COMPUTER CONTROL OF GLUTAMIC-ACID PRODUCTION BASED ON FUZZY CLUSTERIZATION OF CULTURE PHASES

An algorithm for the clusterization of culture status into several culture phases (lag, growth, transient, and production) was developed for the flexible control of glucose concentration in glutamic acid fermentation. Each phase has its own different type of control strategy. Phase recognition was carried out based on fuzzy sets theory. For the recognition, three measurement variables (culture time, CO2 evolution rate, and the amount of NH3 feed) were selected as variables of fuzzy antecedent, and several fuzzy levels for each variable were defined by a table of membership functions. The inference was carried out by a min-max fuzzy algorithm every three minutes, and the tubing pump for glucose feed was controlled to let the amount of feed follow the inferred amount. The control strategy was able to transit smoothly from one phase to another during the cultivation and the glucose concentration was successfully kept constant.