Data-driven formation control for underactuated unmanned surface vehicles with model uncertainty by frequency-adaptive event-triggered scheme

This article investigates the problem of data-driven formation tracking control under distributed event-triggered mechanisms for underactuated unmanned surface vehicles (USVs) subjected to unmeasurable velocities and model uncertainties. Specifically, to address the issue of communication efficiency among formation members, a frequency-adaptive event-triggered mechanism is proposed. This event-triggered mechanism enables the adjustment of triggering frequency based on the magnitude of errors and accounts for the occasional communication instability in marine environments, while ensuring a minimum inter-event time. Second, a distributed reference state monitor (DRSM) under the novel event-triggered mechanism is developed based on local position information from neighboring USVs. The DRSM enables each USV to estimate its desired state information without relying on global path information. Furthermore, a data-driven dynamic control strategy is proposed based solely on position information. A neural-based data-driven controller is designed to estimate model uncertainties and unknown control input gains using real-time and historical data, without requiring prior knowledge of system parameters. Finally, the effectiveness of the proposed approach is validated through rigorous stability analysis and numerical simulations.