Keywods
FEGate for ship (FFS), pSeven, machine learning, approximate model, Nastran, structural analysis, API (Application Programming Interface).

Abstract
Ingeneral, when performing structural strength analysis of ships, FEM models arecreated and analyzed with commercial tools and solvers.
This process oftenrequires iterative tasks such as modifying the existing model for additionalanalysis of reinforced structures. This process is very time consuming andinefficient, and there is a great demand for automation.
Inthis paper, we automate modeling tasks using FEGate for ship's API andintegrate the entire process of structural strength analysis into one platformthrough pSeven, a PIDO (Process Integration and Design Optimization) tool.
The standardized integrated process is linked with Design of Experiment (DoE) toautomatically generate various analysis cases for evaluating the structuralstrength of the ship and accumulate feasible data.
As aresult, pSeven's machine learning algorithm was trained on the accumulated datato propose how to use a predictive model that can be evaluated in real time forthe structural strength analysis of ships that require repetitive manual tasks.