振荡神经网络的建模与动态联想记忆

Dynamic Associative Memory with Oscillatory Neural Networks

本文提出一种全程耦合映射格子式的振荡神经网络模型，该模型具有丰富的非线性动力学行为，在其联想记忆的数值模拟中，观察到同步振荡和时空混沌现象。该网络的动态特性有助于区分具有相同吸引子的不同模式。

A model of oscillatory neural networks (ONN) similar to globally coupled map lattices is proposed. Numerical simulation suggests that the dynamic feature of the ONN is beneficial to the identification of different patterns which shared the identical attractor in the retrieval of autoassociative memory. The ONN shows some phenomena of nonlinear dynamics including synchronized oscillation, spatiotemporal chaos and et al.