基于RBF神经网络的渔船尾部型线两目标优化  

Two-objective Optimization of Fishing Boat Stern Form Line Based on RBF Neural Network

作　　者：韩翔希[1,2] 张恒 冯志强[1,2] 邱昂 余建星[1,2] 符妃 HAN Xiangxi;ZHANG Heng;FENG Zhiqiang;QIU Ang;YU Jianxing;FU Fei(Qinzhou University Guangxi Engineering Technology Research Center of Marine Digital Design and Advanced Manufacturing ,Guangxi Qinzhou 535099,China;Qinzhou University Qinzhou Key Laboratory of Marine Advanced Design and Manufacturing Foundation,Guangxi Qinzhou 535099,China;Wuchang Shipbuilding Industry Group Co.,Ltd.,Wuhan 430060,China;Guangdong Shipping Science Research Institute,Guangzhou 510000,China)

机构地区：[1]钦州学院广西船舶数字化设计与先进制造工程技术研究中心,广西钦州535099 [2]钦州学院钦州学院钦州市船舶先进设计制造重点实验室,广西钦州535099 [3]武昌船舶重工集团有限公司,武汉430060 [4]广东省航运科学研究所,广州510000

出　　处：《船舶工程》2018年第10期35-39,58,共6页Ship Engineering

基　　金：钦州市科学研究与技术开发计划项目:北部湾先进远洋渔船船体型线优化技术研究(20164405);北部湾先进远洋渔船水动力性能优化技术研究(2018KY0600);钦州学院高级别培育项目:海洋立管涡激振动机理及其抑制研究(2016PY-SJ08);广西重点学科船舶与海洋工程资助;广西船舶数字化设计与先进制造工程技术研究中心课题资助;钦州市船舶先进设计制造重点实验室课题资助.

摘　　要：为了开发节能、减排的渔船新船型,提出一种基于仿真设计的优化方法,对一艘灯光渔船的艉部型线进行优化。采用三角变换法修改船型的几何形状,将优化拉丁超立方试验设计生成的样本方案用于RBF神经网络模型的构建,最后以总阻力性能和桨盘面伴流不均匀度为优化目标,采用NSGA2对该渔船船型进行两目标优化,结果表明该优化方法可以用于渔船新船型的开发。To develop,new fishing boats for energy-saving and emission-reducing,the paper proposes an optimization method based on simulation design to optimize the stern form lines of a light fishing boat.The triangular transformation method is used to modify the geometry of the ship.The samples generated with Optimal Latin hypercube method is used to construct the RBF neural network model.Finally,with the objectives of total resistance performance and the wake uniformity on the propeller disk,the NSGA2 is used to do the two-objective optimization of the hull form of the fishing boat.The results show that the optimizing method can be used for the development of new fishing boats.

关 键 词：RBF神经网络 船型优化 渔船 总阻力性能 伴流均匀度 

分 类 号：U661.31[交通运输工程—船舶及航道工程]