Finite-Time Passivity for Coupled Fractional-Order Neural Networks with Multistate or Multiderivative Couplings

Fractional calculus can help to better model the dynamics of neural networks due to its hereditary and memory properties, thus a growing number of authors focus on the dynamical behaviors of fractional-order neural networks (FONNs) in recent years [1–18]. Specially, in view of the fact that passivity can effectively handle stabilization and stability, much attention has been paid to the passivity for FONNs [19–23]. By exploiting matrix theory and inequality techniques, Ding et al. [19] established several robust passivity criteria for two types of uncertain FONNs, that is, the cases with time-varying parameter uncertainty and with interval uncertainty. Sau et al. [20] formulated an asymptotic stability criterion for a delayed FONN by leveraging the Razumikhin fractional-order theorem, and derived a sufficient condition to guarantee the passivity for such network on the basis of the presented stability criterion.