Exponentially Decaying Memory Recovery Feedback Controller Design for Quasi-Synchronization of Complex-Valued Discrete-Time Neural Networks

In this article, exponential quasi-synchronization of nonidentical discrete-time complex-valued neural networks (CVNNs) is investigated in the sense of simultaneous existence of parameter mismatches. Under memory-based state-feedback controller, quasi-synchronization of CVNNs is examined directly via nondecomposition approach by introducing parameter-dependent reciprocal convex matrix inequality. To obtain less conservatism, an augmented Lyapunov–Krasovskii functional with delay-product-type term is constructed which includes the impact of delay variation information into account. Using the inequality technique, the sufficient conditions are derived in terms of linear matrix inequalities. Finally, numerical examples are provided with the simulation results to validate the derived theoretical conclusions. Furthermore, the comparative results show the efficiency of the obtained outcomes.