摘要
提出了一种利用TSK(Takagi-Sugeno-Kang)模糊系统建立生化变量预估模型的方法,用于生化过程的工艺参数预测。利用TSK模糊系统的非线性逼近能力,以谷氨酸发酵过程为研究对象,建立了基于TSK模糊逻辑系统的生化变量预估器。以工厂现场采集的发酵过程参数大量数据为样本,训练TSK模糊系统生化变量预估器,并对模型的预估精度进行检验。仿真结果显示了预估模型的有效性,可以有效地预测谷氨酸发酵过程中生化变量的估计值。TSK模糊系统生化变量预估模型能预估生化过程中工业发酵罐的放罐时间,预估模型的状态预报对正常罐批具有足够高的预报精度,预报误差如果偏大亦可作为异常罐批的早期警示信息。有鉴于此,TSK模糊逻辑系统可望开辟生化过程参数预估的新途径。
A novel estimation model of biochemical process has been presented based on TSK fuzzy system. It has been testified through simulation that the TSK is able to predict the biochemical variables precisely; the estimation mode has a good predict capability because of the strong nonlinear learning ability of TSK fuzzy system. The good predicting results were obtained. In fact, a new tool for estimation, prediction in biochemical production process has been provided.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2006年第4期343-346,共4页
Computers and Applied Chemistry
基金
荣毅仁教育基金资助项目(2002-0320)
关键词
模糊系统
TSK
生化过程
预估模型
fuzzy system, TSK, biochemical process, estimation model
