General criteria for asymptotic and exponential stabilities of neural network models with unbounded delays (Q555357)

This paper studies a general class of Cohen-Grossberg-type autonomous neural network models with infinite distributed delay: \[ \dot{x}_i(t) = - a_i(x_i(t)) \left[b_i(x_i(t)) + \sum_{j=1}^n\sum_{p=1}^P f_{ij}^{(p)}\left(\int_{-\infty}^0 g_{ij}^{(p)}(x_j(t+s)) d \eta_{ij}^{(p)}(s)\right)\right],\,\, i=1,\cdots,n. \] This model is treated as a particular case of a much more general family of DDEs of the form \[ \dot{x}_i(t) = - \rho_i(t,x_t)[b_i(x_i(t)) + f_i(x_t)],\quad i=1,\cdots,n, \] where \(\rho_i, b_i, f_i\) are continuous real functions, \(\rho_i\) are positive functions, and \(x_t\) is defined by \(x_t(s) = x(t+s)\) for \(s\leq 0\). The authors give sufficient conditions for the global asymptotic and global exponential stability of an equilibrium point. The criteria are simple to verify and do not involve the use of Lyapunov functionals, and are directly applicable to most of the autonomous neural network models with unbounded delay investigated in the recent literature.