Data underlying the conference proceeding: Decoupled Hull Resistance Prediction: A Computationally Aware and Physically Plausible Approach for Data-Driven Surrogates

This dataset supports the publication <em>Decoupled Hull Resistance Prediction: A Computationally Aware and Physically Plausible Approach for Data-Driven Surrogates</em>. It contains simulated hydrodynamic resistance data for systematically varied ship hull forms—created using Free-Form Deformation (FFD)—and generated via high-fidelity Computational Fluid Dynamics (CFD) simulations. The objective of the research is to develop and benchmark surrogate models that accurately predict total resistance across different hull geometries and operating conditions, while ensuring physical plausibility and generalization. The dataset includes geometric, hydrostatic, and stability parameters, operating conditions (e.g., Froude number), and the corresponding resistance outputs. It is intended to support studies in surrogate modeling, generalization under data scarcity, and shape–performance relationship analysis in marine hydrodynamics.

本数据集为论文《解耦船体阻力预测：面向数据驱动代理模型的计算感知与物理可信方法》提供支撑。该数据集包含经系统参数化调整的船体外形的模拟水动力阻力数据，这些外形通过自由形状变形（Free-Form Deformation, FFD）生成，并经由高保真计算流体动力学（Computational Fluid Dynamics, CFD）仿真得到。该论文的研究目标是开发并对标代理模型（surrogate model），使其能够在不同船体几何外形与运行工况下精准预测总阻力，同时确保模型具备物理可信性与泛化能力。本数据集涵盖几何参数、静水力学参数与稳性参数，运行工况（例如弗劳德数（Froude number）），以及对应的阻力输出结果。本数据集旨在支撑代理模型建模、数据稀缺场景下的泛化研究，以及船舶水动力学领域中的外形-性能关联关系分析等相关研究。