that RSM and ANN-GA were feasible to optimize the fermentation medium for the production of atrazine-degrading strain DNS32, and ANN-GA had a much better optimizing ability and modeling ability.

关 键 词：阿特拉津降解菌DNS32 发酵培养基 响应面法 人工神经网络 遗传算法 优化 

分 类 号：X172[环境科学与工程—环境科学]