考虑员工学习效应的MTO/MOS指派模型及算法研究  

Assignment Model In Manual Operating System Making To Order Considering Learning Effect And Algorithm Research

作　　者：于秀丽[1,2] 张毕西[1] 李逸帆[3] 李弘[3] 

机构地区：[1]广东工业大学管理学院,广东广州510520 [2]广东工业大学心理学教学研究所,广东广州510520 [3]广东工业大学应用数学学院,广东广州510520

出　　处：《运筹与管理》2014年第1期226-233,共8页Operations Research and Management Science

基　　金：国家自然科学基金资助项目(71271060);广东省自然科学基金资助项目(S2012010009278);广东省普通高校人文社会科学研究基地重大项目(10JDXM63005)

摘　　要：订单式生产(MTO)人工作业系统(MOS)是我国中小制造企业广泛采用的生产系统模式。在MTO/MOS中,一线员工生产技能对生产绩效具有直接的影响。随着重复操作次数的增多,员工的生产效率不断提高,即产生学习效应。本文考虑员工学习效应来优化一线员工的配置,强调员工初始技能和学习能力的个体差异对完工期的影响,以期缩小理论研究与生产实践之间的差距。本文首先提出一个基于员工初始技能、员工的学习能力、工艺难度和订单批量的员工技能动态变化函数。然后,以最小化完工期为目标,建立一个优化模型。由于员工指派问题属于"完全NP-Hard"问题,为了求解本文所提出模型,本文提出Bootstrap方法对问题进行求解。最后,基于一个算例分析,验证该模型及算法的有效性。MTO （Make-To-Order）and MOS（ Manual Operation System）is the general production system model of the small and medium manufacturing enterprises. The worker＇ skills have an important influence on the perform- ance in such a system. With the growing in times of repetition, learning effect happens and workers＇ efficiencies continuously improve. We propose an optimization model of the front-line workers＇ Assignment considering learning effect, paying full attention to the individual difference in order to narrowing the gap between theory re- searches and production practices. Taking the lead time as the objective function, the model is based on work- ers＇ previous experience, learning ability, technology difficulty and batch size. Due to workers＇ assignment is NP complete problem, the Bootstrap approach to solve the problem is proposed. A case study indicates that the proposed model can provide a promising worker assignment approach and the proposed algorithm is effective for combinational optimization problem.

关 键 词：订单式生产 人工作业系统 学习效应 员工指派 

分 类 号：F2[经济管理—国民经济]