混沌系统的快速模糊控制算法

讨论了用模糊推理方法实现混沌系统的控制问题。利用递推模糊聚类算法实时对系统的输入空间进行模糊划分，利用卡尔曼滤波算法确定参数。在此基础上，给出了模糊模型的在线辨识算法。该方法不需要被控混沌系统的解析模型，控制的目标可以为周期轨道，也可以为连续变化的目标函数，在模型参数发生摄动和存在噪声情况下控制仍然有效。针对Henon吸引子的仿真结果，表明该方法的有效性和可行性。

Fast Algorithm for Fuzzy Control of Chaotic System

The control of chaotic system using fuzzy inference is discussed.The input space of fuzzy system is partitioned by means of real time recursive fuzzy clustering,and the parameters of fuzzy model are confirmed by Kalman filtering.Based on this method,the on-line identifying algorithm is proposed.The control objective can be either periodic orbits or continuous variable functions without the need of an analytic model.The method is still effective when there are parameter perturbation and measurement noise.Simulation results with Henon attractor show the effectiveness and feasibility of the proposed method.