论文部分内容阅读
为了提高类球红细菌F3-40辅酶Q10的发酵效价,对发酵培养基进行优化。通过单因子优化实验确定培养基中4种重要成分的浓度范围:葡萄糖25～40g/L、味精5～9g/L、硫酸铵3～7g/L、玉米浆粉5～9g/L。在此基础上采用均匀设计实验对4种成分进行组合优化,分别采取二次多项式逐步回归分析法、人工神经网络与遗传算法耦合法对均匀设计实验结果进行优化分析。结果表明,人工神经网络与遗传算法耦合法取得较好的优化效果,显著提高辅酶Q10的发酵水平,最终辅酶Q10的发酵水平达到245mg/L,比二次多项式逐步回归分析优化法(221 mg/L)、单因子优化实验(211 mg/L)、优化前(150 mg/L)分别提高了10.86%,16.11%,63.33%。
In order to increase the fermentation potency of Rhodobacter sphaeroides F3-40 coenzyme Q10, the fermentation medium was optimized. The concentration of four important components in the medium was determined by single factor optimization experiment: 25-40 g / L glucose, 5-9 g / L MSG, 3-7 g / L ammonium sulfate and 5-9 g / L corn syrup. On this basis, the uniform design experiments were used to optimize and combine the four components. The quadratic polynomial stepwise regression analysis and the artificial neural network and genetic algorithm were used respectively to optimize the experimental results of uniform design. The results showed that artificial neural network (ANN) combined genetic algorithm (GA) could achieve a better optimization effect and significantly increase the fermentation level of coenzyme Q10. The final fermentation level of coenzyme Q10 reached 245 mg / L, which was higher than that of the second order polynomial stepwise regression analysis (221 mg / L), single factor optimization (211 mg / L), pre-optimization (150 mg / L) increased by 10.86%, 16.11% and 63.33% respectively.