考虑碎冰阻力和静水阻力的高效船型多目标优化方法  

Efficient ship hull multi-objective optimization method considering ice resistance and calm water resistance

作　　者：王世超 刘刚[1] WANG Shichao;LIU Gang(School of Naval Architecture Engineering,Dalian University of Technology,Dalian 116024,China)

机构地区：[1]大连理工大学船舶工程学院,辽宁大连116024

出　　处：《中国舰船研究》2024年第6期97-107,共11页Chinese Journal of Ship Research

基　　金：基础加强计划资助项目(2020-JCJQ-ZD-229-00)。

摘　　要：[目的]针对碎冰环境对船舶性能的影响,以及基于经验公式的传统优化方法在碎冰阻力优化方面的局限性,基于CFD&DEM方法提出一种精确优化碎冰阻力和静水阻力的船型优化方法。[方法]首先,基于CFD和CFD&DEM方法计算静水阻力和碎冰阻力,提出一种创新的混合多岛遗传算法(HMIGA),用于模拟真实环境下的碎冰场;然后,结合XGBoost模型建立高效代理模型,并执行NSGA-III算法进行优化求解;最后,以KCS标准模型为例进行验证。[结果]结果显示,优化后船型的碎冰阻力降低了10.58%,静水阻力降低了2.32%;优化船型所承受的峰值载荷更少,同时通过产生波浪推开浮冰,还可进一步降低冰阻力。[结论]所提方法综合考虑了流场和碎冰场的随机性对优化结果的影响,能更精确、有效地改善船舶的碎冰阻力与静水阻力。HMIGA算法和XGBoost模型的引入可提升方法的实际应用效果,从而为未来碎冰环境下船舶的优化设计提供指导。[Objective]To address the impact of ice-covered environments on ship performance and the limitations of traditional optimization methods based on empirical formulas for ice resistance,a precise ship design optimization method based on CFD&DEM is proposed to optimize both ice resistance and calm water resistance.[Methods]First,calm water resistance and ice resistance are calculated based on the CFD and CFD&DEM methods,and an innovative hybrid multi-island genetic algorithm(HMIGA)is introduced to simulate realistic ice fields.Next,an efficient surrogate model is established using XGBoost,followed by the execution of the NSGA-III algorithm for optimization.Finally,the method is validated using the KCS standard model.[Results]The results show that the optimized ship design achieves a 10.58%reduction in ice resistance and a 2.32%reduction in calm water resistance.The optimized ship experiences lower peak loads and further reduces ice resistance by generating waves to push away floating ice.[Conclusions]The proposed method comprehensively considers the effects of flow field and ice field randomness on the optimization results,leading to more accurate and effective improvements in ship ice resistance and calm water resistance.The introduction of HMIGA and XGBoost enhances the practical application of the method,providing valuable guidance for the future optimization design of ships operating in ice-covered environments.

关 键 词：船舶设计 船型优化设计 多目标优化 计算流体力学 离散元方法 混合多岛遗传算法 集成学习 

分 类 号：U661.31[交通运输工程—船舶及航道工程]