Classification of Design Variables Based on Constraint Conditions and Structural Optimal Design - Application to Structural Optimal Design of Midship Section. Issue 1 (January 2014)

• Record Type: Journal Article
• Authors: Hirakawa, Shinichi; Maki, Mitsue; Kitamura, Mitsuru; Nonami, Ryota
• Is Part Of: Ship technology research. Volume 61:Issue 1(2014)
• Page Start: 34
• Page End: 47
• Publication Date: 2014-01
• Subjects: Structural Optimal Design -- Principal Component Analysis -- Genetic Algorithm; Shipbuilding -- Periodicals; 623.82072
• DOI: 10.1179/str.2014.61.1.003
• Languages: English
• ISSNs: 0937-7255

Abstract

Abstract : A structure optimization of midship section is under consideration in this paper. A double hull oil tanker is selected for optimization. Since there are many design variables describing the midship section, usually, design variables are divided into some groups concerning design and manufacturing. When the classification of design variables is not proper, there is a problem that the optimal solution cannot be obtained. Classification method of design variables proposed here is intended to overcome this drawback. Influence of design variables on constraint conditions is estimated based on principal component analysis and used for classifying design variables. The classification of design variables obtained by the proposed method is different from a conventional group. Selected problems show that the proposed method can produce a better solution than the conventional method.