Dynamic Analysis and FPGA Implementation of a New Fractional-Order Hopfield Neural Network System under...

Area of Science:

• Neuroscience
• Chaos Theory
• Electrical Engineering

Background:

• Fractional calculus enhances neural network modeling of human brain temporal memory.
• Investigating fractional-order neural networks reveals complex dynamics beyond integer-order models.

Purpose of the Study:

• To propose a magnetically controlled, memristor-based, fractional-order chaotic system using the Hopfield neural network (HNN).
• To analyze the system's dynamics under electromagnetic radiation and validate its implementation.

Main Methods:

• Utilizing the Adomian Decomposition Method (ADM) to solve the fractional-order system.
• Conducting dynamic simulations to explore system behaviors.
• Implementing the fractional-order HNN system on a Field-Programmable Gate Array (FPGA).

Main Results:

• The system exhibits rich dynamics including chaos, quasiperiodicity, and direction-controllable multi-scroll attractors.
• Analogous symmetric dynamic behaviors emerge with altered radiation parameters while maintaining a constant fractional order.
• Experimental validation on FPGA confirms the theoretical findings.

Conclusions:

• The proposed fractional-order Hopfield neural network system accurately simulates complex brain-like memory effects.
• The system's dynamic behaviors are controllable and sensitive to electromagnetic radiation.
• FPGA implementation demonstrates the practical feasibility of this advanced neural network model.