Attraction region and stability analysis of Cohen-Grossberg neural networks

In this paper, exponential stability (local exponential stability and global exponential stability) is analyzed for Cohen-Grossberg neural networks. Simple sufficient conditions in form of matrix measure are derived for ascertaining the local exponential stability or the global exponential stability for Cohen-Grossberg neural networks. Moreover, for locally exponentially stable equilibrium point, its attraction region is estimated; for the globally exponentially stable equilibrium point, its convergence rate is assessed. The proposed results can be used for the evaluation of fault-tolerance capability and the estimation of exponential decay rate of the networks, hence they are of practical significance for designing Cohen-Grossberg neural networks.