Exponentially finite-time dissipative discrete state estimator for delayed competitive neural networks via semi-discretization approach

This article is concerned with the investigation of finite-time boundedness and exponential (Q, S, R)-dissipative performance for a class of discretized competitive neural networks (CNNs) with time-varying delays. Initially, by employing the semi-discretization technique, a discrete analog of the continuous-time CNNs is formulated, which preserves the dynamical behaviors of their continuous-time counterpart. An appropriate state estimator is developed for the discretized CNNs so that the dynamics of the associated estimation error system attain finite-time exponential (Q, S, R)-dissipative performance. Further, to obtain a tighter summation bound, two novel weighted summation inequalities are proposed, which linearize the quadratic summable terms occurring in the finite difference of the considered Lyapunov-Krasovskii functional. Finally, to refine our prediction, an illustrative example is provided that demonstrates the sustainability and merits of the proposed method.

Automated summary

This article focuses on the investigation of finite-time boundedness and exponential (Q, S, R)-dissipative performance for a class of discretized competitive neural networks (CNNs) with time-varying delays. By using the semi-discretization technique, a discrete analog of the continuous-time CNNs is formulated and a state estimator is developed to achieve finite-time exponential (Q, S, R)-dissipative performance. Two novel weighted summation inequalities are proposed to obtain a tighter summation bound. An illustrative example is provided to demonstrate the sustainability and merits of the proposed method.