一类混沌神经网络的观测器广义投影同步设计  

A class of Observer-Based Generalized Projective Synchronization of Chaotic Neural Networks

作　　者：王兴元[1] 孟娟[1] 

机构地区：[1]大连理工大学,大连116024

出　　处：《应用力学学报》2008年第4期656-659,共4页Chinese Journal of Applied Mechanics

基　　金：国家自然科学基金(30170259);辽宁省教育厅高等学校科学技术研究计划(20040081)

摘　　要：研究了混沌神经网络的广义投影同步问题。基于改进的非线性状态观测器理论和极点配置技术,设计了一种新的广义投影同步方案。该方案可以实现不同混沌神经网络之间的广义投影同步。通过对细胞神经元系统和三阶Winner-Take-All竞争型混沌神经元系统的数值仿真试验,进一步验证了所提方案的有效性。该方案具有全局性,与混沌神经网络的初始条件无关,且可以灵活地调整同步误差的收敛速度。The generalized projective synchronization of chaotic neural networks is investigated. Based on the modified nonlinear state observer algorithm and the pole placement technique, a new synchronization scheme is designed. The generalized projective synchronization of different chaotic neural networks can be achieved with the proposed method. The numerical simulations of the neuron system and the three order Winner-Take-All competed neural network further demonstrate the effectiveness of the proposed scheme, with is global and independent on the initial values of chaotic neural networks. and the convergence rate of synchronization errors can be regulated flexibly.

关 键 词：非线性状态观测器 极点配置技术 广义投影同步 混沌神经网络 

分 类 号：R318.19[医药卫生—生物医学工程]