秸秆水解液发酵产琥珀酸工艺的神经网络分析

Neural Network Analysis for Fermentation Production Technics of Succinic Acid from Crop Straw Hydrolysate

在基本培养基的基础上,通过单因素试验发现厌氧气室中的二氧化碳体积含量、氢气体积含量及培养基中的生物素浓度对发酵有影响。神经网络分析表明,当最佳条件为二氧化碳体积含量62%、氢气体积含量5.4%及生物素微量浓度7.8mmol/L时,发酵产酸最高为75.8g/L,神经网络具有较好的仿真及预测能力。

The results of single factor test showed that volume percentages of CO2 and HE as well as biotin concentration in basic medium have obvious effects on fermentation of crop straw hydrolysate for producing succinic acid. The neural network analysis showed that 75.80 g/L yield of succinic acid is gained under the conditions of 62% CO2, 5.4% HE and 7.8 mmol/L biotin. The neural network has the considerable simulation and prediction abihties.