ABSTRACT
This article introduces a multi-disciplinary optimization method for early-stage ship design. The limited knowledge available about a vessel in the early stages of its development limits the applicability of traditional optimization methods in such stages. This epistemic uncertainty stems from several sources, three of which are considered in this article. First, the uncertainty caused by limited information of a vessel model which is not fully defined. Second, the uncertainty which stems from the difficulty identifying precise design performance targets and comparing designs with similar performance. The third source of uncertainty relates to the difficulty identifying precise bounds on the validity of analysis tools. A multi-disciplinary optimization method is introduced which uses type-2 fuzzy logic to model the vague information inherent to early-stage design and incorporate human expertise directly into the optimization to handle this uncertainty. The method is illustrated on the optimization of a planing craft with respect to resistance and seakeeping.
