New reliable nonuniform sampling control for uncertain chaotic neural networks under Markov switching topologies

Abstract

This paper studies the stochastic exponential synchronization problem for uncertain chaotic neural networks (UCNNs) with probabilistic faults (PFs) and randomly occurring time-varying parameters uncertainties(ROTVPUs). To reflect more realistic control behaviors, a new stochastic reliable nonuniform sampling controller with Markov switching topologies is designed for the first time. First, by taking into full account more information on sawtooth structural sampling pattern, time delay and its variation, a novel loose-looped Lyapunov–Krasovskii functional (LLLKF) is developed via introducing matrices-refined-function and adjustable parameters. Second, with the aid of novel LLLKF and relaxed Wirtinger-based integral inequality (RWBII), new synchronization algorithms are established to guarantee that UCNNs are synchronous exponentially under probabilistic actuator and sensor faults. Third, based on the proposed optimization algorithm, the desired reliable sampled-data controller can be achieved under more larger exponential decay rate. Finally, two numerical examples are given to illustrate the effectiveness and advantages of the designed algorithms.