Memristor-Based Neural Network Circuit of Full-Function Pavlov Associative Memory With Time Delay and Variable...

Area of Science:

• Neuroscience
• Materials Science
• Computer Engineering

Background:

• Traditional memristor-based Pavlovian associative memory requires simultaneous stimuli.
• Introducing time delays in associative learning is crucial for more complex cognitive functions.

Purpose of the Study:

• To design and verify a memristive neural network circuit that implements Pavlovian associative memory with a time delay.
• To explore how varying time delays affect learning rates in associative memory.

Main Methods:

• A novel memristive neural network circuit was designed.
• The circuit incorporates synapse, voltage control, and time-delay modules.
• Simulations were used to verify the circuit's functionality.

Main Results:

• The circuit successfully demonstrated Pavlovian associative memory with a time delay.
• The learning rate was shown to be adaptable based on the stimulus time interval.
• Functions like learning, forgetting, and time-delay learning were implemented.

Conclusions:

• The developed memristive neural network effectively models associative memory with time-delayed stimuli.
• This work provides a foundation for creating more sophisticated brain-inspired computing systems.
• The circuit's adaptability in learning rates offers new possibilities for neuromorphic engineering.