Observer based guaranteed cost control for Markovian jump stochastic neutral-type neural networks

摘要

Based on observer framework, this paper examines the reliable sampled-data actuator fault controller for stochastic neural networks of neutral-type along with Markovian jump parameters and time-varying delays. In particular, the main intention of this work is to obtain the reliable state feedback guaranteed cost controller with satisfactory mixed H-infinity and passivity performance index. Particularly, a more generalized randomly occurring fault model is incorporated in the given quadratic cost function, where the faults of each and every actuator are represented by stochastic variables with some satisfactory probability conditions. By applying a proper Lyapunov-Krasovskii-functional along with neuron activation functions, a newly established conditions are captured to assure the stochastic stability of the considered model. Further an adequate performance index can be compared by solving the developed linear matrix inequalities. Furthermore, the effectiveness and the impact of the cost function based control scheme is tested through simulation results. (C) 2020 Elsevier Ltd. All rights reserved.