Coexisting firing patterns and attractor selection in memristive synapse coupled heterogeneous neurons

Abstract

In this paper, we present a novel heterogeneous neuron network involving the coupling of a two-dimensional (2D) continuous Rulkov model with a 2D Hindmarsh-Rose (HR) neuron via memristive synapse. The present memristive coupled neuron network can produce an infinite number of coexisting hidden attractors with the same shape but located at different positions in the phase space, resulting in the interesting phenomenon of hidden homogeneous extreme multistability. Firstly, we study complex dynamical behaviors in detail through bifurcation diagrams, Lyapunov exponents, time series and phase portraits. Secondly, to select the desired firing pattern, a control method based on the feedback term is applied to the model. It is observed that the addition of this space-dependent feedback term to the dynamic equation of this model effectively eliminates all undesired firing patterns among the coexisting ones, and drives the heterogeneous neuron network to transition from coexisting hidden firing patterns to the expected firing pattern. Finally, Multisim circuit simulation is performed and the results are in line with numerical simulation.