On the adequacy of Adams-Bashforth sampled-data models for characterizing complex underwater-vehicle dynamics with noisy measurements

Abstract

In this paper the adequacy of high order interpolation-based approaches to describe highly perturbed complex dynamics in discrete time was analyzed. The analysis establishes features of the approaches related to modularity, consistency with the model order and the sampling times, and accuracy in disturbed contexts with noisy measurements. A detailed study of the sensitivity of local prediction errors under a high signal-to-noise ratio is carried out with analytical expressions in dependence of physical coefficients of the vehicle. The different interpolation-based approaches were illustrated with simulations using from an AUV-like (Autonomous Underwater Vehicle) system with a few degrees of freedom (DoF), to a ROV (Remotely Operated Vehicle) model of 6 DoF with complex navigation paths.