青霉素发酵过程中溶解氧的变结构预估控制

来源期刊：中南大学学报(自然科学版)2003年第4期

论文作者：于乃功 阮晓钢

文章页码：363 - 367

关键词：模糊神经网络;变结构预估控制;青霉素;发酵;溶解氧;控制

Key words：fuzzy neural network; variable structure predictive control; penicillin; fermentation; dissolved oxygen; control

摘    要：溶解氧是青霉素生产发酵过程中的重要参数之一,其浓度与菌体细胞的生长速度密切相关.在青霉素发酵过程溶解氧变化机理的基础上,将模糊神经网络和变结构预估控制方法相结合,研究了基于模糊神经网络的青霉素发酵过程溶解氧的变结构预估控制方法,设计了青霉素发酵过程溶解氧的变结构预估控制系统.实验结果表明,对青霉素发酵过程中的溶解氧浓度进行变结构预估控制,可以使溶解氧浓度的调节达到满意的效果,既降低氧消耗,又避免出现氧抑制现象,使发酵过程始终处于较优的状态.

Abstract: Dissolved oxygen is one of the important parameters in penicillin fermentation process, which influences the growth metabolism of penicillin biomass cells directly. Combining fuzzy neural network with the variable structure predictive control method, this paper designed a variable structure predictive control system of dissolved oxygen in penicillin fermentation process based on mechanism of penicillin fermentation process. This system was tested in a factory′s penicillin fermentation process, and the experimental results show that using this method to control dissolved oxygen of penicillin fermentation process can make dissolved oxygen keep a good value, and not only reduce oxygen consumption but also avoid oxygen restraint.