In recent years, with the advancement of high-tech weapon systems and shipbuilding technical art improvement, the type of warfare has changed from massive displacement of warships, such as cruiser and destroyer, to the "small, quality, and powerful" such as Littoral Combat Ship (LCS). To implement the missions effectively in a period of peace/ warfare, the tasks assigned by the ships are more variously included as following: sea cruising, humanitarian rescue, sea defending, protecting fish boat and counter-terrorism. Therefore, most of the advanced countries have also developed Unmanned Surface Vessels (USV), which were applied for multi-function small boats, for example, application for hydrological observation, information reconnaissance, mine-hunting and shooting target on the sea. Therefore, this study would experiment in circulating water channel and applied Taguchi's DOE methods to discuss and compare with the resistance performance of monohull and multi-hull. The primary predicted values and parametric conditions for a minimum drag should be acquired by Taguchi's method first. Then, we applied a neural network to construct the nonlinear metamodel and genetic algorithm to search for the optimum factorial levels. The result shows that the two factors and their interaction has a significant effect on the drag of the multi-hull model. Finally, we would carry out the confirmation experiment to make sure the optimum condition is a robust design. In this study, we would use the ship model test and Taguchi dynamical analysis to acquire and validate the robust design of the small surface vessels configuration design.
展开▼