时滞反馈Lorenz系统的神经网络滑模自适应同步控制及在保密通信中的应用

Neural Network Sliding-mode Adaptive Synchronization Control for Delayed Feedback Lorenz System and Its Application in Secure Communication

为了研究不确定Lorenz混沌系统同步控制在保密通信中的应用,首先设计了时滞反馈Lorenz混沌系统,并通过Poincare映射和功率谱分析了其混沌动力学特性。在此基础上,提出了不确定时滞反馈Lorenz混沌系统的神经网络滑模自适应同步控制策略。应用径向基(RBF)神经网络逼近混沌系统的不确定项,基于该径向基神经网络的输出再利用滑模控制和自适应控制相结合的方法提出了单维同步控制器的设计。最后,将所设计的同步控制方法应用于保密通信。仿真结果表明,本文所提出的神经网络滑模自适应同步控制方法可以实现混沌系统同步并可应用于保密通信,且具有较强的抗干扰能力。

To study the synchronization control of uncertain Lorenz systems and its application in secure communication,the time-delay feedback Lorenz chaotic system was designed and its chaotic dynamics was analyzed by the Poincare map and the power spectrum.The neural network sliding-mode adaptive synchronization control scheme was proposed for the delayed feedback Lorenz chaotic system.The RBF neural network was used to approximate the chaotic system uncertainty.Combining with the output of neural networks,the sliding mode control and adaptive control technique were applied to achieve a single dimension signal controller.On the basis of the proposed synchronization control scheme,the application in secure communication was developed.The experimental results showed that the proposed neural network sliding-mode adaptive synchronization control scheme can realize chaotic synchronization and be used in secure communication,which has strong anti-interference ability.