摘要
将小波神经网络应用于动态非线性连续时间系统的辨识 ,同时为了使神经网络的训练达到全局最优和加速小波神经网络训练的收敛速度 ,提出了信赖域算法 ,并研究了信赖域算法的收敛性 .随后进行了算例仿真 ,证明了所提辨识方法的有效性 .
The wavelet neural networks are used for identification of continuous time nonlinear systems. For the improvement of convergent speed and the global optimization in the training of wavelet neural networks, the trust region algorithm is proposed for the training. Then we study the convergence properties about the trust region algorithm. Finally, the simulation results are given to illustrate the efficiency of the method proposed.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2002年第5期709-712,共4页
Control Theory & Applications
基金
国家自然科学基金 ( 6 97740 12 )资助项目
