摘要
目前,金霉素发酵生产过程的补糖速率控制由人工操作完成,导致发酵过程的总糖浓度不稳定而影响发酵效率。为此,提出了一种基于培养基总糖浓度软测量的补糖速率优化调节新方法。通过发酵罐可测变量选取、数据的模糊化处理、递归模糊神经网络构建和网络滚动训练,建立总糖浓度在线软测量模型。并且在总糖浓度软测量的基础上,将发酵罐尾气中CO2含量的变化过程划分为4个控制阶段,依此对补糖速率实现优化调节。现场的试验结果表明该方法的实施有利于菌体生长和金霉素合成,提高了发酵效率并节约了生产成本。
The glucose feed rate of Chlortetracycline （CTC） fermentation process is currently controlled manually in bi- ochemical plant, which results in the instability of total sugar concentration and affects the fermentation efficiency of the fermentation process. Therefore,we present a new optimal adjustment method for the glucose feed rate based on the soft sensor of the total sugar concentration of fermentation broth. We establish an online soft sensor model for the total sugar concentration by means of selecting the measurable variables, blurring the data, constructing the recurrent fuzzy neural network and rolling training the network. On the base of the soft sensor for total sugar concentration, the changing process of the CO2 content in the fermentor exhaust gas is divided into four control phases,with which the optimal ad- justment of the glucose feed rate can be automatically accomplished. The field experiment results demonstrate that the method is beneficial to the cell growth and CTC synthesis,improves the fermentation efficiency and reduces the produc- tion cost.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2014年第2期468-474,共7页
Chinese Journal of Scientific Instrument
关键词
金霉素发酵
补糖速率
优化调节
软测量
Chlortetracycline （ CTC ） fermentation
glucose feed rate
optimal regulation
soft sensor
