劣化效应下混装生产与维护同步作业的多模式鲁棒调度理论及方法

As the production workload and maintenance requirements increase within mixed-model assembly lines, it is extremely urgent to ensure both the maintainability of equipment at the time of production and the production continuity during maintenance. The goal is to synchronize assembly operations with maintenance actions, and reduce capacity waste and economic losses. However, the collision of single production mode and multiple maintenance requirements, and the uncertain deterioration effects on schedule performance and equipment conditions, make the operations synchronization difficult..For operations synchronization of production and maintenance within assembly lines under deterioration effects, this project proposes a mode of operation assignment as the key point, and thereafter explores the theory and methodology of a multi-mode robust scheduling, which includes the following three aspects. .(1).Given maintenance requirements, explore a multi-mode operation assignment and switchover method. The proposed method is to ensure the mode interchangeability, production continuity, economy and high efficiency. It also deduces the preconditions of operations synchronization. .(2).Establish a novel deterministic model for operations synchronization, and develop a multi-objective local search algorithm based on intelligent optimization and knowledge discovery. .(3).Analyze the deterioration trends of schedule performance and equipment conditions, develop a robust equivalent approach for the combination of different probability distribution functions, and probe a multi-mode robust scheduling method via robust proactive-reactive scheduling. As a result, the operations synchronization will be achieved with economy, high efficiency and stability..The project will improve the resource utilization, stabilize the production process and reduce the manufacturing costs. Moreover, it will provide theoretical support to the operations management of production and maintenance for mixed-model assembly lines.

随着混装线上生产负荷与维护需求的增加，需确保生产时的设备可维护性和维护时的生产连续性，实现生产与维护同步作业，减少产能浪费和经济损失。然而，生产模式单一与维护需求多样的冲突、调度性能与设备状态的不确定演化，致使同步难以实现。.本项目围绕劣化效应下混装生产与维护同步作业，以操作分配模式为突破口，研究多模式鲁棒调度理论及方法。具体为：①研究给定维护需求下具有一定互换性、可保证生产连续、且经济高效的操作分配多模式设计与切换方法，推导同步作业前提条件。②构建确定条件下多模式同步调度模型，研制融合智能优化与知识提取的多目标求解算法，实现同步作业。③挖掘调度性能与设备状态的劣化规律，研究多类型分布函数的鲁棒对等转换技术，研制基于鲁棒-预反应调度的多模式鲁棒调度方法，实现经济、高效、稳定的同步作业。.研究成果有利于提高资源利用率、稳定生产秩序、降低制造成本，为混装生产与维护管理提供理论支撑。

随着混装线上生产负荷与维护需求的增加，亟需实现生产与维护同步作业，以确保生产时的设备可维护性和维护时的生产连续性，减少产能浪费和经济损失。然而，生产模式单一与维护需求多样的冲突、调度性能与设备状态的不确定演化，致使同步难以实现。为此，本项目探索了一套基于多模式切换的混装生产与维护同步作业鲁棒调度理论体系。具体为：.①考虑到预防维护和工序分配共有的计划属性，提出面向正常生产和预防维护情形、具有一定互换性的工序分配多模式设计方法，系统性总结了共性约束、目标特性、多任务属性、决策者偏好等特征，为同步作业进行模式准备。.②提出了基于最小生产循环单元、考虑工序分配与投产排序的多模式切换方法，建立了基于机器维护水平的预防维护时刻确定和服从泊松分布的随机故障次数估算模型，构建了面向生产过程波动和设备故障维护的鲁棒运行机制，藉由适应性多模式切换促成同步作业。.③设计了知识驱动的多模式鲁棒调度智能算法，通过问题解析知识引导的优化算子研制、隐性知识驱动的优化目标代理模型构建、面向偏好特性的进化策略制定、融合多域知识的自适应调控机制建立，靶向提升算力效能，精准、快速地编制同步作业方案。.本项目提出了考虑预防维护的多模式工序分配并行设计、基于多模式切换的工序分配与投产排序集成优化、考虑预防维护工位设备故障的混装生产鲁棒调度等方法。所设计的知识驱动智能算法在问题适应性上优势显著，基于多模式切换的同步作业将装配线效率提升了1.8%。本项目在国内外期刊发表论文54篇，其中SCI源刊34篇，JCR分区一区22篇，EI中文源刊7篇。撰写专著1本，申请发明专利8项，授权1项，获批软著1项。培养青年教师3名、出站博士后1名、毕业博士2名及毕业硕士13名。.本项目成果可为混装生产与维护管理提供理论支撑，协助企业进行面向预防维护情形的多模式设计和切换，促成劣化效应下混装生产与维护同步作业，达到提高资源利用率、稳定生产秩序、降低制造成本的目标。