Synchronization of Memristively Couples Oscillator Networks – Theory and Emulation

基本学习机制指的是同步模式，其中单个神经元的信号是同步的。同步能力是由不同的耦合机制、互连拓扑和分布时延决定的。在这个项目中，这些都是在振荡器网络方面解决的，这些振荡器网络具有记忆和反应耦合元件，从而产生动态适配的图形。该方法是相辅相成的，它结合了用于网络结构分析的控制论概念，以及与面向电路的波形数字仿真概念确保和使能同步的设计规则的推导，以综合和评估含有记忆二端口的网络的电路相互作用。除了常微分方程式中的有限维振子模型外，还将考虑涉及偏微分方程式的连续体模型，以研究分布式振荡器的大型系综及其自组织。所开发的分析、设计和综合技术将在数值模拟和WAVE数字仿真中进行评估。

Basic learning mechanisms refer to synchronization patterns, where the signals of the individual neurons are synchronized. The synchronization ability results from different coupling mechanisms, the interconnection topology and distributed delays. These are addressed in this project in terms of oscillator networks with memristive and reactive coupling elements resulting in dynamically adapting graphs. The methodical approach is complementary and combines control theoretic concepts for the structural analysis of the network and the deduction of design rules ensuring and enabling synchronization with circuit oriented concepts of wave digital emulation to synthesize and to evaluate electric circuit interactions of networks involving memristive two-ports. In addition to finite-dimensional oscillator models in terms of ordinary differential equations continuum models involving partial differential equations will be considered to address large ensembles of distributed oscillators and their self-organization. The developed analysis, design and synthesis techniques will be evaluated in numerical simulations and wave digital emulations.