工程设计学报2026，Vol.33Issue(2)：204-212,9.DOI:10.3785/j.issn.1006-754X.2026.05.164
基于Kriging代理模型的仿海豚AUV外形优化设计
Optimization design of dolphin-inspired AUV shape based on Kriging surrogate model
摘要
Abstract
Aiming at the contradiction between low resistance and large volume in the shape design of autonomous underwater vehicles(AUVs),an optimization design method taking dolphins as bionic objects is explored to enhance the hydrodynamic performance and mission payload capacity of AUVs.Firstly,nine segments of Myring-type curves were used to parameterize and fit the dolphin contour,thereby establishing a three-dimensional AUV geometric model.Based on the Reynolds-averaged Navier-Stokes(RANS)equation and the standard k-ε model,the total resistance and envelope volume of the initial AUV were obtained through computational fluid dynamics(CFD)simulation.Subsequently,the optimal Latin hypercube sampling method was utilized to generate sample points,and Kriging surrogate models describing the mapping relationship between the total resistance and envelope volume of the AUV and design variables were constructed.Finally,with the objectives of minimizing total resistance and maximizing envelope volume,the Pareto optimal solution set was solved using NSGA-II(non-dominated sorting genetic algorithm-II).After optimization,the total resistance of the AUV decreased by 5.74%and the envelope volume increased by 5.87%at a navigation speed of 2 m/s.Flow field simulation analysis indicated that the optimized shape flattened the pressure gradient at the AUV tail,reducing the pressure difference resistance by 12.56%.At the same time,the velocity gradient at the AUV tail decreased,effectively inhibiting boundary layer separation.The resistance composition showed that the reduction in pressure difference resistance was the main reason for the decrease in total resistance.The horizontal stability index GH>0 indicated that the optimized AUV had dynamic stability.The multi-objective optimization method that integrates parametric modeling,CFD simulation,Kriging surrogate model and NSGA-II provides a reference for the shape optimization of underwater vehicles.关键词
自主水下航行器/仿生设计/计算流体力学/Kriging代理模型/多目标优化Key words
autonomous underwater vehicle(AUV)/bio-inspired design/computational fluid dynamics(CFD)/Kriging surrogate model/multi-objective optimization分类
交通工程引用本文复制引用
唐军,邱东旭,谢远辉..基于Kriging代理模型的仿海豚AUV外形优化设计[J].工程设计学报,2026,33(2):204-212,9.基金项目
国家自然科学基金资助项目(51864015) （51864015）
