摘要
针对由于缺乏可靠的生物传感器,谷氨酸发酵中重要的生物参数—菌体浓度不能在线测量,更不能对其实现控制的情况,提出了基于径向基神经网络的内模控制。既解决了由于谷氨酸发酵内部机理复杂而难以建立菌体浓度模型的难题,又实现了谷氨酸菌体浓度的内模控制;同时也解决了内模控制中逆模型建模的问题。仿真结果表明,该方法实现了谷氨酸菌体浓度的有效控制,鲁棒性强、抗干扰能力好,具有良好的实际应用和推广价值。
The biomass concentration is an important biologic parameter that affects glutamic fermentation process,but it can not be directly detected on-line and real-time due to lacking of effective biologic sensor,so it can not be controlled.Internal model control(IMC)based on RBF neural network is proposed.It eliminates the difficult biomass concentration modeling problem owing to intricate inner mechanism of glutamic fermentation process and achieves IMC for biomass concentration of glutamic,and also solves inverse model modeling problem in IMC.Simulation results show this method realizes effective control for biomass concentration of glutamic,and possesses strong robustness and preferable stiffness against load disturbance.The method has better practical application and generalization values.
出处
《控制工程》
CSCD
北大核心
2009年第S2期18-20,23,共4页
Control Engineering of China
基金
辽宁省教育厅科研基金资助项目(051346)
关键词
内模控制
径向基神经网络
菌体浓度
仿真
谷氨酸发酵
IMC
RBF internal network
biomass concentration
simulation
glutamic fermentation