【作者】 王云；

【导师】 谭建荣； 冯毅雄；

【作者基本信息】 浙江大学 ， 机械设计及理论， 2011， 博士

【摘要】 随着制造业的迅速发展,制造执行系统正成为制造企业利用制造资源、提高快速响应能力的关键。本文研究了面向云制造的制造执行系统中零件实例检索方法、柔性作业车间的资源管理和配置、生产计划和调度,提出了基于CBR的零件实例检索方法、基于粒子群优化算法的柔性制造单元构建方法、基于JIT的加工和装配生产计划调度集成方法和基于SPEA的柔性作业车间分批调度多目标优化方法；并结合企业具体应用将上述理论应用于机床装备的生产管理中,取得了良好的效果。论文的主要工作如下：第一章综述了制造执行系统的层次结构、功能体系和研究现状,阐述了制造业的现状、发展趋势、相关研究技术和云制造特点。在分析现有制造执行系统在云制造环境下不足的基础上给出了论文主要研究内容及技术路线。第二章提出基于CBR的零件检索方法。阐述了零件实例检索研究现状、零件特征的分类原则、识别方法和技术。分析了零件形状特征间的可比性和相似度,提出了复合特征层、基本特征层和简单特征层零件的形状和几何相似性计算方法和基于CBR的多层次零件检索方法,分三个阶段计算零件间的相似度。第三章提出基于粒子群优化算法的柔性制造单元构建方法。分析了不同的制造环境下制造资源构建方法,构建了在云制造环境下多加工车间生产资源布局的柔性制造单元模型,该模型融合物理制造单元和逻辑制造单元的优点,以制造成本最低为目标,并综合考虑了制造资源布局、加工路径选择以及加工批量设置等因素,给出了一种基于粒子群优化算法的制造单元构建方法。第四章提出了基于JIT的加工和装配生产计划调度集成方法。分析了生产计划和调度研究现状,阐述了生产计划和调度的约束、特点、核心问题和方法。针对订单生产型企业JIT准时供货的要求,以交货期为目标,考虑加工和装配工艺约束、加工和装配设备约束,建立了加工和装配计划调度集成模型,该模型为两层混合规划模型,运用了遗传算法和启发式规则,提出了混合启发算法,对产品制造过程进行分解,先通过装配计划计算得出各工件的交货期,然后根据工件的交货期,确定了工件的加工计划。第五章提出了柔性作业车间分批调度多目标优化方法。分析了作业车间调度的研究现状,构建了以制造工期、拖期惩罚、加工成本、批次数量和机器总负荷为目标函数的柔性作业车间多目标调度模型,考虑制造工期、加工成本、拖期惩罚值、批次数量、机器总负荷五个优化目标,应用改进的强度Pareto进化算法(SPEA)求解。该方法将工件分割成具有柔性数量的多个批次,同时使各批次的工艺路线选取及加工顺序得到优化。第六章结合企业项目需求和本文研究成果,开发了面向云制造的机床装备生产制造执行系统("ME-MES"),以实际应用验证了所提理论和方法的正确性和可行性。第七章总结了本文的主要研究内容和成果,并展望了今后的研究方向。更多还原

【Abstract】 With the rapid development of manufacturing industry, manufacturing execution system has become one of the most important keys of using manufacturing resources and improving the ability of rapid response to market. In the cloudy manufacturing environment, part case retrieval method, manufacturing resources management and allocation, production planning and scheudling are studied. Part case retrieval method based on CBR, flexible manufacturing cell formation approach based on PSO algorithm, integrated method of machining and assembly planning and scheduling based on JIT and multi-objective optimization method of flexible job-shop lot-splitting scheduling based on SPEA are put forward and applied in the machine equipment production, and good effects are obtained.The main works are as follows:The first chapter reviews the hierarchy, fuctions and research status of manufacturing execution system, expounds the status, development trend and technology of manufacturing industry and the characteristic of cloudy manufacturing. By analyzing the shortage of existing manufacturing execution system in cloudy manufacturing environment, the research contents and technical route are presented.The chapter 2 puts forwards part case retrieval method based on CBR and expounds the research status of part case retrieval, the classifying principle of part features, recognition method and technology of part features. By analyzing the comparability and similarity of parts form features, the methond of similarity calculation of parts form and geometry in mulriple feature layer, basic characteristic layer and simple feature layer is put forward. A multi-level part case retrieval method is proposed based on CBR. The similarity is calculated with three stages analysis between a new part and old parts in the case library.The chapter 3 puts forward flexible manufacturing cell formation approach based on PSO algorithm. Manufacturing resource formation methods in different manufacturing environment are analyzed. Flexible manufacturing cell model integrating with the advantages of physical manufacturing cell (PMC) and logical manufacturing cell (LMC) in the open shops in cloudy manufacturing with changeable resources layout is set up. Mainly considsing three factors: machine layout, routing and production lot sizing, based on PSO algorithm, manufacturing cell formation algorithm is proposed to minimize total manufacturing cost.The chapter 4 puts forward integrated method of machining and assembly planning and scheduling based on JIT. The research status of production planning and scheduling is analyzed, the restriction, characteristic, key problem and method of production planning and scheduling are expounded. An integrated model of machining and assembly planning and scheduling is constructed in order to deal with manufacturing in a JIT production system of MTO enterprise for delivery just-in-time. Function objective of the proposed model is to guarantee the due dates of production. Restraint functions include machining capacity, assembly capacity, machining and assembly procedures restraints. Feasible planning can be set down considering these restraints. To build up two level mixed programming model, a hybrid heuristic algorithm based on genetic algorithm and heuristics rules is put forward. Integrated model is used in a shop of a MTO company. By decomposing manufacturing planning, due dates of each parts can be achieved by assembly planning and machining sequence of parts can be ascertained according to due dates of each parts.The chapter 5 puts forward multi-objective optimization method of flexible job-shop lot-splitting scheduling. The research status of job shop scheduling is analyzed and the multi-objective FJSP optimization model is set up, concerned with makespan, tardiness penalty, manufacturing cost, count of batch and total workload. The optimal solutions are obtained by using improved strength Pareto evolutionary algorithm (SPEA). The jobs are splitted into flexible size batch, and the batch routing and sequencing are simultaneously optimized by the method.The chapter 6 develops the manufacturing execution system of machine equipment ("ME-MES") in cloudy manufacturing with the project requirement and research results, which shows the validity and feasibility of the new theory and method proposed in the dissertation.The chapter 7 summarizes the key research contents and achievements, and presents conclusions along with recommendations for future research.更多还原