摘要
根据逆最优控制方法,针对非线性系统,提出了利用动态神经网络产生混沌的一种新方法。令动态神经网络复制所需要的混沌系统来设计控制器。鉴于混沌反控制的本质,巧妙地引入了一个跟踪控制。这种方法不需要估计混沌吸引子的Lyapunov指数,从而大大地减少了计算量,这说明这种方法便于实际应用。以Lorenz系统为例,进行了数值仿真,结果表明所提出的控制器是有效的,性能是良好的。
Propose a new approach to generate chaos via dynamic neural networks according to inverse optimal control for nonlinear systems. The controller is derived to force dynamic neural network to reproduce intended chaos. In light of the nature of chaotification, a track control is subtly introduced. This method need not estimate the Lapunov exponent of the chaotic attractor so that computation is reduced greatly, which means this scheme is simple in practical application. Lorenz system and Rossler chaos are used for examples. The numerical simulation results demonstrate the effectiveness and excellent performance of the suggested controller.
出处
《盐城工学院学报（自然科学版）》
CAS
2005年第3期5-8,共4页
Journal of Yancheng Institute of Technology:Natural Science Edition
关键词
混沌
混沌反控制
逆最优控制
动态神经网络
chaos, chotification, inverse optimal control, dynamic neural networks
