钕铁硼多工序生产计划动态协同优化关键技术研究

The collaborative optimization of the NdFeB multi process production plan is very important for the rational organization of production, quality and production efficiency improvement, material consumption and energy consumption reduction in the dynamic environment. Based on the actual production process of NdFeB, the collaborative management model research of NdFeB multi process production planning dynamic collaborative will be researched. And then, a dynamic collaborative optimization framework for production planning of NdFeB multi process is put forward. And then, a dynamic collaborative optimization framework for production planning of NdFeB multi process based on the concept of layer structure is studied. The optimization strategy which based on “layer cooperative sensing” operator of NdFeB single process dynamic production plan is proposed. The optimization strategy which based on the “environmental situation understanding” operator of production planning dynamic collaborative optimization before and after NdFeB production process is proposed. On this basis, the optimization strategy which based on the “cooperative global re-planning” operator of multi process production planning dynamic collaborative optimization is proposed. Creating a cooperation game model which makes the whole and the partial achieves optimal or approximate optimal by the production plan coordinate of NdFeb production single process and multi process to solve the modeling problem of NdFeB multi process production planning dynamic collaborative. The NdFeB multi process production planning dynamic collaborative optimization method is researched which based on improved augmented lagrangian relaxation method. With the background of the actual production process of NdFeB, the simulation and industrial validation of the above methods are carried out, which provide theoretical methods and application basis for the dynamic collaborative optimization of the multi process production planning of NdFeB.

动态环境下对钕铁硼多工序生产计划进行协同优化，对合理组织生产、提高钕铁硼产品质量和生产效率、降低物耗和能耗具有极为重要的意义。以钕铁硼实际生产过程为背景，研究基于层构理念的钕铁硼多工序生产计划动态协同优化架构。进而提出基于“层协作感知”算子的钕铁硼单工序生产计划动态优化策略和基于“环境态势理解”算子的钕铁硼生产前后工序生产计划动态协同优化策略，在此基础上，提出基于“协同全局重规划”算子的钕铁硼生产多工序生产计划动态协同优化策略。将钕铁硼生产单一工序的生产计划模型与多工序生产计划模型进行协同，建立一个整体与局部均达到最优或较优的合作博弈模型，解决钕铁硼多工序生产计划动态协同建模问题，并研究基于改进增广拉格朗日松弛法的钕铁硼多工序生产计划动态协同优化方法。以钕铁硼实际生产过程为背景，开展上述策略和方法的仿真和工业验证研究，为钕铁硼多阶段生产计划动态协同优化提供理论方法和应用依据。

以钕铁硼、钢铁等实际生产过程为背景，研究动态环境下多工序生产计划协同优化问题，对合理组织生产、提高钕铁硼等产品质量和生产效率、降低物耗和能耗具有极为重要的意义。本基金主要研究基于层构理念的钕铁硼多工序生产计划动态协同优化架构。进而提出钕铁硼生产前后工序生产计划动态协同优化策略，并研究基于改进增广拉格朗日松弛法的多工序生产计划动态协同优化方法。.以钕铁硼磁性材料、机械制造等智能数字化工厂平台的软、硬件构建为基础，重点研究了产品制造过程中各加工单元、仓储物流系统和车间信息化管控系统的多工序分布式协同与优化，提出了基于分布式并行协同的多工序生产计划动态协同优化架构。针对多工序生产计划动态协同优化中的生产订单库存匹配问题，建立了生产订单库存匹配的数学模型，提出基于启发式算法的库存匹配方法。围绕着钕铁硼企业多工序生产计划动态协同优化中的数据获取与处理，给出了一种改进的Slope One协同过滤推荐算法。成型工序是钕铁硼多工序生产计划动态协同优化的关键，其生产计划编制的好坏直接影响钕铁硼多工序生产节奏是否协同，建立了烧结钕铁硼成型日生产计划编制的多目标混合整数规划模型，提出遗传算法和对偶单纯形算法相结合的混合求解方法。动态协同优化中的作业开始时间延迟通常发生在磁性材料、炼钢和连铸等生产过程中，从而改变了初始计划所基于的约束或假设，开展了多工序重计划方法的研究，使系统迅速对此类中断做出反应，对多工序生产计划的动态协同，对企业生产效率的提高，生产成本的降低以及实现有效的材料和能源利用至关重要。提出了一种拉格朗日松弛方法，该方法具有机器产能松弛和操作优先松弛的特点，以解决柔性车间调度问题。围绕磁性材料、炼钢—连铸过程中出现的柔性作业车间生产重调度问题，提出基于人机协同的炉次加工设备调整、基于线性规划的炉次开工和完工时间调整两部分组成的人机协同重调度策略，给出各部分调度算法。.通过本基金的研究成果，开展了上述策略和方法的仿真和工业验证研究，为钕铁硼等多工序生产计划动态协同优化提供了理论方法和实际应用价值，提高企业生产效率，实现节能降耗。