面向船舶管子加工分组调度的矩阵实数编码遗传算法  

A Matrix Real-Coded Genetic Algorithm for Ship Pipe Workshop Group Scheduling

作　　者：贺苗 管在林[1] 侯国祥[2] 鲁佳俊 HE Miao;GUAN Zailin;HOU Guoxiang;LU Jiajun(Huazhong University of Science and Technology,School of Mechanical Science and Engineering,School of Naval Architecture and Ocean Engineering,Wuhan 430074,China;Huazhong University of Science and Technology,School of Naval Architecture and Ocean Engineering,Wuhan 430074,China;Shenzhen Tencent Computer System Co.,Ltd.,Shenzhen 518054,Guangdong,China)

机构地区：[1]华中科技大学机械科学与工程学院,武汉430074 [2]华中科技大学船舶与海洋工程学院,武汉430074 [3]深圳市腾讯计算机系统有限公司,广东深圳518054

出　　处：《船舶工程》2022年第4期140-145,共6页Ship Engineering

基　　金：国家自然科学基金资助项目(51775216)。

摘　　要：分析了船舶管子加工车间任务重且加工进度会严重影响船舶制造进度的问题,结合管子加工成组技术和管件族制造法,建立了船舶管子的混合流水车间多目标分组加工模型。利用矩阵实数编码遗传算法(MRCGA)能对该加工模型建立有效的分组调度方案。该算法能保证在迭代过程中子代个体基因的交叉、变异的合法性和可行性,相较于遗传算法(GA)具有较好的寻优能力和稳定性。结果表明,调度目标最大完工时间降低了46.98%,很大程度上提高了船舶管子加工分组调度车间的生产效率。The problem that the processing progress of ship pipe processing workshop would seriously affect the ship manufacturing progress is analyzed.Considering the pipe group technology and pipe fitting family manufacturing,and the multi-objective group processing model of the mixed flow workshop of the ship pipe is established.A matrix real-coded genetic algorithm for this group scheduling plan is proposed,which could ensure the feasibility and validity of the individual gene in the process of crossing and mutation.Compared with the original genetic algorithm(GA),MRCGA shows better target optimization and robustness ability,and the maximum completion time of the scheduling target reduces by 46.98% and greatly improves the production efficiency of the ship pipe processing group scheduling workshop.

关 键 词：管子加工 成组技术 混合流水车间 矩阵实数编码遗传算法 遗传算法 

分 类 号：U671.3[交通运输工程—船舶及航道工程]