Genetic Algorithm-Based Clustering Method to Formulate Standard Specifications for Merchant Ship Preliminary Design

In the preliminary design of merchant ships, shipbuilders generally modify some of the standard specifications to fulfill shipowner needs, which is time-consuming owing to the complex techno-economic constraints of ship design. Therefore, an appropriate standard specifications formulation method is necessary to improve the efficiency of the preliminary design. In this study, we performed genetic algorithm-based clustering to determine the subtypes of a specific type of merchant ship and formulated the standard specification for each subtype. To demonstrate the effectiveness of the proposed method, experiments were performed using 98 specification documents to formulate the standard specifications. The results showed that feature relations among each determined subtype were significantly simpler than those of the main type; thereby, the formulated standard specifications were desirable in the preliminary design of merchant ships.