Symmetric multi-scroll attractors in magnetized Hopfield neural network under pulse controlled memristor and pulse current stimulation

Abstract

In this paper, a new method for generating symmetric multi-scroll attractors is proposed based on a magnetized Hopfield neural network (HNN). Firstly, the mathematical model of magnetized HNN by introducing an original memristor to describe the effect of electromagnetic radiation is established, and its dynamic characteristics are analyzed. Then, utilizing two types of multi-level-logic pulse functions, the complex attractor structure changes of the magnetized HNN under external pulse controlled memristor, pulse current stimulation and their combined effects are discussed respectively. The numerical simulation results show that the magnetized HNN under pulse controlled memristor or/and pulse current stimulation can produce multi-scroll chaotic and multi-scroll periodic strange attractors with symmetric structure. The number of attractors can be changed by setting the different series of multi-level-logic pulse functions. Finally, the magnetized HNN circuit under pulse controlled memristor or/and pulse current stimulation is realized by using discrete analog electronic devices. The experimental results verify the validity of the numerical simulation results.