混沌系统的神经网络预测控制

提出一种基于预测控制的神经网络控制方法，将模型未知时的混沌运动控制到不稳定的不动点（UFP）处，该控制系统不需要UFP的位置及其局性态等知识，它包括观测器、带反馈校正的神经网络在预测器和在线训练的神经网络控制器，其方法简便，收敛速度比现有同类方法快得多，同时还分析了控制系统的稳定性，并证明了神经网络控制器的收敛性，理论推导和仿真结果都表明了该方法的有效性。

Neural network predictive control of chaotic systems

Aneuralnetworkcontrolalgorithmbased on predictive control is proposed. When the chaotic model is unknown, the control system stabilizes a chaotic orbit into an unstable fixed point (UFP) without using the knowledge of the location of the UFP and the local linearized dynamics at the UFP. The control system includes a watcher, an on line neural predictor with feedback adjustment and an on line trained neural controller. The proposed algorithm is simple and its convergence speed is much higher than existing similar algorithm. The stability of the control system is analyzed and the convergence property of the neural controller is proved. The theoretic derivation and simulations demonstrate the effectiveness of the algorithm.