Novel Criteria on H∞ State Estimation for Static Neural Networks With Hybrid Delays Using Improved Matrix‐Valued Polynomial Inequality

In this article, the issue of state estimation for static neural networks (SNNs) accompanied by hybrid delays is concerned. First, a modified Lyapunov‐Krasovskii functional (LKF) is established by introducing a cubic matrix‐valued polynomial regarding time‐delay and more information on nonlinear activation function. Second, to coordinate with the features of the newly proposed LKF, an improved third‐degree matrix‐valued polynomial inequality and its simplified form are applied to avoid complex manual calculations as done in some works. Third, based on the proposed LKF and inequality technique, some further enhanced state estimation criteria are achieved for guaranteeing the asymptotical stability of the concerned system and satisfying given performance. At last, two simulations examples are done to show the validity of the developed method.