Non-fragile state estimation for second-order memristive neural networks with unbounded time-varying delays

摘要

Without converting the second-order memristive neural networks (SMNNs) into the usual first-order systems, this article works on the non-fragile state estimation of SMNNs with unbounded time-varying delays. The memristor is treated as uncertainty by recommending some measurable functions. According to Lyapunov functional approach, inequality techniques, and combining Barbalat lemma, a new Lyapunov functional is proposed to straightway discuss the asymptotic stability for the error system, and several sufficient conditions are obtained to assure the existence of the preset estimators. The nonfragile estimators are designed and several explicit descriptions of the estimators are given by solving linear matrix inequalities. To overcome the uncertainty of state variables, the problem of parameter mismatch is as well considered. Finally, two examples are given to verify the availability of the results.