Hull form reliability-based robust design optimization combining polynomial chaos expansion and maximum entropy method

In the ship design optimization process, the neglect of the unavoidable uncertainty of parameters in the actual navigation and experimental observations, may lead to a bad result with some hidden dangers in practical applications. Considering the influence of the uncertainty, a new and effective hull form reliability-based robust design optimization (RBRDO) framework, including the hull form modification module and RBRDO module, has been developed and tested in the present work. Radial basis function method is utilized as the parametric hull surface modification technique to generate a series of smooth hull forms while combining polynomial chaos expansions (PCE) method with maximum entropy method (MEM) to conduct the uncertainty analysis for the prediction of the mean and the standard deviation of the objective and the failure probability of constraints. To verify the validity of the method, hull form design optimization of the bow of KCS model is implemented under the influence of the uncertainty. Numerical results indicate that the proposed RBRDO framework is effective compared with traditional Monte Carlo method. Meanwhile, compared with traditional DO case, RBRDO case has higher adaptability to the environmental uncertainty with the lower failure probability, which ensure the robustness and reliability of the optimal hull form.