不确定环境下模具制造车间前摄与反应式调度方法研究

发布时间：2018-08-22 15:31

【摘要】：模具是装备制造业的重要组成部分,模具生产水平的高低,已经成为衡量一个国家产品制造水平高低的重要标志。虽然企业不断引进高端数控加工装备,但在高端精密模具制造技术水平方面与发达工业国家仍存在较大的差距。加工精度不高、制造周期普遍过长、订单交货拖期严重是制约我国模具企业国际竞争力的三大关键因素。究其原因,主要是缺乏一种适应模具制造特点的有效生产控制方法。面向资源的单件工程订货型生产方式使得模具制造过程中充满着大量各种不确定性因素,这给模具车间制订合理可行的作业计划带来了一定的困难。而在基准方案执行当中,突发事件的频繁冲击将使得初始制订的基准调度方案可执行性不强。因此,必须采取有效的预先防范手段,以及合理的重调度方法与控制策略,确保生产计划按时按质完成。本文在国家自然科学基金(50675039、50875051)和国家863计划资助项目(2006AA04Z132)的联合资助下研究不确定环境下的模具制造车间前摄与反应式调度问题。根据模具制造单元的特点,对不确定因素进行了分析及建模,详细探讨了不确定环境下模具制造车间调度决策机制,重调度驱动机制,建立了相关调度模型和求解算法。模具柔性流水车间前摄与反应式调度决策支持系统的实施,将有助于提高企业生产计划与控制水平,增强企业市场竞争力。本文的主要工作包含以下几个方面： 1、提出了以主关键件为控制指针的模具生产组织结构,建立了不确定环境下模具柔性流水车间模型,并针对三种主要不确定因素建立了数学模型,最后提出了模具制造车间前摄与反应式调度框架； 2、基于工时离散概率模型,分析工时组合情况。在此基础上,提出了以最大化稳定度为优化目标的前摄性调度模型,提出了一种变宽集束搜索求解算法,通过与定宽集束搜索算法的对比分析,表明在求解质量和计算耗时方面均具有优势； 3、定义了工序延期概念,提出了工序累积延期的计算方法。针对模具柔性流水车间存在隐性不确定因素和显性不确定因素的影响,提出了基于累积延期和事件触发混合驱动的反应式调度机制,建立了基于Plant Simulation的仿真模型,分析了混合驱动机制的属性,并与周期性重调度、事件驱动重调度进行了对比分析,结果表明该混合驱动机制能获得较低的反应式调度频率,并且在急单插入频率较高的情况下还能保持良好的调度性能； 4、基于上述调度衡量指标,提出了模具柔性流水车间两阶段反应式调度模型,包含局部修补调度和反应式调度两个阶段。当混合驱动机制触发反应式调度,首先执行反应式调度模型的第一阶段：局部修补调度。在局部修补阶段,将稳定性指标作为首要考虑因素,工件在机器上的加工序列不变,对于受影响工序在时间轴上延后其开始加工时间。采用约束传播树表示,提出了基于约束传播树的受影响工序局部修补算法。将所提算法与传统右移(Right Shift)重调度方法比较,分析了在机器故障和急单插入情况下局部修补算法具有一定优越性； 5、通过局部修补算法求解得到调度方案后,通过设定一个评价模型,包括拖期情况分析,拖期惩罚成本分析,以辅助调度人员进行决策,由计划调度员根据情况交互启动第二阶段。在第二阶段的反应式调度：双目标反应式调度。构建了兼具有效性和稳定性的双目标反应式调度模型,将基于欧几里得拥挤距离的非支配排序与基于聚类算法的精英归档策略相结合,提出一种改进的多目标遗传算法进行反应式调度模型求解,并与NSGA-Ⅱ算法进行比较,仿真结果显示,由于采用了基于欧几里德拥挤距离的非支配排序,MMOGA算法求得的非劣解分布比NSGA-Ⅱ算法更均匀,更靠近近似Pareto最优前端； 6、分析了企业基本业务流程和生产计划递阶协调控制模式,设计了系统的功能模块和数据库,开发了一套模具柔性流水车间前摄与反应式调度决策支持系统,用于辅助生产总调协调整个生产车间活动,及时对突发事件作出反应,应用效果良好。
[Abstract]:Die is an important part of equipment manufacturing industry. The level of mold production has become an important symbol to measure the level of a country's product manufacturing. Not high, manufacturing cycle is generally too long, the order delivery delay is a serious restriction on the international competitiveness of China's mold enterprises three key factors. The main reason is the lack of an effective production control method to adapt to the characteristics of mold manufacturing.
The resource-oriented ordering mode of one-piece engineering makes the mold manufacturing process full of a lot of uncertainties, which makes it difficult for the mold workshop to make a reasonable and feasible operation plan. Therefore, it is necessary to take effective precautionary measures, as well as reasonable rescheduling methods and control strategies to ensure that the production plan is completed on time and on time.
In this paper, the problem of forward and reactive scheduling in mold manufacturing workshop under uncertain environment is studied under the joint support of National Natural Science Foundation (50675039, 50875051) and National 863 Program (2006 AA04Z132). According to the characteristics of mold manufacturing cell, the uncertain factors are analyzed and modeled, and the uncertain environment is discussed in detail. The decision-making mechanism and the rescheduling driving mechanism of the mold manufacturing workshop are established, and the relevant scheduling model and solving algorithm are established.
The main work of this paper includes the following aspects:
1. The mould production organization structure with the main key parts as the control pointer is put forward, and the flexible flow shop model of the mould under the uncertain environment is established. The mathematical models are established for three main uncertainties. Finally, the forward and reactive scheduling framework of the mould manufacturing workshop is proposed.
2. Based on the discrete probability model of man-hour, the time combination is analyzed. On this basis, a proactive scheduling model with maximum stability as the optimization objective is proposed, and a variable-width clustering search algorithm is proposed.
3. The concept of process delay is defined, and the calculation method of cumulative process delay is proposed. Aiming at the influence of implicit and explicit uncertainties in flexible flow shop, a reactive scheduling mechanism based on cumulative delay and event-triggered hybrid drive is proposed. A simulation model based on Plant Simulation is established and analyzed. Compared with periodic rescheduling and event-driven rescheduling, the results show that the hybrid mechanism can achieve lower reactive scheduling frequency, and can maintain good scheduling performance in the case of high insertion frequency.
4. Based on the above scheduling metrics, a two-stage reactive scheduling model for flexible flow shop is proposed, which includes two stages: local repair scheduling and reactive scheduling. As the primary consideration, the machining sequence of the workpiece on the machine is unchanged, and the processing time of the affected process is delayed on the time axis. The constrained propagation tree is used to represent the local repair algorithm of the affected process. The local repair algorithm under the condition of machine failure and urgent single insertion has certain advantages.
5. After the scheduling scheme is solved by local patching algorithm, an evaluation model is set up to assist the dispatcher in decision-making, including the analysis of tardiness and the analysis of tardiness penalty cost. A two-objective reactive scheduling model with validity and stability is proposed. Combining non-dominated scheduling based on Euclidean congestion distance with elite archiving strategy based on clustering algorithm, an improved multi-objective genetic algorithm is proposed to solve the reactive scheduling model. The simulation results are compared with the NSGA-II algorithm. The simulation results show that the proposed algorithm is effective and stable. By using non-dominated sorting based on Euclidean congestion distance, the non-inferior solutions obtained by MMOGA algorithm are more uniform than those obtained by NSGA-II algorithm and closer to the approximate Pareto optimal front-end.
6. The hierarchical coordination control mode of enterprise basic business process and production plan is analyzed, the function modules and database of the system are designed, and a set of decision support system for flexible flow shop scheduling is developed, which is used to assist production production coordination to coordinate the activities of the whole production shop, respond to emergencies in time and is effective in application. Good fruit.
【学位授予单位】：广东工业大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TP301.6;TH186

【相似文献】