Event-Triggered and Intermittent Emulation of Safety-Critical Controllers for Nonlinear Systems

This paper explores the emulation of safety-critical controllers for nonlinear systems within the constraints of limited computational resources and system data. We introduce innovative dynamic event-triggered and intermittent control strategies to digitally implement any given safety-critical controller constructed using a high-order control barrier function and its resultant Lyapunov function. These strategies simultaneously ensure forward invariance, asymptotic convergence, and prevent the Zeno phenomenon while providing an explicit characterization of the minimal dwell time. Our approach offers significant advantages in terms of resource efficiency compared to continuous-time safety-critical controllers. Additionally, we develop an extended framework to address the same control problem in the presence of unknown dynamics, enabling the application of a learning-based approach. We illustrate the effectiveness of the proposed design strategies through both an academic example and a practical problem in adaptive cruise control.