基于改进GA⁃BP的邮轮建造仓储物流风险预警研究  

Improved GA⁃BP Based Early Warning of Warehousing and Logistics Risk in Cruise Ship Construction

作　　者：王海燕[1] 王湾 侯华保 邓鹏翥 姜前昆 WANG Haiyan;WANG Wan;HOU Huabao;DENG Pengzhu;JIANG Qiankun(School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan 430070,China)

机构地区：[1]武汉理工大学交通与物流工程学院,湖北武汉430070

出　　处：《中国航海》2022年第3期57-64,共8页Navigation of China

摘　　要：针对邮轮建造物资装卸搬运、检验、入库、储存四个阶段产生的风险评估需求,从“人、机、料、法、环”构建邮轮建造仓储物流风险评价指标体系,结合对应风险评估结果,建立以风险预警误差最小为目标的网络模型,嵌入改良圈算法、正交试验设计改进传统GA⁃BP神经网络算法以改进算法全局搜索能力和收敛速度。结果表明,改进GA⁃BP神经网络算法比GA⁃BP和BP神经网络算法预警误差更小且拥有更好的预警性能。Risks associated with the process of loading/unloading and transfer,inspection,receiving and storage of materials for construction of cruise ships are studied in terms of machinery,materials,law and environment.A risk evaluation index system is devised.The network model is constructed for minimum error in early warning of the risk.The circle modification algorithm and experimental design method are introduced into the basic GA-BP neural network to improve its global searching ability and convergence speed.The design is verified through experiments.

关 键 词：邮轮建造仓储 风险预警指标体系 改进GA⁃BP神经网络模型 正交试验设计 改良圈算法 

分 类 号：U6-9[交通运输工程—船舶与海洋工程]