Dual-Channel Control Structure-Agnostic Event-Triggering Framework
Deniz Kurtoglu, Tansel Yucelen, Eloy García, Dzung M. Tran, David W. Casbeer · IFAC-PapersOnLine · 2025
Over the past decades, the feedback control domain has expanded beyond purely model-based regulators to include decision rules predicated on artificial intelligence algorithms, real-time optimization tools, or humans interacting with physical systems. Unlike classical control laws, whose structures are explicitly defined, these decision rules typically lack a closed-form representation. Consequently, traditional event-triggering schemes, which require knowledge of the structure of the control law for stability analysis, may fail to guarantee closed-loop stability when such decision rules govern the system. To address this scientific gap, the contribution of this paper is a novel dual-channel control structure-agnostic event-triggering framework. Specifically, the proposed framework operates independently of any specific control law representation, relies solely on open-loop system dynamics, and guarantees closed-loop stability. In addition, we perform a perturbation analysis to derive an analytical condition on the allowable uncertainty in the open-loop system model, where this condition quantifies the reliance on imperfect system knowledge. Finally, two illustrative numerical examples—one involving a human subject and one based on a reinforcement learning policy—are also provided to demonstrate the efficacy of the proposed framework.