摘要
从径向基函数网络的硬件实现和实时应用的角度出发，给出了RBF网络的一种新型混合递推学习算法．该算法既具有良好的数值性质又易于并行实现．把RBF网络用于非线性系统在线辨识，仿真结果显示了本文方法的有效性．
A new hybrid recursive learning algorithm for the radial basis function (RBF) neural networkis proposed from the viewpoint of hardware implementations and on-line applications. The new algorithm hassuperior numerical properties and can be implemented in parallel easily. Finally, recursive identification ofnonlinear systems using RBF network is investigated. The simulations show that the algorithm is very effective.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
1998年第2期272-276,共5页
Control Theory & Applications
关键词
神经网络
学习算法
系统辨图
BRF网络
radial basis function
neual network:learning algorithm
system identification
