摘要
发酵培养基配比的优化是含有多变量非线性目标函数的优化。以青霉素发酵为例,通过考察青霉素效价产率的大小,用MATLAB统计工具箱的“regress”函数对培养基配比进行多元非线性回归,以含有约束的回归方程作为目标函数,调用MATLAB优化工具箱的“fmincon”函数,对其进行优化,确定了适宜的培养基配比。将优化条件得到的预测值与实测值比较,当培养基配比为乳糖8．0％、花生饼粉4．0％及苯乙酰胺0．213％时,目标产物的效价产率为1226．0,与优化结果1210．9比较接近。
A regress equation, which relates the potency based on penicillin＇s yield to the Concentrations of culture medium, was generated. This regression model was obtained by fitting the experimental data to full second-order polynomial model using the regress function in MATLAB＇s Statistics Toolbox. Optimal concentrations of the medium components within the range tested for the maxmal potency were obtained by numerical techniques using the fmincon function in MATLAB＇s Optimization Toolbox. In order to test this model the penicillin were produced in the optimum medium for the potency i. e. 8.0% lactose, 4.0% peanut cake powder, 0.213% phenylαce. The result obtained was compared to the corresponding prediction; the potency of 1226.0 was close to the predicted value of 1210.9.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2006年第7期627-630,共4页
Computers and Applied Chemistry
关键词
青霉素
培养基
优化
MATLAB
penicillin, culture medium, optimization, MATLAB
