Practically Predefined-Time Stabilization of Stochastic Fuzzy Memristive Neural Networks Under Deception Attacks

Area of Science:

• Control Theory
• Neural Networks
• Network Security

Background:

• Fuzzy memristive neural networks (FMNNs) are crucial for complex computations.
• Ensuring stability in FMNNs under stochastic disturbances and random deception attacks (RDAs) is a significant challenge.
• Existing methods for predefined-time stabilization have limitations.

Purpose of the Study:

• To introduce and investigate the concept of practically predefined-time stabilization in probability (PPDTSP) for FMNNs.
• To develop a novel Lyapunov-type criterion for PPDTSP.
• To design a simplified control scheme for achieving PPDTSP under disturbances and RDAs.

Main Methods:

• Introduction of a novel Lyapunov-type criterion for PPDTSP.
• Development of a simplified, practically predefined-time control scheme.
• Mathematical analysis and theoretical derivations for stability guarantees.
• Numerical simulations for validation.

Main Results:

• A novel Lyapunov-type criterion for PPDTSP is proposed, which is more general than existing criteria.
• A simplified control scheme is designed to achieve PPDTSP for FMNNs under stochastic disturbances and RDAs.
• The proposed method ensures stabilization within a practically predefined time.
• Special cases of predefined-time stabilization in probability (PDTSP) are derived.

Conclusions:

• The proposed PPDTSP concept and criterion effectively address the stabilization of FMNNs under challenging conditions.
• The developed control scheme provides a practical solution for enhancing the resilience of FMNNs.
• Numerical simulations confirm the validity and effectiveness of the theoretical findings.