Exponential Stability Analysis of Time-Varying Delay Neural Networks by the Creation of Experimental Lyapunov Functionals

Abstract

This study examines neural network systems exponential stability, a critical characteristic that guarantees the bounded ness and long-term convergence of network states. Using Linear matrix inequality and Lyapunov functional methods, we prove sufficient conditions for exponential stability. The system's state trajectory converges exponentially to an equilibrium point under suitable conditions, as demonstrated by the stability criteria derived from the network's weight matrices, activation functions, and time delays. Our results are validated by theoretical proofs and numerical simulations, showing the usefulness of the findings for control systems, computational neuroscience, and deep learning architectures.