Robust stability for stochastic Hopfield neural networks with time delays (Q867963): Difference between revisions

The Hopfield neural network is described by a stochastic delay differential equation of the form \[ dx(t)=(-Ax(t)+W l(x(t-h)))dt+(Cx(t)+Dx(t-h))dw(t). \] Here \(A,W,C,D\) are matrices with certain properties, \(l\) is a Lipschitz continuous neuron activity function and \(w\) is a Brownian motion. It is shown that certain explicit matrix inequalities imply that the equilibrium solution is robustly, globally, asymptotically stable in the mean-square sense.