Filter Design of Delayed Nonlinear Discrete-time Markovian Neural Networks Systems with Missing Measurements

This paper presents a L-2-L-infinity filtering scheme for nonlinear discrete-time Markovian jumping neural networks with time delay under missing output measurements. By constructing appropriate Lyapunor-Krasovkii functional and utilizing some linear matrix inequality techniques, the mean square stability of the stochastic estimation error systems is guaranteed and a sufficient condition is established to ensure the given L-2-L-infinity, filtering performance. What's more, we provide the design approach of the filter when the delayed states in output measurements are involved or output measurements of the systems can be fully obtained. The gain matrix of the filter can be derived from the solution of a set of linear matrix inequalities. Finally, the simulation proves the availability of the proposed approach.