Complex dynamics in a Hopfield neural network under electromagnetic induction and electromagnetic radiation

Area of Science:

• Neuroscience
• Complex Systems
• Electrical Engineering

Background:

• Neural networks operate in complex electromagnetic environments due to potential differences.
• Electromagnetic induction and radiation can influence neural network behavior.

Purpose of the Study:

• To emulate electromagnetic induction and radiation effects on a Hopfield neural network (HNN).
• To investigate the dynamic behaviors and stability of the HNN under electromagnetic influences.

Main Methods:

• Utilized hyperbolic-type and quadratic nonlinear memristors to model electromagnetic effects.
• Analyzed the HNN's dynamics through numerical simulations (Matlab, Multisim).
• Validated simulation results with breadboard experiments.

Main Results:

• The HNN exhibits an unstable origin equilibrium point.
• Complex dynamics, including transitions from regular to chaotic and back to regular motion, were observed.
• Coexisting bifurcations leading to bi-stable patterns were identified with changes in synaptic weights.

Conclusions:

• Memristor-based emulation effectively models electromagnetic influences on HNNs.
• The HNN displays rich and controllable dynamic behaviors influenced by electromagnetic coupling.
• Experimental validation confirms the theoretical and simulation findings.