机构地区: 湖南大学电气与信息工程学院
出 处: 《电子与信息学报》 2005年第1期143-145,共3页
摘 要: 针对混沌系统模型误差,该文提出一种非线性鲁棒自适应辨识和控制新方法,目标是通过下面两个步骤将混沌系统镇定到不动点:首先利用动态神经网络对系统进行辨识,然后在辨识估计基础上设计控制器将混沌状态引导至期望目标位置;并且对系统的稳定性能进行了严格数学分析;Duffing方程的数值仿真实验证明了所提出方法的有效性。 A new robust adaptive identification-based control of chaotic system with uncertain parameters in view of modeling error is proposed in this paper. The objective is to adjust the unknown chaos to a fixed point. It is fulfilled by taking following two steps: a dynamical neural network is used as system identifier, then a controller based on identification estimates is established to direct the chaos states towards desired target. Also, rigorous mathematical proof is given to analyze the stability properties of the system. Finally, the effectiveness of the proposed method is demonstrated by the Duffing equation.
领 域: [自动化与计算机技术] [自动化与计算机技术]