福建师范大学学报（自然科学版）2025，Vol.41Issue(6)：1-12,12.DOI:10.12046/j.issn.1000-5277.2025040007
多涡卷忆阻双神经元网络动力学分析与电路实现
Dynamical Analysis and Circuit Realization of a Multiscroll Memristive Two-Neuron Network
摘要
Abstract
Compared with ordinary chaotic attractors,multiscroll attractors exhibit higher com-plexity and greater application potential.However,most existing neural networks capable of genera-ting multiscroll attractors are high-dimensional and characterized by complex connectivity.In this study,a memristor with a piecewise arctangent function is used to replace the fixed weight in a two-neuron network.Based on this,a memristive two-neuron network capable of generating complex multiscroll chaotic attractors is constructed.The designed network incorporates an external stimulus current that induces chaotic attractors,and the number of attractors varies with changes in the piecewise function of the memristor.By adjusting the parameters,the multiscroll attractors can be decomposed into different numbers of coexisting attractors,depending on the design of the piecewise function.In addition,when only a single scroll exists in the system,the amplitude can be modula-ted by parameter variation.A digital circuit suitable for engineering applications is further designed and implemented,and the multiscroll attractors are successfully captured on an oscilloscope.Final-ly,the NIST test is performed to verify the practical applicability of the system.关键词
混沌/多涡卷吸引子/忆阻器/Hopfield神经网络/共存吸引子Key words
chaos/multiscroll attractors/memristor/Hopfield neural network/coexisting attractors分类
物理学引用本文复制引用
赖强,徐雨笛..多涡卷忆阻双神经元网络动力学分析与电路实现[J].福建师范大学学报（自然科学版）,2025,41(6):1-12,12.基金项目
国家自然科学基金项目(62366014) （62366014）
