New results on exponential input-to-state stability analysis of memristor based complex-valued inertial neural networks with proportional and distributed delays

Abstract

The present work accumulates the Exponential input-to-state stability (EISS) criteria of memristor based delayed complex-valued neural networks (DMCNN) associated with an inertial term and time-varying delays. Here two varieties of time-varying delays are provided, namely proportional and distributed delays. In this study, the delayed memristor neural networks (MNN) is constructed on the basis of second order complex-valued space. In addition, the sufficient conditions are proposed to ensure the EISS by using the combination of non-smooth analysis, set-valued maps, Lyapunov-Krasovskii functional having double integral terms and Kirchhoff’s matrix tree theorem, moreover we employ Cauchy-Schwarz inequality & some inequality techniques. At the end of this work, the hypothesis has been established with an illustrative example along with the simulations.