Multi-objective optimization of an autonomous underwater vehicle shape based on an improved Kriging model

Through numerical simulations, this article aims to minimize the direct sailing resistance and diving resistance and maximize the envelope volume of an autonomous underwater vehicle (AUV) with a Myring type shape. After determining the design parameters and resistance simulation methods, a parameterized analysis of the AUV shape is performed in CATIA software integrated with STAR-CCM + software to obtain an approximate model of the AUV shape. The multi-objective optimization is solved using an improved non-dominated sorting algorithm (INSGA-II). The effectiveness of INSGA-II is verified on common test functions and in AUV shape optimization models. The AUV shape-optimization model is based on an improved Kriging model that uses point sampling and point filling strategies. The Pareto solution set of the multi-objective optimization of AUV shape was better optimized to varying degrees from that of the initial solution. Partial solutions to all objective functions were optimized, verifying the effectiveness of the improved Kriging model and INSGA-II.