A CFD Study of Hydrodynamic Characteristics of an Autonomous Underwater Vehicle in Calm Water

Authors:

Abstract
In this study, inspired by the lateral-line sensing in fish behavior, the hydrodynamics and pressure-based motion state estimation of a full-submerged AUV body was numerically investigated. The unsteady Reynolds-averaged Navier-Stokes (URANS) equation with the k – ω SST turbulence model was solved using the commercial computational fluid dynamics (CFD) software Star-CCM+. First, a series of pre-planned trajectory simulations in open water conditions were performed, including straight-line, turning and time-varying speed profiles (sinusoidal trajectory cases), to examine the pressure and velocity distributions along the AUV hull. The complex local flow near the AUV body and their influence on the portstarboard pressure variation were discussed. Additionally, on the basis of this database, two regression-based motion state detection models, a pressure variation model (PVM) and a Gaussian process regression (GPR) model, are examined to estimate the advancing speed and drift angle from pressure measurements. Second, static drift and pure sway simulations are carried out in a near-wall configuration, and the resulting near-wall pressure measurements are used to develop two regression-based wall-distance detection models that estimate the vehicle’s drift angle and distance to the wall. It was confirmed that present wall distance detection model could well predict the motion state of AUV and the distance from the wall.