Global exponential stability analysis of neural networks with an interval time-varying delay via an extended free-matrix-based double integral inequality

In the paper, we study exponential stability of neural networks with an interval time-varying delay. Firstly, we propose an extended free-matrix-based double integral inequality (FMDI) to estimate the double integral terms in the derivative of Lyapunov-Krasovskii functional (LKF). Secondly, we compare the extended FMDI and the FMDI and show that the former encompasses the latter as a special case. Finally, to show the advantages of the extended FMDI over Wirtinger-based double integral inequality (WBDI), we use the two different inequalities to investigate the exponential stability of delayed neural networks and derive new exponential stability criteria based on the same LKF. Moreover, the conservatism comparison of the criteria are illustrated through one numerical example.