Research And Realization Of Intelligent Monitor And Control System For Lincomycin Fermentation Process

China is one of the largest producers of antibiotics. There is a great marketdemand for lincomycin which is widely used as a kind of antibiotic. The mechanismof biochemical reactions in fermentation process is complex and some biologicalparameters are difficult to take the online measurement in real time. Based on theabove technology status, the level for detection and control and the degree ofautomation have not reached a high position. In order to improve the fermentationrate and product quality, reduce the consumption of raw material and energy, increasemarket competitiveness for enterprises, establishing an intelligent monitor and controlsystem for lincomycin fermentation process is urgently needed. Meanwhile, with thedemand of real-time predictive diagnostics for fermentation state, abnormalitydiagnosis subsystem for fermentation process which is combined with expert systemmethod in the field of artificial intelligence is established. By utilizing the system, theusers are assisted to diagnose abnormal conditions and deal with fermentationproblems.Firstly, phase characteristics for lincomycin fed-batch fermentation process wereanalyzed in depth and the key state parameters of fermentation process were collectedand analyzed. Then the collection methods and control strategies for biochemicalparameters were determined, and fermentation parameters which were difficult totake online measurement directly were calculated online by the soft measurementtechnology.After the analyses of fermentation process characteristics, the composition andstructure of expert system were studied. The expertise was obtained and knowledgebase of lincomycin fermentation was established. Then the reasoning policies forinference engine were set and human machine interface for experts and users wasdesigned. Through the above researches, the prediction and diagnostics ofabnormalities for lincomycin fermentation process were achieved. To further enrichthe rules of expertise in knowledge base, dynamics modeling for biochemicalreactions in fermentation process was studied and model-based knowledge rules were built. The forecast of fermentation process status and trends was improved by themodel-based knowledge rules.At last, aiming at the features for management of fermentation process, thedesign and implementation of the fermentation process monitor and control systemwas introduced in detail. The fermentation tank temperature, pressure, sterile air flow,pH values, dissolved oxygen concentration in the fermentation broth and some otherparameters were treated as the main monitoring parameters. The environmentalparameters on-site were collected by using sensor technology. By Siemens PLC, thereal-time data were processed and control signals were exported. Workstationprograms were developed through the Delphi development platform. Modbusprotocol was used to achieve the real-time communication between PLC andworkstation computers, and then the collection of real-time data and the configurationof control parameters were realized.