论文部分内容阅读
提出了一类带有离散时间 FIR/ IIR滤波器的递归 RBF神经网络 ,用离散时间 FIR/ IIR滤波器代替通常的 RBF神经网络中的线性输出权值 ,以适用于离散动力学系统的辨识和控制以及混沌时间序列预测 .本文给出的学习算法简单 ,可以避免传统的递归算法的不稳定性 .将该类神经网络用于动力学系统的建模 ,收到很好的效果 .
A class of recursive RBF neural networks with discrete-time FIR / IIR filters is proposed. The discrete-time FIR / IIR filters are used to replace the linear output weights in ordinary RBF neural networks to apply to discrete-time dynamical system identification and Control and prediction of chaotic time series.The learning algorithm presented in this paper is simple and can avoid the instability of the traditional recursive algorithm.The neural network is used to model the dynamic system and received good results.