Application of PSO and SVR to the optimization shape of bulbous bow

• Main Authors: Jyun-ChengLiou, 柳軍承
• Other Authors: Shih-An Yang
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/50049948959518989598

碩士 === 國立成功大學 === 系統及船舶機電工程學系 === 103 === The ship owners of marine transport have asked the shipyards to design the best energy saving ships at a design speed and draught. The shipyards must comply with this requirement as swift as possible. However, the computational fluid dynamics for designing the hull form is computer time consuming, especially combined with the optimization method. This stringent requirement has been currently one of tough challenges for the researchers and shipyards. This thesis develops a numerical optimization system of single and multiple objective by integrating PSO (particle swarm optimization), SVR (support vector regression), SHIPFLOW, seakeeping code, and NURBS code. PSO is used to solve the shape optimization problem of the bulbous bow, and SVR is used to approximating the CFD calculation. The author expects to develop a highly efficient and accurate numerical system for optimizing the bulbous bows. The outcome will enhance the shipyards’ design ability.