存在返工的定点装配车间生产计划与调度的集成优化研究及应用

The rework of the non-conforming products does exist in the manufacturing system. Especially in the fixed-position assembly workshop, the rework characterized by uncertainty leads directly to the production schedule infeasible, and furthermore affects the execution of production plan. Therefore, in this project, focused on the fixed-position assembly workshop, the integrated optimization of production planning and scheduling under circumstances of parts and products rework will be investigated. In order to reduce the influence of rework disturbance on production management, the methods of delaying rework process, adding the detecting unit and setting the dedicated process time of rework job are proposed. The models of the integrated production planning and scheduling under circumstances of parts and products rework in the fixed-position assembly workshop are established, and the constraints of available starting time of operations and assembly groups are proved. By studying the structure and computational complexity of the integrated model, fast and effective solution techniques are developed. By applying the above-mentioned results, the integrated optimization systems of production planning and scheduling for fixed-position assembly workshop with rework are designed and developed, and a large number of examples are operated and analyzed to inspect the feasibility of the algorithms. Furthermore, the theoretical system of the integrated optimization of production planning and scheduling under uncertainty is improved. By cooperating with Jincheng Software Limited Corporation, Nanjing, the above-mentioned system will be integrated to the ERP (enterprise resource planning) system and MES (manufacturing execution system) of the corporation as the planning and scheduling subsystem, and applied to practical production to generate significant economic benefits.

不合格品返工情况在制造系统中客观存在，尤其在定点装配车间，返工不确定性直接导致预先制定的调度不可行，进而影响生产计划执行。因此，本项目以定点装配车间为对象，研究零部件返工及产品返工情况下生产计划与调度集成优化问题，提出将返工延后处理、增设单元检测点、设置返工处理时间的方法，以降低返工对生产管理的干扰，建立零部件返工及产品返工情况下定点装配车间生产计划与调度集成优化模型，并证明工件与装配班组的开工时间应满足的约束条件；通过对模型的结构复杂性与计算复杂性的研究，提出解决该类问题快速而有效的求解技术。应用上述成果，研发存在返工的定点装配车间生产计划与调度的集成优化系统，对大量算例进行运算和结果分析以检验上述算法的有效性，进而完善补充不确定条件下生产计划与调度的集成优化理论。与南京金城软件有限公司合作，将上述系统集成到该公司ERP、MES系统中作为计划与调度子系统，并应用于实际，产生显著经济效益。

根据航空发动机装配车间特点凝练出存在返工的定点装配车间特征，并提出将存在的部件返工、不合格产品返工、售后产品返工三种不确定情况下的定点装配车间作为项目研究对象，研究其生产计划与调度集成优化问题。.针对产品返工导致的不确定可重入情况，证明定点装配车间中工件及装配班组开工时间应满足的约束条件，并据此建立不确定可重入定点装配车间随机生产计划与调度集成优化问题的双层随机规划模型，证明可行调度解的递推公式，提出具有双层结构的交替迭代式改进遗传算法求解双层模型。.针对装配过程中部件不合格导致的返工不确定情况，提出在产品装配过程中增设单元检测点，建立基于场景分析的定点装配车间生产计划与调度集成优化模型，证明调度解在所有场景下的可行性，推导出可行调度解在各场景下装配班组完工时间计算公式，并提出双层遗传-文化基因算法求解模型。.针对不合格产品返工不确定情况，提出返工延后处理的方法，建立其混合整数规划模型以求取返工延后处理的粗生产计划，然后根据每周期实际返工不合格品数量，建立定点装配车间生产计划与调度集成优化模型，并提出求解该问题的遗传-文化基因算法，证明文化基因算法的局部搜索条件。.针对来自于车间外部的不合格售后产品返厂维修不确定情况，提出在每周期设置恰当的空闲时间进行返工维修处理，采用混合模式下两阶段处理方法：求解预留的空闲时间，构建带空闲时间的生产计划与预测调度集成优化模型，提出双层交替迭代遗传-候鸟迁徙优化算法求解，证明算法的邻域搜索可行性及启发式搜索规则；对执行预测调度时产生的返厂维修情况，利用空闲时间进行反应式调度，最终实现计划调度同时优化。.研发存在返工航空发动机装配车间生产计划与调度的集成优化管理系统，并由南京金城智能科技有限公司负责推广应用，有望产生显著经济效益。.本项目的研究不仅进一步完善了定点装配车间生产计划与调度集成优化理论，而且成果可用于多种类型的定点装配车间，有良好的应用前景。