摘要
给出了Wiener型非线性系统的一种新的辨识方法。采用Laguerre函数与静态BP网络相结合构成Wiener型非线性系统的模型,并给出此模型参数递推辨识算法。这种模型参数辨识不需要实际系统的阶次和时延先验知识,对实际系统阶次和时延变化有较强的鲁棒性。数字仿真验证了该方法的有效性。
A new identification method for the nonlinear systems of Wiener type is proposed. The model of nonlinear systems of Wiener type is made up of Laguerre function networks with static BP (SBP) networks, and also, its recursive identification algorithm is given. The parameter identification of this model needs no a priori knowledge on the order degree and the time delay of the identified nonlinear systems. The digital simulation result shows the efficiency of the proposed method.
出处
《系统仿真学报》
CAS
CSCD
2002年第8期1053-1055,共3页
Journal of System Simulation
基金
安徽省自然科学资金(99043203)资助项目
