Improved Delay-Dependent Stability Analysis for Neural Networks with Time-Varying Delays

This paper considers the stability analysis problem of delayed neural network systems (DNNs). First, an appropri-ate Lyapunov-Krasovskii Functional (LKF) is chosen to obtain Linear Matrix Inequality (LMI) conditions, then applying appropriate relaxation techniques such as Finsler's lemma as well as a new augmented reciprocally convex inequality, to relax as much as possible the LMI-based stability conditions. Finally, to clarify the validity of the proposed approach, a numerical example is proposed to emphasize the improved reduction in conservatism delivered compared to other relevant recent studies.