摘要
赖氨酸发酵过程是一个时变、非线性、强耦合多变量系统。为了有效的控制直接反映发酵品质的重要生化过程参数,如菌体浓度、残糖浓度、产物浓度等,实现高性能的解耦控制的目标,将逆系统方法与神经网络相结合,提出了一种基于神经网络逆系统的赖氨酸发酵过程解耦控制方法。在一定程度上解决了传统解析逆系统解耦控制方案过于依赖过程模型和对模型参数的变化过于敏感的不足。在实验中,以发酵罐和嵌入式开发系统为平台对控制方法进行了验证。实验结果表明该解耦控制方法能够对菌体浓度、残糖浓度、产物浓度等重要的生化参数进行有效的控制,适应过程模型的不确定性和参数的时变性,具有较强的鲁棒性。
The lysine fermentation is a system of multivariate,nonlinear and seriously coupled system which has more parameters.In order to effectively control those important parameters such as biomass,glucose concentration,lysine thickness that directly represents the fermentation quality and to achieve the goal of decouple.A method is come up with to decouple and control lysine fermentation process based on neural network inverse,by which inversion system method integrate the neural network.To some extend,this solution overcome the disadvantage of over-depending on the course module and is hypersensitive to parameters variation.In the experiment,this method is verified by the fermentor and embed system.Experimental results show that this method can control those parameters well and adopt to model uncertainty and time-varying parameters over lysine fermentation with strong robustness.
出处
《微计算机信息》
2012年第9期93-95,共3页
Control & Automation
基金
基金申请人:嵇晓辅
项目名称:基于神经网络逆微生物发酵解耦控制的研究
基金颁发部门:教育部高等学校博士学科专项科研基金
编号:20080299010
关键词
逆系统
神经网络
发酵过程
解耦控制
invertible systems
neural networks
fermentation
decoupling control
