基于RBF神经网络系统辨识研究

Study on the System Identification of the RBF Neural Network

从神经网络系统辨识原理出发 ,利用 Matlab神经网络工具箱中提供的一组输入输出数据对 ,应用 RBF网络进行系统辨识 ,建立 RBF神经网络辨识模型 ,模拟其输出曲线 ,并且对辨识结果做了对比分析 ,进而实现了 RBF网络的最佳逼近性质。

The paper starts from the system identification p ri nciple of neural networks,using a group of I/O data provided by the neural netwo rk toolbox for Matlab, applying the RBF network to process system identific ation ,setting up the RBF neural network identification model,simulating the output cu rve,be sides compared and analyzed for the result of identification,realizing further t he best approximation character of the RBF network.