Hull Shape Design Optimization with Parameter Space and Model Reductions, and Self-Learning Mesh Morphing

In the field of parametric partial differential equations, shape optimization represents a challenging problem due to required computational resources. this contribution, data-driven framework involving multiple reduction techniques is proposed reduce such burden. Proper orthogonal decomposition (POD) and active subspace genetic algorithm (ASGA) are applied for dimensional original (high fidelity) model an efficient based on property. The parameterization directly mesh, propagating generic deformation map surface (of object optimize) mesh nodes using radial basis function (RBF) interpolation. Thus, topology quality preserved, enabling application POD-based reduced order modeling techniques, avoiding necessity additional meshing steps. Model performed coupling POD Gaussian process regression (GPR) in fashion. validated benchmark ship.