基于多目标进化算法和决策技术的螺旋桨优化设计研究  

Optimal Design of Propeller Based on Multi-objective Evolutionary Algorithm and Decision Technology

作　　者：杨路春[1,2] 杨晨俊[1] 李学斌[2] YANG Luchun;YANG Chenjun;LI Xuebin(State Key Laboratory of Ocean Engineering,Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration,Shanghai Jiao Tong University,Shanghai 200240,China;Wuhan Second Ship Design and Research Institute,Wuhan 430000,China)

机构地区：[1]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 [2]武汉第二船舶设计研究所,武汉430000

出　　处：《中国造船》2019年第3期55-66,共12页Shipbuilding of China

基　　金：海装十二五预研项目“XXXX螺旋桨水动力性能预报”(31511060101)

摘　　要：论文采用RNSGA-Ⅱ-SBJG优化算法,开展考虑兼顾推进效率、空泡、激振力及桨叶强度等要求的螺旋桨优化设计。该算法基于实数编码的非支配排序多目标遗传算法RNSGA-Ⅱ,并引入跳变基因算子以提升基因多样性、加快寻优进程。此外,还引入灰色关联度分析方法,通过计算Pareto解集与理想解的关联度,对优化解集进行评价排序。应用上述方法,对某集装箱船螺旋桨进行了船桨匹配和船机桨匹配情况下的优化设计。数值结果表明,Pareto解集分散良好,基于灰色关联度分析的优化解评价方法能够有效和合理地选取优化设计方案。Algorithm RNSGA-II-SBJG is a variant of the real-coded non-dominated sorting genetic algorithm(RNSGA)and is adopted to optimize propeller design with consideration of requirements in propulsive efficiency,cavitation,excitation force and blade strength.Simulated binary jumping gene(SBJG)operator is introduced to improve genetic diversity and expedite searching process.To determine the optimal design from the Pareto solution set,a grey relational analysis approach is employed.The proposed method is applied to optimize the design of a container-ship propeller under the requirement of hull-propeller matching and hull-engine-propeller matching respectively.Numerical results indicate that the Pareto solutions are well dispersed,and optimal solutions can be effectively identified by the grey relational analysis.

关 键 词：螺旋桨 设计 优化 RNSGA-Ⅱ-SBJG 跳变基因 灰色关联度分析 

分 类 号：U661.313[交通运输工程—船舶及航道工程]