Unmanned underwater vehicles (UUVs) are essential tools for marine exploration, research, and surveillance. Accurate state estimations are critical for effective navigation, but conventional methods, such as Doppler velocity logs (DVLs) and inertial navigation systems, are expensive and vulnerable to environmental conditions. Inspired by the biological lateral line system in fish, this study proposes an enhanced pressure variation model (PVM) that estimates the velocity and drift angles using the data from pressure sensors. The improved model introduces unified regression coefficients and accounts for nonlinear flow effects, reducing the reliance on motion-specific parameters and increasing the adaptability across various maneuvering conditions. The model was validated by conducting extensive computational fluid dynamics (CFD) simulations across multiple motion scenarios. The enhanced PVM achieved high estimation accuracy while maintaining robustness under different dynamic conditions. The contributions of this study include the following: (1) a refined estimation framework using a unified coefficient model, (2) a low-cost, environment-resilient alternative to traditional systems, and (3) verified reliability through CFD-based performance evaluations. Future work will focus on experimental validation, extending the performance to large-angle motions, and developing a real-time data processing module to enable in situ application. This approach supports more autonomous and reliable UUV navigation for marine robotics and underwater missions.

When a submarine moves near the free surface, the lift and drag characteristics that act on it are different compared to when in deep water; for example, waves on the free surface cause submarine motions that are not seen in deep water conditions and lead to changes in speed, fuel efficiency, safety, and maneuverability. To accurately predict the maneuverability of a submarine, it is necessary to consider how both maneuvering and seakeeping performance are affected by free-surface effects during the design stage. In this study, the unified maneuvering and seakeeping dynamic model is proposed. In the maneuvering performance analysis, hydrodynamic forces in the horizontal plane were calculated using STAR-CCM+. In the seakeeping performance analysis, the 6-DOF motions of the submarine and the mean wave drift forces in the horizontal plane were calculated using Ansys AQWA. Since the maneuvering motion component has a relatively long period and the seakeeping motion component has a relatively short period, the unified maneuvering and seakeeping dynamic model for a submarine moving near the surface was established using a two-time-scale approach. Using the established unified maneuvering and seakeeping dynamic model, turning circle simulations were performed in both calm water and in waves. In calm water, there were no significant differences as depth was varied. However, in irregular waves, significant differences were found in the trajectories and motion variables as depth varied. These findings underscore the necessity of accounting for sea surface conditions when operating near the free surface to ensure safety and avoid potentially hazardous scenarios during submarine operations.

The free surface resolution in current load calculations increases costs significantly when using computational fluid dynamics (CFD), even though the small Froude number of a dynamic positioning (DP) vessel suggests minimal wave effects. This study examined the impact of including and excluding the water-air interface in CFD simulations to identify a more efficient method for load computation. For each current condition, this study analyzed two virtual domains: one modeled the free surface using the Volume of Fluid method, while the other treated the free surface as a symmetry plane based on the double body assumption. The Reynolds-averaged Navier-Stokes (RANS) algorithm with the Shear Stress Transport k-ω model was used, and the results were validated against the empirical formula and experimental data. These findings showed that the free surface had a negligible effect on the resulting hydrodynamic force at the targeted low Froude numbers. Despite this, implementing free-surface modeling costs a hundred times more than the simulations using the steady RANS approach. The substantial increase was mainly due to the low convergence rate of the unsteady solver for a low-propagating flow. The study enhanced the credibility of the steady RANS solver without free-surface resolution in numerically calculating the current load on a DP vessel.