Stability analysis of fractional-order neural networks: An LMI approach

This paper analyzes the stability of fractional-order neural networks (FNNs) without and with delay by the fractional Lyapunov direct method and the fractional Razumikhin-type theorem, respectively. S-procedure is applied to handle the nonlinear constraints to obtain a wider parameter selection of the systems. For FNNs without delay, the improved less conservative conditions of the existence and uniqueness of the equilibrium point and the global Mittag-Leffler stability are all derived in the form of linear matrix inequalities (LMIs). Moreover, an LMI-based uniform stability condition of FNNs with time delay is established, which simplifies and extends some previous work. Finally, the validity of the presented results is indicated by some numerical examples.