Scholarlink

<p>The field of artificial intelligence is currently witnessing a surge of interest in the area of neural network systems. In this paper, a novel cellular neural network system (CNNs) is devised through the estimation of electromagnetic radiation effect with the aid of a memristor. Using detailed theoretical analysis and accurate numerical simulations, complex nonlinear behaviors are induced and controlled in the CNNs by varying the parameters of the memristor and the parameters of the system. The rich and complex dynamical behaviors of CNNs are deeply explored by using plotting bifurcation diagrams, maximum Lyapunov exponential spectra, Sample entropy with two parameters, and coexisting attractors. The offset boosting of the system and having multi-stability properties are found. It is well shown that the introduction of memristor as electromagnetic radiation can have a strong effect on the CNNs. Subsequently, the circuits of the CNNs were designed and simulated, and the constructed CNNs were also implemented on the DSP platform. The consistency between the simulation and implementation results provides evidence for the existence and feasibility of CNNs. The CNNs proposed in this paper can provide a reference for the field of image encryption.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Implementation of a cell neural network under electromagnetic radiation with complex dynamics

  • 期刊论文

  • 发表日期：

    2025年4月5日

查看全文
  • Tao Ma,
  • Jun Mou,
  • Wanzhong Chen

摘要

The field of artificial intelligence is currently witnessing a surge of interest in the area of neural network systems. In this paper, a novel cellular neural network system (CNNs) is devised through the estimation of electromagnetic radiation effect with the aid of a memristor. Using detailed theoretical analysis and accurate numerical simulations, complex nonlinear behaviors are induced and controlled in the CNNs by varying the parameters of the memristor and the parameters of the system. The rich and complex dynamical behaviors of CNNs are deeply explored by using plotting bifurcation diagrams, maximum Lyapunov exponential spectra, Sample entropy with two parameters, and coexisting attractors. The offset boosting of the system and having multi-stability properties are found. It is well shown that the introduction of memristor as electromagnetic radiation can have a strong effect on the CNNs. Subsequently, the circuits of the CNNs were designed and simulated, and the constructed CNNs were also implemented on the DSP platform. The consistency between the simulation and implementation results provides evidence for the existence and feasibility of CNNs. The CNNs proposed in this paper can provide a reference for the field of image encryption.