Stabilization of G-Stochastic Complex Networks via Asynchronous Hierarchical-Triggered Intermittent Control

For providing a more accurate description of probabilistic and model uncertainties, this paper studies the stabilization of G-stochastic complex networks (SCNs-G), where the normality assumption on noise is removed by accounting for non-Gaussian G-noise. Unlike prior intermittent event-triggered control (IE-TC), a novel asynchronous hierarchical-triggered intermittent controller (AHTIC) is designed, where the asynchronous self-triggered law determines the switching between intermittent control (IC) intervals and rest intervals, while the asynchronous dynamic event-triggered mechanism (E-TM) governs the sampling rule within the control intervals. The hierarchical-triggered feature enables the proposed AHTIC strategy to respond adaptively to unanticipated events while improving energy efficiency. Note that the sublinear expectation framework that the SCNs-G stems from, results in the existing IE-TC stabilization results inapplicable. Subsequently, a novel G-Lyapunov functional with assisting functions is constructed, and a new analytical technique is introduced to tackle the challenges arising from sublinear expectations. Specifically, the stability analysis relies on the estimation of sublinear expectations for several integrals relating to sampling errors, dynamic terms, and system states, which in turn enables the derivation of less conservative stability conditions. By means of Lyapunov functional method, and G-stochastic analysis techniques, two types of exponential stabilization criteria, including mean-square and quasi-sure exponential stability, for the control system are developed. Finally, the effectiveness of theoretical results is validated by numerical simulations on autonomous land vehicle networks. Note to Practitioners—Existing studies on IE-TC for stochastic systems are mainly conducted within the framework of linear expectations, and the design of IC strategies all relies on time-triggering mechanisms. This paper proposes an AHTIC strategy. Specifically, the trigger logic is organized into two layers: the first layer is an asynchronous self-triggered mechanism that determines the switching between active and resting intervals of IC; the second layer is an asynchronous event-triggered rule that governs the sampling mechanism during the active control intervals. This strategy exhibits significant advantages in improving control flexibility and reducing the consumption of communication resource. By virtue of this strategy, the stabilization of G-stochastic complex networks is investigated, employing the Lyapunov functional method and G-stochastic analysis techniques. Since G-stochastic complex networks are defined under the sense of nonlinear expectations, this poses considerable challenges to theoretical analysis. A novel analytical framework is proposed to establish exponential stability criteria. Finally, a simulation study on autonomous land vehicle networks is presented to validate the effectiveness of the proposed approach.