抗生素生产过程补料时机的Fuzzy识别与预报

本文就抗生素发酵生产过程,提出了一种补料时机的Fuzzy识别、预报的建模方法,并根据20m~3发酵罐工业数据的分析,选取了以总糖、总糖变化率、氨基氮和氨基氮变化率四个主要参数作为输入集因素,运用本文提出的方法,建立了一个实际的抗生素发酵生产过程补料时机的 Fuzzy 识别、预报模型。经现场调试效果良好,与富有经验的行家相比,准确率达90%以上。

FUZZY RECOGNITION AND PREDICTION OF FEEDING TIME FOR AN ANTIBIOTIC PRODUCTION PROCESS

A method of modelling for fuzzy recognition and prediction of feeding time is proposed for antibiotic fermentation processes.By using this method and in accordancewith the analysis of an industrial fermentation process information,four main para-meters,i.e.,sugar concentration,change rate of sugar concentration,amino nitrogen conc-entration and change rate of amino nitrogen,are selected as input set factors,and a modelof fuzzy recognition and prediction of feeding time is developed.The validity of thismodel will be examined by comparing the results of computer simulation with thosefrom experienced expert.It is found that a good dicision will be given by the model.