Abstract: To solve the Capacity Configuration Problem (CCP) of large-scale finite buffer production line,an optimal model with throughput constraint was built to minimize the equipment investment costs.Aiming at the problem that the finite buffer production line cannot be expressed by closed expression,a decomposition-coordination optimization method was put forward based on large-scale theory to overcome dimension explosion problem.This method decoupled original system into several subsystems,and the coordinator was used to update coordination variables for optimizing the equipment configuration.The branch and bound method was used to search optimal solutions of subsystem,and the system performance indexes was calculated with queuing network model.Several numerical examples with different parameter settings were analyzed,and the equipment configuration of examples was brought into simulation model to verify the effectiveness.

Key words: large-scale, queuing network, decomposition-coordination, capacity configuration, branch and bound

摘要： 针对大型有限缓冲生产线的设备配置问题,以最小化设备投资成本为目标,建立具有系统产出率约束的优化模型。由于有限缓冲生产线无法用封闭形式表达式描述,提出一种基于大系统理论的分解协调优化方法解决维数爆炸问题,该方法将原系统解耦为多个子系统,利用协调器更新协调变量优化设备配置。使用分支定界算法搜索子系统的最优解,并使用排队网近似求解系统性能指标。通过算例分析不同参数设定下使用所提方法得到的优化结果,并将其带入仿真模型,验证了所提方法的有效性。

关键词: 大系统, 排队网, 分解&mdash, 协调, 设备配置, 分支定界算法