Lag Exponential Synchronization of Delayed Memristor-Based Neural Networks via Robust Analysis

Abstract

In this paper, the lag synchronization of delayed memristor-based neural networks (MBNNs) via robust analysis is studied. The MBNNs are neural networks closely related to the state variables. Therefore, the traditional linear feedback control may not achieve the goal of lag synchronization between the master system and the slave system. Under the definition of Filippov’s solution, we convert the varying weight coefficients of the MBNNs into interval perturbation which is the first time to consider positive real uncertainty and simultaneously avoid discussing the problem of parameter mismatch. Based on the Lyapunov–Krasovskii functional and an improved convex combination inequality, some new lag synchronization criteria are established in the form of linear matrix inequalities. Compared with some existing works, the robust analysis approach can improve the synchronization performance. Finally, numerical examples are provided to show the reliability and effectiveness of the results presented.