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

Area of Science:

• Nonlinear Dynamics
• Complex Systems
• Computational Neuroscience

Background:

• Existing multiscroll Hopfield neural networks (HNNs) suffer from high sensitivity to parameter changes and increasing complexity.
• These limitations hinder the practical application of multiscroll HNNs for generating chaotic signals.

Purpose of the Study:

• To propose a robust full-parameter control method for a novel memristive chain HNN (MCHNN).
• To address the limitations of existing HNNs by developing a simpler and more practical multiscroll attractor generator.

Main Methods:

• Detailed theoretical and numerical analysis of a newly designed memristor's electrical characteristics.
• Investigation of the MCHNN's multiscroll attractor structures using equilibrium point and stability analysis.
• Exploration of complex dynamics, including multistability and attractor transformations, through parameter and initial state variations.

Main Results:

• The proposed MCHNN exhibits a simpler chain topology compared to traditional HNNs.
• The MCHNN successfully generates diverse multiscroll attractors with complex dynamics, including multistability and orbit transformations.
• A digital experimental platform validated the MCHNN's feasibility for generating usable multiscroll chaotic signals.

Conclusions:

• The novel MCHNN offers a more robust and simpler approach to constructing multiscroll chaotic systems.
• The MCHNN's ability to generate diverse chaotic dynamics and its practical implementation demonstrate its potential.
• The MCHNN shows promise for applications such as pseudorandom number generation due to its high randomness.