摘要
在轻工发酵过程中,酵母细胞浓度是一个重要的过程参数,但目前的检测手段多为离线测量,不能用于自动控制。发酵液的电容率与酵母细胞浓度之间存在某种关系。RBF神经网络在间接测量方面有着逼近能力强、训练方法简单易行等特点。结合温度、pH值和液位等易测参数,利用RBF神经网络来逼近这种关系,很好地解决了酵母细胞在线测量的问题。
In the process of cider production, the thickness of barm is a very important parameter,but there isn't a kind of measurements which can measure it on line. The traditionmeasurements aren't online and can't be applied in automatic control. There is someuncertain relation between the barm thickness and capacitive ration of solution of barm.The RBF nets have ability of simulation this relation. Combining with other generalparameters of the process, RBF nets can forecast the thickness well and truly.
出处
《制造业自动化》
2002年第6期44-47,共4页
Manufacturing Automation
