吸引子
计算机科学
人工神经网络
纸卷
记忆电阻器
Hopfield网络
网格
拓扑(电路)
平衡点
物理神经网络
控制理论(社会学)
物理
人工智能
时滞神经网络
人工神经网络的类型
电气工程
数学
非线性系统
数学分析
量子力学
历史
控制(管理)
几何学
考古
工程类
出处
期刊:Physica Scripta
[IOP Publishing]
日期:2025-07-17
卷期号:100 (8): 086002-086002
被引量:3
标识
DOI:10.1088/1402-4896/adf153
摘要
Abstract Grid-symmetric multi-scroll attractors (GSMSA) are of great significance to the field of neural networks due to their complex and flexible characteristics, but there have been few studies on GSMSA to date. In this work, a novel strategy is proposed to generate GSMSA in the magnetized Hopfield neural network (MHNN). First, the model of MHNN composed of four neurons is established. Through the equilibrium point and stability analysis, it is found that the MHNN model can produce rich dynamic behaviors, and the numerical simulations are performed to verify that the MHNN can generate controllable multi-scroll attractors. Second, a memristive self-connection synapse is added to make the system generate grid multi-scroll attractors (GMSA). Next, the multi-level-logic pulse function IMLP, which acts on the electromagnetic memristor, is introduced to generate more complex and flexible GSMSA. Finally, the feasibility of GSMSA is verified by Multisim simulation software.
科研通智能强力驱动
Strongly Powered by AbleSci AI