Novel robust stability criteria for uncertain stochastic Hopfield neural networks with time-varying delays (Q2371021)

Stochastic robust stability of a class of \(n\)-dimensional stochastic Hopfield neural networks for the neural state vector \(x\) \[ dx(t) = [-A(t) x(t) + M(t) f(x(t-\tau (t)))] \,dt + [H_0(t) x(t) + H_1(t) x(t-\tau (t))] \,dW(t) \] driven by a \(m\)-dimensional Wiener process \(W\) with time-varying delays \(\tau (t)\), connection weight matrices \(M(t)\) and norm-bounded parameter uncertainties of \(A\), \(M\), \(H_0\) and \(H_1\) is investigated. \(f\) with \(f(0)=0\) represents the neuron activation function which is supposed to be bounded and Lipschitz continuous with constant Lipschitz-matrix \(G\). The time-delay factors \(\tau\) are unknown and time-varying with known bounds. Based on Lyapunov-Krasovskii functional and stochastic analysis approaches, some new stability criteria are presented in terms of linear matrix inequalities (LMIs) to guarantee the delayed neural network to be robustly stochastically asymptotically stable in the mean square for all admissible uncertainties. Numerical examples using MATLAB illustrate their approach, in particular the effectiveness and less conservativeness of the developed techniques. To prove the main results, the concepts of Schur complement and scalar-product inequalities involving vector-valued integrals are exploited.