Organization Of Ship Design Information Using Cramer’s V And Genetic Algorithm
Year: 2023
Editor: Harold (Mike) Stowe; Tyson R. Browning; Steven D. Eppinger; Jakob Trauer; Christopher Langner; Matthias Kreimeyer; Ola Isaksson; Massimo Panarotto; Arindam Brahma
Author: Yusuke Sato (1), Kazuhiro Aoyama (1), Hideaki Ozawa (1), Kenji Takahashi (2), Shingo Maeda (3), Naoki Hirai (3)
Series: DSM
Institution: 1: The University of Tokyo, Japan; 2: Imabari Shipbuilding Co., Ltd.; 3: Nihon Shipyard Co., Ltd.
Page(s): 068-076
DOI number: 10.35199/dsm2023.08
Abstract
In ship design, existing specifications are repeatedly modified in response to shipowner requests. In this process, there are components that are simultaneously changed for design-related reasons, but the relationships among components have been established based on the tacit knowledge of designers owing to a large number of ship components. In this study, we identify specifications that change simultaneously by using Cramer's V. By leveraging the macro-level granularity, the designers can perform efficient checks and create a Design Structure Matrix (DSM). Additionally, the components are grouped by performing DSM clustering, which reflects whether specification changes occur within or outside the department. A case study is conducted using the actual specifications of a merchant’s vessel to demonstrate the effectiveness of the proposed method. Consequently,accurate relationships among components are extracted, and the departments that must collaborate with each other are identified.
Keywords: Shipbuilding, specification change, Cramer’s V, DSM clustering