New result on reliable H∞ performance state estimation for memory static neural networks with stochastic sampled-data communication

This work examines the H(infinity )performance state estimation problem for memory static neu- ral networks (MSNNs) with reliable state feedback stochastic sampled-data control (SSDC). The purpose of presenting this study is to determine whether the H(infinity )performance and criteria with less conservatism for stability could be gained by SSDC for MSNNs or not. Firstly, we suppose that the sampling interval values follow Bernoulli distribution and the probability of occurrence are teadfast constant, then generalize it to a more universal form. Secondly, on basis of considering the sampling input delay and its sawtooth structure characteristics, a modified augmented Lyapunov-Krasovskii functional (LKF) is constructed on account of the free-matrix-based integral inequality (FMBII) together with generalized free-weighting-matrix (GFWM) inequality, which can reduce the conservatism of H(infinity )performance criteria. Thirdly, the expected estimator gain matrix can be designed in the light of the solution to linear matrix inequalities (LMIs). Finally, an numerical example is given to check the superiority of the proposed MSNNs control design technique. (C) 2019 Elsevier Inc. All rights reserved.