Dynamic analysis and reliable mechanical optimization application of ring HNN effected with a memristive neuron

Area of Science:

• Nonlinear dynamics
• Computational neuroscience
• Materials science

Background:

• Memristor-based neural networks simulate biological synapses and exhibit complex dynamics.
• Their inherent randomness is promising for mechanical optimization.
• Locally active memristors are key components in advanced neural network models.

Purpose of the Study:

• To design and analyze a ring Hopfield neural network model incorporating a memristive neuron (RHNN-MN).
• To investigate the influence of memristive neurons on the network's nonlinear dynamics.
• To explore the application of the model's chaotic sequences for reliable mechanical optimization.

Main Methods:

• Construction and analysis of a locally active memristor model.
• Investigation of memristor integration dynamics within the RHNN-MN.
• Analysis of complex dynamics (bifurcations, chaos) by varying memristor parameters and network weights.
• Development and testing of a novel chaos optimization algorithm.

Main Results:

• The RHNN-MN model exhibits complex nonlinear dynamics, including multiple bifurcations, quasi-periodicity, and chaos.
• The memristor parameters and network weights significantly influence the system's dynamics.
• The proposed chaos optimization algorithm demonstrates superior performance in solving mechanical problems.

Conclusions:

• The memristive neuron significantly impacts the dynamics of ring Hopfield neural networks.
• The chaotic sequences generated by the RHNN-MN are effective for mechanical optimization.
• This research provides valuable theoretical and practical insights for designing novel neural network models and enhancing mechanical optimization techniques.