基于RBF神经网络的混沌时间序列预测

PREDICTION OF CHAOTIC TIME SERIES BASED ON RBF NEURAL NETWORKS

本文提出将径向基函数(RBF)神经网络应用于混沌时间序列的预测,设计了一个三层RBF网络结构。对于三个典型的混沌系统,在不同的噪声水平下,采用RBF网络模型分别进行了预测研究。仿真结果表明,采用RBF网络进行混沌时间序列的预测能够取得比现有其它方法更好的效果。

In this paper, we present that Radial Basis Function neural network can be used in the prediction of chaotic time series. For such purpose, we designed a three-layer RBF network structure. The time series prediction on three typical chaotic systems is investigated with RBF network models at various noise levels. Simulations show that the results will be better than the other methods when RBF neural network is used in prediction of chaotic time series.