A Data-Driven Reduced Order Optimization Approach for Cruise Ship Design

This Master Thesis describes the implementation of a data-driven shape optimization pipeline in a naval architecture application. We investigate the use of Reduced Order Methods (ROMs) in order to improve the efficiency and applicability of the techniques, also in an industrial setting. The above-mentioned pipeline is applied to a realistic cruise ship in order to reduce the total drag applied. We begin by defining the design space, generated by deforming an initial unoptimized shape in a parametric way using Free Form Deformation (FFD). The evaluation of the performance of each new hull is determined by simulating the flux via finite volume discretization of a biphase (water and air) fluid. In order to improve the efficiency of the simulation over a new shape, we use ROMs, in particular Proper Orthogonal Decomposition with Interpolation (PODI). Finally, a genetic optimization algorithm is used to explore the design space, using PODI for efficient evaluation of the flows. Other applications of ROMs are considered apart from the efficient resolution of the numerical problem, which are the approximation of dynamical systems via Dynamic Mode Decomposition (DMD) and reduction of the dimensionality of the parameter space.