An Improved Exponential Stability Criterion of Generalized Neural Networks With Intermittent Large-Delay Periods

This paper focuses on the exponential stability analysis of generalized neural networks (GNNs) with time-varying delays, where the time delays include several intermittent large-delay periods (LDPs). First, the improved piecewise augmented Lyapunov-Krasovskii functional (LKF) candidate is constructed based on the delay-dependent state information. Then, some advanced inequalities are adopted to estimate the derivative of LKF candidate. Combining with switching techniques, an exponential stability criterion with less conservativeness is proposed. At last, a numerical example is used to verify the advantage of the proposed criterion.