Finite-time synchronization for fractional-order memristor-based neural networks with discontinuous activations and multiple delays

Abstract

This paper addresses the finite-time synchronization problem for fractional-order memristor-based neural networks (FMNNs) with discontinuous activations, in which multiple delays are considered. Firstly, on the basis of set-valued mapping as well as differential inclusions theory, the synchronization issue of drive-response systems is considered as the stabilization of the error system. Then, the state feedback controllers, which contain both discontinuous part and time-delayed part, are designed to analyze the finite-time synchronization of the concerned network model. Making use of the stability theorem of fractional-order systems with multiple time delays, some fractional derivative inequalities and comparison theorem, several sufficient criteria are established for confirming that the synchronization error of the concerned system can reach zero within a limited time. Additionally, the settling time can be optimized by adjusting controller parameter. Finally, the effectiveness of synchronization strategies is validated through the simulation results.