Event-Triggered Data-Driven Control of Autonomous Vehicle With Unknown System Model

This paper proposes an event-triggered data-driven controller online data-driven controller synthesis algorithm of autonomous vehicle with unknown system model. To overcome the difficulty of establishing models for autonomous vehicle system, a novel design framework of the dual integrator structure is constructed to transform the original autonomous vehicle system into an equivalent data model. Based on the constructed data model, a data-driven algorithm is developed to directly construct a stabilizing controller from real-time data without using the autonomous vehicle system model. Furthermore, this result is extended to an event-triggered control scheme to further solve the problem of the limited communication resources. A significant difference from the existing method is that an implementation method based on online data is proposed to avoid storing historical data, and the persistence of excitation (PE) condition used to ensure the convergence of the algorithm is relaxed into a milder interval excitation (IE) condition. The simulation results demonstrate the superiority of the theoretical results.