A new approach to non-fragile state estimation for continuous neural networks with time-delays

In this paper, the non-fragile state estimation problem is investigated for a class of continuous neural networks with time-delays and nonlinear perturbations. The estimator to be designed is of a simple linear structure without requiring the exact information of the activation functions or the time-delays, and is therefore easy to be implemented. Furthermore, the designed estimator gains are allowed to undergo multiplicative parameter variations within a given range and the non-fragility is guaranteed against possible implementation errors. The main purpose of the addressed problem is to design a non-fragile state estimator for the recurrent delayed neural networks such that the dynamics of the estimation error converges to the equilibrium asymptotically irrespective of the admissible parameter variations with the estimator gains. By employing a combination of the Lyapunov functionals and the matrix analysis techniques, sufficient conditions are established to ensure the existence of the desired estimators and the explicit characterization of such estimators are then given via solving a linear matrix inequality. Finally, a simulation example is used to illustrate the effectiveness of the proposed design method.