Introducing Uncertainty in Multidiscipline Ship Design

Mathematical formulations of reliability‐based design and robust design have been developed for addressing the presence of uncertainty during optimization. When reliability is introduced in the design process, the constraints that are influenced by uncertainty are converted into probabilistic constraints. This ensures that the optimal solution will satisfy the constraints in the presence of uncertainty within a prescribed reliability level. Robustness is introduced by modifying the objective function to depend on the mean and variance of the response of the objective function. This paper presents a formulation for applying reliability‐based design and robust design within a network of parallel optimizations in order to account for uncertainty in multidiscipline design optimization. The method for multidiscipline design optimization under uncertainty is applied to a conceptual ship design. The model of a bulk carrier from the literature is used in this analysis. The optimization results are compared with previous data and they demonstrate that reliable and robust optimization can be applied effectively to ship design, and that it is important to consider uncertainty during the multidiscipline ship design process.