摘要
在电解铝的生产过程中,针对氧化铝浓度不能在线测量的难题,采用了软测量技术来估计氧化铝浓度;运用径向基函数(RBF)神经网络构建了软测量的数学模型,利用神经网络的非线性逼近能力和自学习功能来适应不断变化的槽况,从而较准确地估计出氧化铝浓度;以某铝厂为实际应用背景,结果表明该软测量模型能够有效实现氧化铝浓度的在线检测。
With regard to difficult problems of online measurement of alumina concentration,soft-measurment technique was applied in this paper.The mathematic model was erected by RBF neural network which is provided with good nonlinear approximation and adaptive abilities for different electrolyzer conditions.Finally,the application result in some electrolysis plant shows that the method is efficient.
出处
《计算机测量与控制》
CSCD
北大核心
2010年第11期2472-2474,共3页
Computer Measurement &Control
关键词
电解铝
氧化铝浓度
软测量
RBF神经网络
aluminum electrolysis
alumina concentration
soft sensing
RBF neural network
