Event-Triggered Formation Control of Unknown Autonomous Underwater Vehicles Subject to Arbitrarily Large Time-Varying Communication Delays

This paper presents a novel dual-loop event-triggered control framework designed to facilitate the formation control of unknown autonomous underwater vehicles (AUVs) operating under the constraints of arbitrarily large time-varying communication delays. The control scheme is executed in two distinct phases. Initially, an event-triggered distributed predictor is developed to estimate the position reference for each AUV, mitigating the impact of large, time-varying communication delays. Subsequently, a reinforcement learning-based approach is employed to derive an event-triggered optimal control policy for nonlinear autonomous underwater vehicle subject to unknown dynamics. The control performance of the proposed formation controller is analyzed and the Zeno-behaviors of the triggering functions are excluded. The effectiveness of the proposed control strategy is substantiated through comparative simulation studies, demonstrating its superiority over existing methods in terms of formation performance.