Further results on H∞ state estimation of static neural networks with time-varying delay

Abstract

Abstract This paper deals with the H∞ state estimation problem for static neural networks subject to time-varying delays. First, an augmented Lyapunov-Krasovskii functional (LKF) is constructed for the use of the second-order Bessel–Legendre integral inequality. Second, by introducing some novel techniques to absorb the time-varying delay, the bounded real lemma (BRL) is expressed as matrix inequalities linearly dependent on the time-varying delay rather than in its quadratic form. By employing the linear convex approach, a less conservative BRL condition is derived for the estimation error system. Based on this condition, the state estimators can be calculated by solving a set of linear matrix inequalities. Finally, an example is used to illustrate the effectiveness of the proposed method.