A New Result on Stability Analysis of Recurrent Neural Networks with Time-Varying Delay Based on an Extended Delay-Dependent Integral Inequality

This paper studies the stability issue of recurrent neural networks (RNNs) with time-varying delay. Firstly, an extended delay-dependent integral inequality that contains more free matrices is presented, which is an extension of some existing delay-dependent integral inequalities. Secondly, by employing the extended delay-dependent integral inequality, a tight upper bound of the Lyapunov-Krasovskii functional (LKF) derivative is estimated, then a new criterion on stability analysis of delayed RNNs is obtained. Finally, simulation results are provided to verify the superiority of the presented method.