Extreme Multistability in a Hopfield Neural Network Based on Two Biological Neuronal Systems
Lilian Huang, Yue Zhang, Jianhong Xiang, Jin Liu
Abstract
Based on different biological neuronal systems, various memristive neuron and neuron network models are generated. In this brief, a three-neurons-based Hopfield neural network with two biological neural mechanisms is investigated, with one of three neurons exposed to electromagnetic radiation and one of the synaptic weights replaced by a memristor. The effects of electromagnetic radiation and memristive synaptic weight on neural networks are investigated. It is found that coexisting behaviors exist in the neural network by adjusting the memristive synaptic weight. Moreover, infinitely many coexisting behaviors are observed by changing the initial electromagnetic radiation. That is, both of them are critical to the system and have the potential to complicate it. Furthermore, circuits of the neural network are implemented in Pspice and DSP board, and the experimental results are in agreement with numerical ones.