摘要
混沌系统的建模与辨识是混沌控制的基础。提出一种动态线性子系统与ＲＢＦ神经网络并联的增广ＲＢＦ神经网络模型，该模型不仅对动态非线性系统具有良好的逼近能力，而且网络学习速度很快。对Ｈｅｎｏｎ系统时间序列的仿真预测结果表明，增广ＲＢＦ网络能有效地用于混沌系统辨识。
Modeling and identifying chaotic system is the basis of chaos control. An extended radial basis function (ERBF) neural network model in which a linear dynamics subsystem and a RBF neural network are connected in parallel has been presented in this paper, which possesses not only a very good capability in approximating nonlinear dynamics systems, but also a rapid speed of network learning. Simulation results for the Henon system time series prediction show that the extended RBF neural network can be applied effectively to identify chaotic system.
出处
《北京化工大学学报（自然科学版）》
CAS
CSCD
1999年第1期45-48,共4页
Journal of Beijing University of Chemical Technology(Natural Science Edition)
