Robust full-parameter control method: Constructing multiscroll HNN via memristor

Existing methods for constructing multiscroll Hopfield neural network (HNN) result in the drawbacks of high sensitivity of model form to scroll count and growing complexity, posing bottlenecks and inconveniences for practical applications. To address this issue, this paper proposes a robust full-parameter control method to construct a novel memristive chain HNN (MCHNN) yielding multiscroll attractors, based on a newly designed memristor. By theoretical and numerical methods, the electrical characteristics of the memristor are analyzed in detail. Notably, the MCHNN features a simpler chain topology, yet it can yield multiscroll attractors with diverse structures, which are investigated via equilibrium point and stability analysis. Furthermore, configuration variations in the network parameters and initial state induce rich and complex dynamics, including extremely complex homogeneous and heterogeneous multistability, coexisting multiscroll attractors, scroll orbit region transformations, distribution boosting and ripple attractor. To validate the practical feasibility of the proposed method so as to yield real and usable multiscroll chaotic signals, a digital experimental platform is constructed to implement the MCHNN. Additionally, the MCHNN is applied to developing a pseudorandom number generator with high randomness.