微生物发酵乙醇浓度在线检测模型及其应用  

Modeling and implementation of on-line measurement of ethanol concentration in animalcule fermentation

作　　者：赵利强[1] 王建林[1] 于涛[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《北京化工大学学报（自然科学版）》2012年第1期98-102,共5页Journal of Beijing University of Chemical Technology(Natural Science Edition)

基　　金：中央高校基本科研业务费项目(ZY1110)

摘　　要：利用半导体气敏传感器,结合气液平衡关系原理,提出了一种微生物发酵乙醇浓度在线检测模型的建立方法。该方法利用LM(Levenberg-Marquardt)算法对模型进行参数辨识,建立了液体乙醇浓度和测量电压信号两个物理量之间直观的非线性对数模型,并采用虚拟仪器技术实现了该模型算法。将模型应用于谷胱甘肽发酵中乙醇浓度的检测,实验与分析表明了该模型的有效性,检测绝对误差小于0.1%,重复性好,能够很好地满足微生物发酵工程应用需求。A method of establishing an on-line measurement model for animalcule fermentation has been proposed based on the logarithmic model and the vapor-liquid equilibrium principle.The Levenberg-Marquardt algorithm was adopted for identification of the model parameters.An intuitive on-line measurement model was established between the liquid ethanol concentration and the measured voltage signal and the algorithm was realized using virtual instrument technique.The model was employed in the detection of ethanol concentration during glutathione fermentation.The experimental results indicated the high efficacy of the model algorithm as its absolute error was less than 0.1%.The model has good reproducibility and can also meet the demands required for application in animalcule fermentation engineering.

关 键 词：乙醇浓度 在线测量 系统辨识 LM算法 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]