Model-Based Synthetical Optimization Analysis on Navigation Performance of Unmanned Surface Vehicle

Abstract

Unmanned Surface Vehicle (USV) has very complex forces and states when sailing at sea. While the optimization of USV via regression equations usually has a lower precision. Rapidity and seakeeping tests have been conducted based on a USV model of JUST. The rapidity test is mainly ship model resistance test and the seakeeping tests include wave added resistance test, hydrostatic pitching test, pitching and heaving test under the wave. Response surfaces fitting vectors of rough water resistance, pitching significant value, heaving significant value is established while using VC++ language to write second-order response surface fitting programs after dimensionless conversion of test data. Mathematical model of USV navigation performance synthetical optimization is proposed, including systems of rapidity, maneuverability and seakeeping. A layered parallel genetic algorithm (L-P-GA) optimization program is written in VC++ language, which can optimize a variety of conditions with different design speed and displacements. The results indicate that the optimization method has a higher precision and it can provide an opportunity to optimize the navigation performance of USV effectively.