Research On The Embedded Control System Of Fed-Batch Culture In The Production Of Monosodium Glutamate

This topic,research on the embedded control system of fed-batch culture in the production of monosodium glutamate,is a part of the project,an annual output of100000tons of monosodium glutamate clean production technology integration (application) demonstration project which is supported by national science and technology.Glutamic acid fermentation is a typical multi variable,strong coupling,nonlinear industrial process.All the time since,for fermentation production which plays an important role in the process of feeding quantity control is rely on operator’s experience and knowledge.This management mode should be influenced of operator ability,quality and professional knowledge.So properties of feeding optimization and control is different from man to man,and the difference is very big.In recent years,artificial intelligence technology and fuzzy reasoning technology based on expert knowledge and computer science has been applied in many biological processes.This paper presents a method based on the embedded control system of fermentation control system,combined with artificial neural network to establish the process of glutamic acid fermentation fed-batch prediction model,realized the fed-batch control of the fermentation of glutamic acid.According to the complexity and nonlinearity of the glutamate fermentation,using artificial neural network establishes dynamic mathematical model for the feed.In embedded control system, the TQ2440ARM development board served as the core control module,based on embedded QT4platform,establishes the prediction model of fermentation process.The model for the amount transmits to execute module (51SCM system) through wireless digital module CC1100or serial port.SCM system completes the feeding timing and feeding peristaltic pump driving,thereby realizing the fed-batch control.The distance of wireless data transmission is restricted,serial transmission need to site wiring,so the control system provides wireless and serial two ways to transmit data,.it is Convenient for user chosen.Single core control board can control multiple front-end execution module.Extended control scale (increasing control module) simply modify the core control board procedures.The whole system is very easy to expand.The experiment shows that this model can estimate the amount of fermentation for glutamic acid accurately,and the control system transmits data stably.