摘要
对于具有相似性的一类不确定复杂系统,应用高阶神经网络逼近各个子系统的互联项,设计了控制器,即提出了难以解决的互联项问题的高阶神经网络表示方法.该方法通过在线调节神经网络的权值来确保闭环系统的稳定性.由于复杂系统的结构相似性,降低了控制器设计过程中的计算量,使得工程上较易实现.仿真算例表明了所提出方法的有效性.
A novel robust adaptive controller is designed for a class of uncertain complex systems with similarity. This approach uses high-order neural networks to deal with interconnections of subsystems. It assures the stability of closed-loop systems by updating the weights of NN on line. It needs less calculation and is easy to realize in engineering because of the similarities of complex systems. The simulation example shows the effectiveness of proposed method.
出处
《控制与决策》
EI
CSCD
北大核心
2006年第8期929-932,936,共5页
Control and Decision
关键词
高阶神经网络
相似复杂系统
鲁棒自适应控制
High-order neural networks
Complex systems with similarity
Robust adaptive control
