摘要
讨论径向基函数网络（ＲＢＦ网络）的网络结构和基本算法，在此基础上提出了鲁棒自适应ＲＢＦ网络方法。仿真结果表明。
A modeling approach of polypropylene melt index(MI) based on robust and adaptive radial basis function(RBF) neural networks is proposed in this paper. In this method, adaptive algorithm and robust regression are incorporated to deal with time-varying property and error data respectively. Test results have shown its better capability to predict polypropylene MI than BP neural networks.
出处
《控制与决策》
EI
CSCD
北大核心
1999年第4期339-343,共5页
Control and Decision
基金
国家重点自然科学基金
国家863/CIMS主题应用基础研究基金
