上海交通大学 海洋工程国家重点实验室； 高新船舶与深海开发装备协同创新中心， 上海 200240
通讯作者:王刚成（1991-），男，河南省信阳市人，硕士生，主要研究方向为绿色智能船舶技术. E-mail: gangcheng111@163.com.通信作者：马宁，男，教育部****，****，博士生导师，E-mail: ningma@sjtu.edu.cn.基金资助:教育部重大专项课题（GKZY010004）Fast Collaborative Multi-Objective Optimization for Hydrodynamic Based on Kriging Surrogate Model
WANG Gangcheng，MA Ning，GU XiechongState Key Laboratory of Ocean Engineering; Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Shanghai Jiao Tong University, Shanghai 200240, China
摘要/Abstract
摘要： 引入Kriging代理模型，对船舶兴波阻力、垂荡和纵摇运动幅值进行了多目标快速优化.优化过程中，以Rankine源非线性势流理论和三维频域面元法为基础，利用非支配排序遗传算法(NSGA-II)并结合船体曲面修改函数对船体型线进行多目标优化设计，将高精度Kriging代理模型引入到优化进程中以解决迭代造成的耗时难题.数值分析结果表明：该优化方法能够使船型最优，在设计吃水、服务航速下有效减少航行阻力并提高耐波性能，且Kriging代理模型方法可以大幅提高船型优化效率.最后，基于雷诺平均(RANS)方程的计算流体动力学(CFD)方法对模型尺度λ为 31.599 的最优船型的阻力结果进行了对比验证，以证明优化结果的可靠性.
关键词: 多目标优化, Kriging代理模型, 阻力, 耐波性, 雷诺平均方程
Abstract: The wave drag, heave and pitch performance of the ship at design speed are optimized by introducing Kriging surrogate model. During the optimizing process, hull lines are optimized by utilizing the non-dominated sorting genetic algorithm and surface modification functions based on Rankine source method and 3-D Green function method. In addition, the high-precision Kriging surrogate model is applied to approximate optimization objectives in order to solve the time-consuming problem in the iterative process. Numerical results show that the present optimization approach can be used to optimize ship hull forms for reducing drag and improving seakeeping performance, and Kriging surrogate model can improve optimization efficiency considerably. Finally, the drag optimization results at model scale λ=31.599 are validated via CFD method based on Reynolds-averaged Navier-Stokes (RANS) equations.
Key words: multi-objective optimization, Kriging surrogate model, drag, seakeeping, Reynolds-averaged Navier-Stokes (RANS)
PDF全文下载地址:
点我下载PDF
