Extended dissipative state estimation for memristive neural networks with time-varying delay

Area of Science:

• Control Theory
• Neural Networks
• Nonlinear Systems

Background:

• Memristor-based neural networks (MNNs) are crucial for advanced computing.
• Time-varying delays in MNNs complicate state estimation.
• Extended dissipative state estimation is vital for system analysis.

Purpose of the Study:

• To develop novel criteria for extended dissipative state estimation in MNNs with time-varying delays.
• To ensure robust and accurate state estimation for these complex networks.
• To provide a unified framework for various state estimation types.

Main Methods:

• Utilizing nonsmooth analysis and differential inclusions.
• Constructing a new Lyapunov-Krasovskii functional.
• Applying set-valued maps and novel integral inequalities.

Main Results:

• Derivation of new extended dissipative state estimation criteria.
• Demonstration of the criteria's applicability to l2-l∞, H∞, passive, and dissipative state estimation.
• Validation of effectiveness and reduced conservatism through numerical examples.

Conclusions:

• The proposed criteria offer an effective and less conservative approach to extended dissipative state estimation for MNNs.
• The methodology provides a flexible framework adaptable to different state estimation requirements.
• Numerical simulations confirm the theoretical findings and practical utility.