神经元网络系统的斑图动力学行为分析及控制

项目名称： 神经元网络系统的斑图动力学行为分析及控制

项目编号： No.11472238

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 数理科学和化学

项目作者： 史雪荣

作者单位： 盐城师范学院

项目金额： 66万元

中文摘要： 斑图动力学由于其广泛的工程应用背景而成为当前研究的热点课题之一。本项目拟利用复杂网络的相关知识、非线性理论分析、数值模拟与仿真等方法，深入探讨神经元网络系统的斑图动力学行为及诱发机制，分析系统斑图的复杂特征，从而灵活、有效地对斑图进行选择和控制。首先，讨论时滞、脉冲耦合对神经元网络系统斑图动力学行为的影响机制。其次，研究分布式刺激和不同的拓扑结构对神经元网络斑图动力学行为的影响及诱发机理，探索网络的耦合系数、耦合强度对斑图动力学的影响，考察神经元网络达到混沌同步的阀值，实现对混沌同步的有效控制。再次，研究网络边界条件对斑图的影响和选择机制，探索边界效应对系统时空斑图的影响。最后，根据神经元网络系统的斑图动力学行为讨论神经元网络的突触可塑性，探讨突触可塑性与学习记忆的关系，特别是长时程增强与学习记忆的关系。该研究结果有助于人们深入理解神经元网络系统的复杂现象及其转化模式。

中文关键词： 斑图动力学；神经元网络；时滞；脉冲；突触可塑性

英文摘要： Pattern dynamics is one of the hot topics of current research due to its extensive engineering background. The knowledge of complex network, nonlinear theory analysis and numerical simulations will be used to further explore the pattern dynamics of neural network and its induced mechanism, to analyze the complex features of pattern dynamics of the system, so as to select and control the pattern flexibly, effectively. Firstly, the impact mechanism of time-delay or pulse coupling on neural network is discussed. Secondly, the effects of distributed stimulation or different topology structures on pattern dynamics of neural network and its induced mechanism are researched. The impact of the coupling coefficient, coupling strength is discussed and the threshold to achieve chaotic synchronization of neural networks is studied which makes chaotic synchronization be controlled effectively. Thirdly, the impact and selection mechanisms of boundary conditions to pattern are investigated. The impact of boundary effect on the form of spatiotemporal pattern is probed. Finally,according to the pattern dynamics of neural network,the synaptic plasticity of neural network is discussed. The relationship between synaptic plasticity and learning memory, especially, between the Long-term potentiation and learning memory, is explored. The results are helpful to further understand the complex phenomena and their transformation model of the neural network.

英文关键词： pattern dynamics;neural network;time-delay;pulse;synaptic plasticity