Autonomous Surface Vehicle, Systems Identification, Standard Maneuvers
Abstract
Modeling the dynamics of surface vessels presents a multifaceted challenge, stemming from intricate mechanical factors and hydrodynamic phenomena. The conventional identification maneuvers often diverge significantly from standard operations, impacting both the quality of the model and its learning process. This research is centered on delineating the dynamic model of the Autonomous Surface Vehicle (ASV) AeroCat by employing standard operating maneuvers and leveraging the Bat Algorithm (BA) as an optimizer. Through successive parametric estimation stages, the efficacy of the models was evaluated, revealing a commendable grasp of the ASV dynamics. The adoption of a multi-stage estimation approach led to a noteworthy reduction in Mean Squared Error (MSE) by 40.48% compared to a single-stage method. This preliminary investigation underscores the potential of standardized maneuvers in yielding high-quality models, warranting further exploration due to their simplicity and promising outcomes.
