| 基于神经网络误差修正的混沌控制与同步 |
| |
| 引用本文： | 董恩增,陈增强,袁著祉.基于神经网络误差修正的混沌控制与同步[J].控制工程,2006(Z1). |
| |
| 作者姓名： | 董恩增 陈增强 袁著祉 |
| |
| 作者单位： | [1]南开大学信息技术科学学院 [2]天津 |
| |
| 基金项目： | 国家自然科学基金资助项目(60374037,60574036)，高校博士点专项基金资助项目(20050055013)，教育部新世纪人才计划资助项目(NCET) |
| |
| 摘 要： | 针对混沌系统非线性强、多变量耦合等特点,提出了一种基于神经网络误差修正的自适应多变量混沌系统的广义预测控制算法,用线性广义预测控制器控制混沌系统,用神经网络对模型预测误差进行修正。算法中辩识过程模型用递推最小二乘法(RLS)、神经网络权值用Davidon最小二乘法(DLS)训练。这种算法对被控混沌系统的先验知识要求较少,无需知道被控系统的精确模型,数值仿真显示可实现混沌系统的宽范围控制与同步。
|
| 关 键 词： | 神经网络 混沌控制 混沌同步 递推最小二乘法 Davidon最小二乘法 |
Adaptive Predictive Control and Synchronization of Chaotic System Based on Neural Network Errors Compensation |
| |
| DONG En-zeng,CHEN Zeng-qiang,YUAN Zhu-zhi.Adaptive Predictive Control and Synchronization of Chaotic System Based on Neural Network Errors Compensation[J].Control Engineering of China,2006(Z1). |
| |
| Authors: | DONG En-zeng CHEN Zeng-qiang YUAN Zhu-zhi |
| |
| Abstract: | According to strong nonlinearity and the multi-variable coupling characteristic of chaotic systems,an adaptive predictive control method for control and synchronization of chaotic systems based on neural network errors compensation is proposed.Linear generalized predictive controller is used to control chaotic systems,and the model predictive errors are compensated by neural network.The process model is identified by recursive least square(RLS),and the neural network connection weights are identified by Davidon least square(DLS).By this method,the chaotic systems can track discretionary reference signal.The method can actualize effective control and synchronization of chaotic systems without knowing their precise models.The effectiveness of the proposed method is verified by numerical simulations. |
| |
| Keywords: | neural network chaotic control chaotic synchronization recursive least square method Davidon least square method |
| 本文献已被 CNKI 维普 等数据库收录！ |
