Chebyshev神经网络辨识器

基金项目:

湖南省教委科研项目

Identifier Based on Chebyshev Neural Network

摘要: 基于逼近论，将一组Ｃｈｅｂｙｓｈｅｖ正交多项式取代ＢＰ网络中的Ｓ型函数，构成一种新的神经网络模型。理论分析和仿真实验表明，该网络可逼近任意非线性系统，且建模容易，收敛速度快，学习次数远远少于ＢＰ网络。
- 正交多项式 /
- 神经网络 /
- 系统辨识
Abstract: Based on the approximation theory, this paper presents a new neural network model based on Chebyshev orthogonal multinomial. By theoretical analysis and simulation experiments we show that the neural network has excellent performance of approximating nonlinear systems and the learning speed is high than BP networks.
- Orthogonal multinomia,l Neural network, Identificatio