Finite-time peak-to-peak analysis for switched generalized neural networks comprised of finite-time unstable subnetworks

Abstract

This research is concerned with finite-time stability and peak-to-peak performance analysis for the discrete-time switched generalized neural networks (SGNNs) with time-varying delay. Compared with the reported results, each individual subnetwork of the SGNNs is considered to be finite-time unstable in the present study. To accomplish the anticipatory objective, the quasi-time-dependent Lyapunov–Krasovskii functional is constructed, and the associated sufficient conditions are simultaneously formulated to confirm that the disturbance-free SGNNs are finite-time stable when the subnetwork satisfies a certain switching time interval. In addition, a prescribed disturbance attenuation level is also achieved for the perturbed SGNNs in the sense of peak-to-peak performance. Finally, the provided simulation example corroborates the effectiveness and applicability of the established finite-time analysis framework in the absence of finite-time stable subnetworks.