Dual-layer dynamic event-triggered formation maneuver control for underactuated autonomous underwater vehicles via edge vector-based method

Abstract

This study addresses formation maneuver control for autonomous underwater vehicles (AUVs) under conditions of limited network bandwidth, constrained actuation resources, and multiple uncertainties. Unlike the existing translation and scaling formation control, this study incorporates rotational motion to enable AUVs to navigate curved channels. A leader-first–follower structure is adopted where only the leaders have access to the formation maneuvering parameters. To address these issues, a dual-layer dynamic event-triggered formation maneuver control method is proposed. Specifically, a dynamic event-triggered mechanism (DETM) is employed to save communication resources and a predictor is used to estimate neighbor information during communication intervals. Next, a novel edge vector-based method, which calculates the edge vector between two neighbor positions, is applied to determine the virtual speed for followers. An extended state observer is introduced to estimate the unknown sideslip effect. Subsequently, a dynamic event-triggered control law is developed for AUVs to perform the formation maneuver. This law uses adaptive neural networks to approximate the composite disturbance caused by input constraints, environmental disturbances, and unknown model parameters, while the design DETM avoids unnecessary executions. Eventually, the system stability and the absence of Zeno behavior are demonstrated, and simulation results substantiate the effectiveness of the proposed control scheme.