摘要
首先提出神经元数目有限的小波神经网络对一大类 L ipschitz函数的逼近能力定理 ;然后对Thau定理进行推广 ,得到几个实用性较强的推广定理 ;最后通过构造一种基于推广 Thau定理的小波神经网络非线性观测器 ,展示出该逼近定理的应用前景。
A novel approximate theorem of wavelet neural networks is presented at first. Then the well known Thau theorem is improved and some practical extensive theories are got. Finally, based on the extensive Thau theorem, a kind of state observers for nonlinear systems using WNN is constructed. Simulation results show the effectiveness of the WNN approximate theorem.
出处
《控制与决策》
EI
CSCD
北大核心
2000年第5期561-564,568,共5页
Control and Decision
基金
国家自然科学基金项目!(6 97740 14)
关键词
小波
神经网络
Thau定理
推广
函数逼近
wavelet neural networks, approximate ability, improvements of Thau theorem, nonlinear observers
