Abstract:Asymptotic behavior is discussed for large-scale general neural networks with time-varying delays by employing the inequality techniques and the properties of nonnegative matrices. Some estimation results are obtained about the attraction domain of equilibrium states and sufficient conditions for asympotic stability of large-scale general neural networks with time-varying delays. Moreover, the exponential attraction domain and the exponential convergent rate of equilibrium states for large-scale general neural networks with time-varying delays are estimated when delays are bounded. These results can be used for evaluation of error-correction capability and synthesis procedure of the networks.

Attraction Domain of Large-Scale General Neural Networks with Time-Varying Delays