基于关键链的多目标动态生产调度方法研究

发布时间：2018-07-27 18:51

【摘要】：生产调度是制造企业生产管理系统的核心,同时也是影响生产效率的关键因素,因此解决好生产调度问题对于缩短生产工期、提高企业竞争力和客户满意度、控制拖期成本等方面具有重要意义。然而,由于生产过程中的不确定因素导致生产调度成为一个十分具有挑战性的工作,如生产新技术的引入导致更短的生产周期和更快的工艺变化、产品组合的更新、不可预测的设备故障、客户交货期的变化以及瓶颈设备的漂移等等,这些动态因素以及它们之间相互作用给制造企业的生产过程带来了大量复杂的动态不确定性,使得制造企业的生产系统性能经常出现波动,生产调度对生产过程的有效调控变得十分困难。论文受国家自然科学基金面上项目“基于变动性的半导体制造系统性能预测与优化方法研究”(批准号:71671026)资助,充分利用工厂物理学变动性理论对工序加工时间变动性的度量为工序加工时间的缩减提供了参考手段,而以客户重要度和订单优先级等为目标计算工序优先级并基于工序优先级生成初始生产计划实现了生产效益最大化的目标,然后通过启发式算法求解生产计划的关键链,同时结合变动性理论与资源利用度的度量对关键链的缓冲区设置方法的改进更符合实际生产的复杂环境,最终通过软件系统实现了基于关键链的多目标生产调度,在保证企业稳定生产的同时实现了生产效益最大化。首先通过工厂物理学变动性理论对工序进行变动性缓冲并提出改进方法以缩短工序加工时间。其次,通过设计针对混流生产特点的优先级规则,实现综合考虑客户重要度、订单以及工序优先级的生产效益最大化的目标,并通过优先级规则对生产工序进行排序生成初始生产计划,并通过关键链识别方法,在资源约束和优先级约束条件下对初始生产计划进行调整并找出其中的关键链。然后,通过工厂物理学变动性理论对生产过程中工序加工时间变动性及其相互影响进行定性和定量的度量,并结合变动性度量结果与工序资源利用度情况,对调整后的生产调度计划进行缓冲区设置,最终构建基于关键链技术的多目标生产调度模型。在以上研究基础上,通过混合遗传操作的微粒群算法对模型进行仿真验证,最后以某空调换热器生产企业混流生产线为实例,设计原型系统进行实际验证,证明该模型能够对混流生产线进行有效调度,提升企业生产效率和经济效益。
[Abstract]:Production scheduling is the core of the production management system in the manufacturing enterprise, and it is also the key factor affecting the production efficiency. Therefore, it is of great significance to solve the problem of production scheduling to shorten the production time, improve the competitiveness of the enterprise and customer satisfaction, and control the cost of the tardiness. Production scheduling has become a very challenging task, such as the introduction of new production technologies that lead to shorter production cycles and faster process changes, product combination updates, unpredictable equipment failures, changes in customer delivery, and drifting of bottlenecks and so on. These dynamic factors and their interaction to manufacturing The production process of the enterprise brings a lot of complex dynamic uncertainty, which makes the production system performance fluctuates frequently. The production scheduling is very difficult to control the production process effectively. The paper is subject to the National Natural Science Foundation Project "the variable based performance prediction and optimization method of the semiconductor manufacturing system." Research on "research" (approval number: 71671026) support, making full use of the variable theory of plant physics to provide a reference for the reduction of process processing time, and the priority of customer importance and order priority and the initial production plan based on process priority. The goal of maximizing the production efficiency is to solve the key chain of production plan by heuristic algorithm. At the same time, the improvement of the buffer setting method of key chain is more in line with the complex environment of the actual production with the change theory and the measurement of resource utilization. Finally, the multi-objective production scheduling based on the key chain is realized through the software system. In order to ensure the stable production of the enterprise, the production efficiency is maximized. Firstly, the process is buffered by the variable theory of factory physics and the improvement method is proposed to shorten the processing time. Secondly, by designing the priority rules for the mixed flow production characteristics, the customer importance, order and work are taken into consideration. The goal of order priority is to maximize the production efficiency, and order the production process by priority rules to generate the initial production plan, and through the key chain identification method, the initial production plan is adjusted under the resource constraints and priority constraints, and the key chain is found out. Then, through the theory of factory physics variability The qualitative and quantitative measurement of the change of process time and its mutual influence in the process of production, and combining the variable measurement results and the utilization of the process resources, set up the buffer setting for the adjusted production scheduling plan, and finally build a multi-objective production scheduling model based on the key chain technology. Through the hybrid genetic manipulation of particle swarm optimization, the model is simulated and verified. Finally, the mixed flow production line of an air conditioner heat exchanger is used as an example. The prototype system is designed to prove that the model can effectively dispatch the mixed flow production line and improve the production efficiency and economic efficiency of the enterprise.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB497

【参考文献】