Event-triggered optimal line-of-sight formation control for underactuated USVs with unknown disturbance

This paper presents an event-triggered optimal line-of-sight (ETO-LOS) control scheme for multiple underactuated USVs with unknown dynamics and external disturbances. A two-stage Identifier-Actor-Critic (IAC) framework is combined with dynamic surface control (DSC) to achieve online optimal policy learning. A unified event-triggered mechanism operating on the norm of the total control vector is introduced, which enforces strictly synchronous updates of both channels so that the actual number of actuator actions corresponds exactly to the number of trigger events, achieving genuine resource savings at the actuator level. An inertia-matrix norm normalization term is further incorporated into the reference signal design to constrain the signal amplitude within a physically feasible range. Lyapunov-based analysis proves that all closed-loop signals are Semi-Globally Uniformly Ultimately Bounded (SGUUB) and Zeno behavior is excluded. Simulation results on the high-fidelity 'Dolphin 2' USV model verify the effectiveness and robustness of the proposed method.