摘要
为求解工业实际生产中无限缓冲与有限缓冲并存的分布式异构柔性流水车间问题,构建了数学规划模型,进而提出一种改进离散人工蜂群(IDABC)算法以最小化最大完工时间。首先,结合机器编码与最早完成时间规则设计基于工厂与工件的二级向量表述调度解,考虑机器选择规则以及阻塞和缓冲的动态修正进行解码,进而混合DNEH启发式法、均衡规则和随机程序提高二级初始种群元胞组质量;然后,对于经雇佣蜂、跟随蜂和侦察蜂3个阶段后产生的新元胞组,设计工厂间插入/交换和工厂内插入3种不同邻域结构以进行变邻域搜索;最后,提出基于优势解的机器搜索策略以避免基于规则的机器分配方法生成单一解的情况。仿真实验测试了不同规模的算例,通过与一些现有启发式算法的对比显示了所提算法获得了更好的近优解且收敛性能表现更佳,随着问题规模的增大,该优势更为明显,这说明了所提算法求解这类问题的有效性与优越性。
To solve the distributed heterogeneous flexible flow shop problem with infinite and limited buffers coexisting in industrial real production,a mathematical planning model was constructed,and then an Improved Discrete Artificial Bee Colony(IDABC)algorithm was proposed to minimize maximum completion time.A two-level vector solution representation based on factories and jobs was designed by combining machine encoding and the earliest completion time rule,and was decoded by considering the machine selection rule and the dynamic corrections of blocking and buffering,and further mixed with DNEH(distributed Nawaz-Enscore-Ham)heuristic,balance rule and random procedure to improve the quality of the two-level cellular groups of initial population.Then,for the new cellular group after three stages of employed bee,following bee and scout bee,three different neighborhood structures including interplant insertion/exchange and intra-plant insertion were designed to perform variable neighborhood search.Finally,the dominant-solution based machine search strategy was developed to avoid the case that the rule-based machine allocation approach might produce a single solution.Simulation experiments tested different scale instances,indicating superior performance of the proposed algorithm in obtaining better near-optimal solutions and exhibiting superior convergence compared to some existing heuristic algorithms.With increasing problem scale,the advantage was more obvious,which demonstrated the effectiveness and superiority of the proposed algorithm for this problem.
作者
轩华
朱林
李冰
XUAN Hua;ZHU Lin;LI Bing(School of Management,Zhengzhou University,Zhengzhou 450001,China)
出处
《计算机集成制造系统》
北大核心
2026年第3期846-861,共16页
Computer Integrated Manufacturing Systems
基金
河南省科技研发计划联合基金资助项目(242103810046)
河南省自然科学基金面上资助项目(252300421945)
国家社会科学基金资助项目(24BJY113)
河南省哲学社会科学规划资助项目(2023BJJ085)。
关键词
分布式异构柔性流水车间
混合缓冲
不相关并行机
改进离散人工蜂群算法
机器分配
distributed heterogeneous flexible flow shop
mixed buffers
unrelated parallel machines
improved discrete artificial bee colony algorithm
machine allocation
