基于BP神经网络与遗传算法的苏云金芽胞杆菌发酵优化

Using BP Neural Network and Genetic Algorithm to Optimize the Fermentation Medium of Bacillus thuringiensis

本文应用BP神经网络与遗传算法优化苏云金芽胞杆菌BRC-ZQL3的发酵培养基,获得最佳发酵培养基配方为:酵母膏0.3928%,玉米粉0.4928%,黄豆饼粉2%,MgSO4·7H2O0.035%,KH2PO40.025%,CaCO30.0496%;用其培养的菌株产孢量为6.3×108/mL,与响应优化的结果相比提高了7.7%。

BP neural network and genetic algorithms methods were used to optimize the fermentation medium of Bacillus thuringiensis. The best fermentation medium is 0. 392 8% yeast extract, 0. 492 8% corn mea1, 2% defatted peanut flour, 0. 035% MgSO4·7H2O, 0.025% KH2PO4, 0. 049 6% CaCO3. The spore production was 6.3 × 10^8/mL under this condition, which was 7.7% higher than that by RSM method.