基于RBF网络的混沌动力系统辨识

提出用RBF神经网络对混沌动力系统进行辨识，设计了一个三层RBF网络结构，仿真实验说明了RBF网络用于学习混沌动力系统时的基本性质。用辨识模型重建吸引子方法定性地评价辨识模型，通过计算辨识模型的Lyapunov指数定量地评价辨识模型的性能，同时推导了RBF网络模型Lyapunov指数的计算公式。仿真结果表明，该辨识模型能很好地逼近原混沌动力系统，准确地体现原混沌系统的动力学特性。

Identification of Chaotic Dynamical Systems Based on RBF Neural Networks

Chaotic dynamical systems can be identified by RBF neural networks.A three layer RBF network structure was designed and the fundamental properties of the RBF networks were clarified to learn chaotic dynamical systems through some numerical experiments.A qualitative evaluation of the identified models was made with the reconstruction of an attractor by the identified models,and a quantitative evaluation of the identified models was made with calculation of the Lyapunov exponents of the identified models,too.The formula of the Lyapunov exponents of RBF networks models is derived.Simulations show that the identified models can approach the original chaotic dynamical systems and extract the dynamical characteristics of the original chaotic systems.