Synchronization of Coupled Neural Networks With Constant Time-Delay Using Sampled-Data Information.

Area of Science:

• Control Systems Engineering
• Computational Neuroscience
• Network Science

Background:

• Coupled neural networks (CNNs) are fundamental models in neuroscience and control systems.
• Achieving synchronization in CNNs with time delays is challenging due to inherent system dynamics.
• Existing methods often require continuous data, which is impractical for many applications.

Purpose of the Study:

• To develop a novel synchronization control method for CNNs with constant time delay using sampled-data information.
• To reduce the conservatism of synchronization conditions and minimize communication energy load.
• To determine the maximum allowable sampling interval for effective synchronization.

Main Methods:

• A distributed control protocol based on sampled-data information from neighboring nodes was designed.
• Lyapunov functional and advanced integral inequalities (Park's, improved free-weight matrix) were employed.
• An optimization problem was formulated to determine the maximum sampling interval.
• Aperiodic sampling control techniques were implemented.

Main Results:

• Sufficient conditions for achieving synchronization in CNNs with constant time delay were derived with reduced conservatism.
• The maximum sampling interval was successfully determined.
• The aperiodic sampling control technique effectively reduced communication energy load.
• Numerical simulations validated the proposed method's capability to achieve synchronization.

Conclusions:

• The proposed sampled-data control method effectively achieves synchronization in CNNs with constant time delay.
• The method offers reduced conservatism and lower communication costs compared to existing approaches.
• The findings provide a practical framework for designing synchronization controllers for delayed neural networks.