Sample-Efficient and Surrogate-Based Design Optimization of Underwater Vehicle Hulls

TL;DR

This study finds that the Bayesian Optimization-Lower Condition Bound (BO-LCB) algorithm is the most sample-efficient optimization framework and has the best convergence behavior of those considered and shows that the DNN-based surrogate model predicts drag force on test data in tight agreement with CFD simulations.

Abstract

Physics simulations like computational fluid dynamics (CFD) are a computational bottleneck in computer-aided design (CAD) optimization processes. To overcome this bottleneck, one requires either an optimization framework that is highly sample-efficient, or a fast data-driven proxy (surrogate model) for long-running simulations. Both approaches have benefits and limitations. Bayesian optimization is often used for sample efficiency, but it solves one specific problem and struggles with transferability; alternatively, surrogate models can offer fast and often more generalizable solutions for CFD problems, but gathering data for and training such models can be computationally demanding. In this work, we leverage recent advances in optimization and artificial intelligence (AI) to explore both of these potential approaches, in the context of designing an optimal unmanned underwater vehicle (UUV) hull. Our study finds that the Bayesian Optimization-Lower Condition Bound (BO-LCB) algorithm is the most sample-efficient optimization framework and has the best convergence behavior of those considered. Subsequently, we show that our DNN-based surrogate model predicts drag force on test data in tight agreement with CFD simulations, with a mean absolute percentage error (MAPE) of 1.85%. Combining these results, we demonstrate a two-orders-of-magnitude speedup (with comparable accuracy) for the design optimization process when the surrogate model is used. To our knowledge, this is the first study applying Bayesian optimization and DNN-based surrogate modeling to the problem of UUV design optimization, and we share our developments as open-source software.

Figures (11)

• Figure 1: A typcial Myring Hull profile. The bow/nose (left) and stern/tail (right) have independently parameterized profiles, and the overall design is axisymmetric as indicated by the horizontal dashed line.
• Figure 2: Demonstrating the influence of $n$ and $\theta$ on the shape of the upper portion of a Myring hull profile. Larger $n$ and $\theta$ tend to increase the volume of the nose and tail, respectively.
• Figure 3: An integrated toolchain incorporating FreeCAD (parametric CAD modeling software) and OpenFOAM (CFD simulation software) in a Python environment (controlling the process flow and running optimizers and samplers) for CAD design generation, drag evaluation, and optimization.
• Figure 4: Meshing in OpenFOAM after blockMesh-based volumetric mesh generation and snappyHexMesh-based refinement. \ref{['fig:meshing:cross']} Cross-sectional view, \ref{['fig:meshing:oblique']} Oblique view. OpenFOAM performs adaptive meshing with a maximum number of allowable cells provided as a parameter by the user (in our experiments, 1 million).
• Figure 5: An example steady state flow, with the flow field colored by \ref{['fig:steadystateflowproperties:vf']} mean axial velocity (meters/second), \ref{['fig:steadystateflowproperties:pf']} pressure (Pascals).