水下航行体推进器安装位置多目标优化设计

Multi-objective Optimization Design of Thruster Installation Position for Underwater Vehicles

摘要:

针对水下航行体推进器布局的多目标优化问题，以泵喷效率、航行体阻力和泵喷消耗功率为优化目标，采用代理模型方法对推进器的纵向安装距离和横向安装距离进行了优化设计。采用面心复合设计 （FCCD）与拉丁超立方抽样（LHS）结合的方式确定初始设计节点。通过数值模拟获得初始设计点目标响应数据，采用克里金（Kriging）、广义回归（PRS）、径向基（RBNN）、Shepard（SHEP）和 PWS 混合加权代理模型对目标变量进行拟合，并基于 Pareto 多目标优化方法优化推进器安装位置参数。研究表明，PWS 代理模型具有最低的拟合误差，优化后的推进器安装位置显著提高了泵喷效率，同时有效降低了航行体阻力和泵喷消耗功率，提升了推进系统的整体性能。另外，典型工况的数值结果与 PWS 代理模型预测结果较吻合，PWS 代理模型优化方法可为水下航行体推进系统布局提供可靠方案。

Abstract:

To address the multi-objective optimization problem of thruster layout for underwater vehicles，this study takes pump-jet efficiency，vehicle resistance，and pump-jet power consumption as optimization objectives. A surrogate model method is employed to optimize the longitudinal and lateral installation distances of the thruster. The initial design points are determined using a combination of face-centered composite design（FCCD）and Latin hypercube sampling（LHS）. Numerical simulations are conducted to obtain the target response data of the initial design points. The Kriging，polynomial response surface（PRS），radial basis neural network（RBNN），Shepard （SHEP），and PWS hybrid weighted surrogate models are applied to fit the target variables. The thruster installation position parameters are then optimized based on the Pareto multi-objective optimization method. The results demonstrate that the PWS surrogate model exhibits the lowest fitting error. The optimized thruster installation position significantly improves pump-jet efficiency while effectively reducing vehicle resistance and pump-jet power consumption，thereby enhancing the overall performance of the propulsion system. Additionally，the numerical results under typical operating conditions are in good agreement with the predictions of the PWS surrogate model. The PWS surrogate model optimization method provides a reliable solution for the layout design of underwater vehicle propulsion systems.