DETERMINATION OF RELATIONAL CLASSIFICATION AMONG HULL FORM PARAMETERS AND SHIP MOTIONS PERFORMANCE FOR A SET OF SMALL VESSELS

Abstract

Data science for engineers is the most recent research area which suggests to analyse large data sets in order to find data analytics and use them for better designing and modelling. Ship design practice reveals that conceptual ship design is critically important for a successful basic design. Conceptual ship design needs to identify the true set of design variables influencing vessel performance and costs to define the best possible basic design by the use of performance prediction model. This model can be constructed by design engineers. The main idea of this paper comes from this crucial idea to determine relational classification of a set of small vessels using their hull form parameters and performance characteristics defined by transfer functions of heave and pitch motions and of absolute vertical acceleration, by our in-house software application based on K-Means algorithm from data mining. This application is implemented in the C# programming language on Microsoft SQL Server database. We also use the Elbow method to estimate the true number of clusters for K-Means algorithm. The computational results show that the considered set of small vessels can be clustered in three categories according to their functional relations of their hull form parameters and transfer functions considering all cases of three loading conditions, seven ship speeds as non-dimensional Froude numbers (Fn) and nine wave-length to ship-length values (lambda/L).